Known Bugs in Khoros 2.2 : 1-July-97
This page contains the complete description of all of the known bug
in Khoros 2.2 and is organized by issue number. Along with the FAQs,
you can check this page to see if your bug is already known and being
worked on for future releases.
I...-.18
Bug Description: Editimage does not correctly create greyscale images
Toolbox: ENVISION
Object: editimage
Full bug descripton:
Verified by Paula on 21-Jan-1997
Follow Danielle's example below...
For example:
% editimage -i ~kadmin/training/k2.1_training/FIGURES/color/xwd/ap1.xwd
Go to "Colormaps" pane of "Options" subform
Click on "Greyscale".
Look at the image; it's way dark!!
For contrast purposes,
% xv ~kadmin/training/k2.1_training/FIGURES/color/xwd/ap1.xwd
Save as greyscale; redisplay
Look at the image; it's a nice greyscale version.
================================================================
I...-.25
Bug Description: kformats not working as a shell script inside cantata
Toolbox: DATAMANIP
Object: kformats
Full bug descripton:
This bug was reported earlier in K2.1.
In the DIP toolbox, there is an script object to convert
KDF to GIF files which call kformats and pbmplus.
Apparently kformats -i xxx -o yyy
is not working inside a shell script inside cantata.
Outside cantata, it works.
The error that output is:
DATAMANIP
kformats
ktransport
kfflush
Category: Bad file number
Only able to flush -1 out of 34 bytes.
NOTE: (DA & MY, May 7, 1997)
a workaround is to have the
user unsetenv the KHOROS_IPC environment variable inside their
shell script. This should be in the FAQ for Khoros 2.1 if it
is not already.
================================================================
I...-.28
Bug Description: Text entry selections on Render have no effect
Toolbox: DESIGN
Object: xvobjects
Full bug descripton:
Verified by Paula on 16-Jan-1997
If the user changes the text value for the background color and hits return
nothing happens. If the user changes the text value and hits the "Render"
button, the background color does change, but the scrollbar does not.
----------------------------------------------------------
If you change the text values of the background colors on the Render
pane, nothing happens, with either hitting return or the Render button. The
sliders do work.
================================================================
I...-.38
Bug Description: feedback loop won't reschedule
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Original Rpt: To reproduce, run cantata with the workspace
~mirage/testimages/data/workspaces/loops/feedback.wk
1 - Hit the run button; it schedules correctly.
2 - Hit the stop button; gull image goes away.
3 - Hit reset; workspace resets
4 - Hit the run button; it refuses to run again.
Verified by PB on 15-Jan-97
Note (DA & MY, May 7, 1997)
Workaround is to interpose an operator glyph between the input glyph
and the glyph that does the feeback loop. For example, absolute
value is a no-op if the image values are all > 0, or you could
use an add glyph to add a constant of 0.
this will only happen
if you connect an input glyph to the loop, not if you connect
an operator glyph to the loop.
================================================================
I...-.87
Bug Description: editimage quits updating after many tries
Toolbox: ENVISION
Object: editimage
Full bug descripton:
After 20-40 changes to a "kdf" image file, "editimage"
no longer automatically updates. Pressing the "Files"
button and then hitting on the "Input Image"
text line crashes the application. A window briefly
(<<1 sec) pops up and states something like "too many
files .....". In addition the following is written to
the screen:
X Error: BadDrawable
Request Major code 62 ()
ResourceID 0x272e2f6f
Error Serial #16072
Current Serial #16116
Also, every 10-20 updates the "editimage" image window
turns black. Hitting on the "Input Image" line
properly displays the image .... ????.
================================================================
I...-.92
Bug Description: scroll bar in the Commands window not accurate
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
1-21-96 -- squeedle
When the make commands are run (i.e. any commands w/ command lines longer than
the text window), the horiz scrollbar shifts to the right, when in fact the
left side is all the way left. The vertical scrollbar is always the same
size, regardless of how many lines of text there are, i.e., if you scroll
all the way down, the bar scrolls way past the actual last line of text.
Original Rpt:
Does not reflect actual position of text window in text stream,
vertically or horizontally. Again, this makes it hard to see
progress of tools such as compiler.
Notes : This should be logged in the approprate widget library
in design
================================================================
I...-119
Bug Description: temporary files not being cleaned up in /var/tmp
Toolbox: BOOTSTRAP
Object: kutils
Full bug descripton:
Original Rpt: Cantata, dataservices operators, and data
testsuites are still leaving temporary files
behind. Not sure why, but this really needs
to be fixed. Eventually cantata starts failing
because there is no more temporary space.
Notes : Need to hack kutils after bronze to warn if you
ever exit a program without unlinking a temporary
file. At least this will allow us to track down
which programs under which conditions are leaving
these temporary files around.
================================================================
I...-122
Bug Description: Puticon does not not have an exit or quit button
Toolbox: ENVISION
Object: NS
Full bug descripton:
Machines tested -
2.0.2 (enchilada)
Original Rpt: Run: % puticon -i image:ball and notice that there
is not quit button. It might be nice to make the
minimum size a little larger such that a quit button
can be displayed for consistency with other putdata
pane objects
================================================================
I...-126
Bug Description: Can't switch between areas when Area menuform is open
Toolbox: ENVISION
Object: xprism
Full bug descripton:
Original Rpt:
When the Area Menuform is open, you cannot switch
between areas. The menuform refers only to the area
you brought it up for, not for the currently
highlighted area.
in xprism, when the Area Menuform is open, you cannot
switch between areas - the highlighted area is indeed
the one you select, but when you change something it
changes only the area you first opened the menuform for.
reproduce by: plot 2 fcns, then go to Options and select
an Area, then open the Area menuform. while this window
is open, select a the other area and change something.
================================================================
I...-127
Bug Description: Entering 3D function in the 2D window may cause crash
Toolbox: ENVISION
Object: xprism
Full bug descripton:
Original Rpt:
Entering 3D function in the 2D window may cause
a crash on some architectures
purify xprism; open the Plot window in xprism and put
in a function of x & y. watch the UMR's in kexpr
================================================================
I...-129
Bug Description: 3D plots often redraw themselves 2 or 3 times
Toolbox: DESIGN
Object: xvisual
Full bug descripton:
Original Rpt: 3D plots tend to redraw themselves 2 or 3
times on exposure events.
We already tried to fix this once (April 16) by
putting this line in the Plot3D Refresh() routine:
if (!xvw_check_visible(xvw_object(widget))) return;
However, it was later discovered that this causes
putplot3 and several of the example programs not to
display their plots initially at all; plots were
refreshed only when the plot type or other attribute
was changed via the menuform.
================================================================
I...-134
Bug Description: Color is not always correct using ReadWrite Visual
Toolbox: DESIGN
Object: xvisual
Full bug descripton:
Original Rpt: Color is not always set properly with a ReadWrite Visual
when the display is out of colors (POST 2.1).
Need to rewrite the colorcell object completely:
Modification of colorcell to be completely controlled by
the RGB values with optional index specification rather
than the current implementation, which has the opposite
case: required index specification and optional RGB value
specification
================================================================
I...-139
Bug Description: Should you be able to change map columns on RGB image?
Toolbox: DESIGN
Object: xvisual
Full bug descripton:
Original Rpt: open test:RGB:ubyte_reg_range and select options, then
select the Display pane. edit fcns dictating colormap
Seems to always do the same
thing no matter what fcn you put in - acts like you put a
sequence of r, g and blue filters on it until it turns
black
================================================================
I...-143
Bug Description: License screen pops up larger than the screen area
Toolbox: DESIGN
Object: xvregister
Full bug descripton:
Original Rpt: After filling out the register program, the license
window pops up and is off the screen at the top and the
bottom. I can't even get to the resize icon with twm to
change it. I would prefer a normal sized window that I
scroll to view all of it.
================================================================
I...-144
Bug Description: xvregister has too many popups when running
Toolbox: DESIGN
Object: xvregister
Full bug descripton:
Machines tested - Any platform (bronze2)
Original Rpt: When running xvregister, there are 4 pop-ups that
are displayed. These should be combined somehow.
================================================================
I...-155
Bug Description: Class contents not always updated properly
Toolbox: DESIGN
Object: spectrum
Full bug descripton:
Statement of problem...
Class contents are not updated when you change the Class name or when you
Add Cluster, but are when you Transfer Cluster or Delete Cluster. It does
not matter if you first close the Class Contents pane & then make the
changes.
================================================================
I...-157
Bug Description: Something wrong with imagine:sevilleta files
Toolbox: GIS
Object: import_img
Full bug descripton:
Statement of problem...
The Import IMAGINE glyph gets stuck in Run mode when using either of the
imagine:sevilleta8* files in the Aliases. It does not get stuck when
using the imagine:usa picture as input, so I am assuming the two files
are the problem.
================================================================
I...-170
Bug Description: Animate pretends to write to unwriteable directory.
Toolbox: ENVISION
Object: animate
Full bug descripton:
######################################################################
Bug Report
Animate pretends to write out a file to a directory even when it
does not have write permissions to that directory. That is, there
is no message saying you can't write there.
================================================================
I...-177
Bug Description: Cantata remote hosts has unpredictable behavior
Toolbox: DESIGN
Object: cantata
Full bug descripton:
######################################################################
Bug Report
Remote hosts behavior is unpredictable and I can't seem to reproduce
the behavior, for instance, trying to start daemons in the remote hosts
list with no .rhosts file in the kpro account produced an error msg as
expected.
However, when a .rhosts was created, the following happened:
trying to start a daemon on glacier (which was not in the .rhosts
file) made cantata hang.
Trying to start the daemon on tucumcari, which has no binaries for
that purpose, produced the expected error message.
But rsh's were started (we checked to see if phantomd processes were
running) on mohawk & enchilada, neither of which are in the .rhosts
file.
we always used the same format,
[machine name]:kpro:rsh to start the daemon
================================================================
I...-182
Bug Description: Fcns dictating colormap won't change back to original
Toolbox: ENVISION
Object: animate
Full bug descripton:
Fcns dictating colormap will not change back to originalC
######################################################################
Bug Report
After changing the function dictating the colormap in the Display Pane
of animate, you cannot get back to the original colormap, even by
specifying M0 as Red, M1 as Green, M2 as Blue, or by erasing the
function or functions & pressing Return.
MORE: actually you can get the original colormap, but you can't use lowercase
m's. --squeedle 22May97
================================================================
I...-183
Bug Description: Animate with -alloc 2 produces black images
Toolbox: ENVISION
Object: animate
Full bug descripton:
######################################################################
Bug Report
Animate with Read/Write colormap makes black or mostly black image.
Look at animate -sequence:baby -alloc 2
================================================================
I...-185
Bug Description: unclear what the bootstrap examples are for
Toolbox: BOOTSTRAP
Object: Examples
Full bug descripton:
######################################################################
Bug Report
Both of these example programs do nothing and are examples of printf
statements, WOW!. If they do something, there should be a readme or
better yet, some output stating what is happening or what to expect.
Also, 2.datatype example has several repeated lines.
================================================================
I...-187
Bug Description: Can't change output file if there isn't write permssion
Toolbox: ENVISION
Object: editcmap
Full bug descripton:
Paula Barth 5/21/97
Verified this on fresca. Do the following:
1) % editcmap -i image:mandril -pseudo -priv 0 -alloc 1 -x -1 -y -1 -width 512
-height 512 -update 1
2) Click the Save button
3) Type /research and click OK
4) You will get two error messages, saying that you can't write to /research.
Click OK on both of them
5) Click on the Save button again (ostensibly to save to a different file).
6) You get two more error messages saying you can't write to /research
--------------------------------------------------------------------------------
Cannot change output file if there is no write permission
######################################################################
Bug Report
If, while in editcmap, you enter a filename for output to a dir which
does not have write permissions, you can never change it to something
else; you just keep getting messages saying you can't write there.
================================================================
I...-192
Bug Description: Ran out of colors unexpectedly when on 24 bit display
Toolbox: ENVISION
Object: cantata
Full bug descripton:
######################################################################
Bug Report
Reporting this per Mark's request: here I am happily running Cantata
on the 24 bit display, & this workspace calls up xprism & putdata.
When I select different colors using the Options pane, it says I am
out of colors - this is because these apps are running at 8 bit depth.
So Mark says that the programs should maybe be designed to always
take advantage of the display, like editimage & animate.
================================================================
I...-199
Bug Description: button allocation rules get confused
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
######################################################################
Bug Report
If you shrink the vertical dimension down enough to where the left and
right buttons are just bars, you can get the right button to get into a
state where it is flashing between creating a full button and just a bar.
================================================================
I...-202
Bug Description: Continuous Run not smart about which glyph to restart
Toolbox: ENVISION
Object: cantata
Full bug descripton:
######################################################################
Bug Report
If you are in Continuous Run mode & you change a file transport
type, the glyph right after the pipe runs, rather than the one
right before, which means 2 things:
The new transport type is not used for that execution (I think?).
If the pipe is not green (temp file has already been routed thru),
this produces an error.
================================================================
I...-212
Bug Description: Don't need 3 err mesgs for bad input value
Toolbox: ENVISION
Object: edithresh
Full bug descripton:
######################################################################
Bug Report
When you enter 1 bad value for the limits in Editcmap, you get 3
error messages. (I entered "bleh")
================================================================
I...-213
Bug Description: Cannot return to display of ROI with no output file
Toolbox: ENVISION
Object: editimage
Full bug descripton:
######################################################################
Bug Report
>From editimage, once you specify an output file for an extracted ROI, you
cannot then erase the name of the file, and just have the ROI displayed
but not written out.
================================================================
I...-214
Bug Description: Got an error box running getimage
Toolbox: ENVISION
Object: getimage
Full bug descripton:
######################################################################
Bug Report
when doing a "getimage -o /usr/tmp/getimage.out -wname" I got the following
error pop-up message:
Could not grab pointer for display
================================================================
I...-221
Bug Description: The range selected in editcmap selected wrong
Toolbox: ENVISION
Object: editcmap
Full bug descripton:
######################################################################
Bug Report
editcmap -i image:gull -pseudo
If you select Color Cell Palette Type and then select a range.
Then change the color & the buttons unselect and do not show the
color change. The color does change, but sometimes they do not
show the new color unless selected.
You can also select individual color cells, i.e. the first and last
and see the behavior.
Not only that, but there is much strange behavior when selecting individual cells
with the LMB, then changing the color with MMB --squeedle 22 May 1997
================================================================
I...-222
Bug Description: putpixel does not consistantly reset offsets
Toolbox: ENVISION
Object: putpixel
Full bug descripton:
######################################################################
Bug Report
putpixel -gui
Set the Input file to image:ball
Run
Then when window comes up,
select Options
Change on Depth, Time or Elements Offset to 2.
You will get a message that that is an invalid value and the Offset
value is set back to 0. If you enter a negative number in for the
offset you get a message,
Invalid double XXX provided ...
The Offset is not set back to 0 when the Ok button is pressed as it
was in the previous case.
NOTE: The Offset is integer item see bug report with short description
"Problem running khelp -gui error prompt wrong". I believe this
is a similar bug.
================================================================
I...-224
Bug Description: The range selected in editcmap hard to do
Toolbox: ENVISION
Object: editcmap
Full bug descripton:
######################################################################
Bug Report
editcmap -i image:gull -thresh
select License
select Other Files
the License button does not fully extent the length of
the pull down.
================================================================
I...-225
Bug Description: can not quit or run stuff when running xvrun -gui
Toolbox: DESIGN
Object: xvrun
Full bug descripton:
######################################################################
Bug Report
xvrun -gui
You can not do anything from the -gui pane other then enter the command
and flip a few toggles, not even quit.
================================================================
I...-226
Bug Description: Changing the range
Toolbox: ENVISION
Object: edithresh
Full bug descripton:
######################################################################
Bug Report
edithresh -i image:ball -priv y
The moving of the range slider jitters and does not always let
you change the lower or upper value to their minimum or maximum
value, repectively.
================================================================
I...-227
Bug Description: Selecting range in editcmap hard to do
Toolbox: ENVISION
Object: editcmap
Full bug descripton:
######################################################################
Bug Report
editcmap -i image:gull -pseudo
If you select Color Wheel Palette Type it is hard to determine what
range is being selected.
================================================================
I...-228
Bug Description: Problem running khelp -gui error prompt wrong
Toolbox: DESIGN
Object: khelp
Full bug descripton:
######################################################################
Bug Report
khelp -gui
select Window width
Enter a number greater than 1000
An error message pops up with
Invalid double XXX provided ...
The Window Width is a integer entry.
================================================================
I...-229
Bug Description: Changing the range in edithresh
Toolbox: ENVISION
Object: edithresh
Full bug descripton:
######################################################################
Bug Report
edithresh -i image:head -update 1.2
When changing the range you get lots of
Cannot set scrollbar value as requested to {VALUE LESS
THAN MINIMUM}; ...
even though the Lower scrollbar is at the minumum. The message
also says it is resetting the minimum, but the value does not
appear in the text box.
================================================================
I...-231
Bug Description: khelp -gui width & height changes have no effect
Toolbox: DESIGN
Object: khelp
Full bug descripton:
######################################################################
Bug Report
khelp -gui
select Window Width
and/or
select Window Height
enter a value(s)
It does not seem to do anything.
================================================================
I...-233
Bug Description: putanimate: Unpredictable behavior when '[' is used
Toolbox: ENVISION
Object: putanimate
Full bug descripton:
Unpredictable behavior for putanimate when '[' is used
######################################################################
Bug Report
putanimate displays some strange behavior when I accidentally entered
a '[' at the end of the path I specified for the input file. It seemed
to go into a -P mode for random other putdata objects (like puticon and
putimage), and would display the first frame using the respective gui
for which it asked questions.
The behavior was found because of a typo in my putanimate.sh script -
see the next-to-last command.
================================================================
I...-234
Bug Description: changing values for -height and -width has no effect
Toolbox: ENVISION
Object: putpalette
Full bug descripton:
######################################################################
Bug Report
Changing the value of the -width and -height options in the command
line has no effect on the size of the displayed window.
================================================================
I...-238
Bug Description: Putplot3 -alloc argument not working in answer file.
Toolbox: ENVISION
Object: putplot3
Full bug descripton:
######################################################################
Bug Report
%putplot3 -A /usr/tmp/plot.ans
For an input file use plot3d:flow
Accept all default values except for:
Enter: (alloc) Color allocation policy --
1 ("Read Only")
2 ("Read Write") , [1] }: 2
Notice I choose 2 here, instead of the default.
When the plot comes up, hit the quit button.
NOW, here's where we run into problems:
%putplot3 -a /usr/tmp/plot.ans
The plot comes up, but you also get an error message:
Program: putdata
Library: kforms
Routine: kvf_modify_form_from_cmd
Invalid or missing list/cycle value after "-alloc" argument
/.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\
ASK ME ABOUT OUR SOFTWARE SUPPORT CONTRACTS!!
Khoral Research, Inc.
Paula Barth 6001 Indian School Road N.E.
paula@khoral.com Suite 200
505 837-6500 ext. 504 Albuquerque, NM 87110
http://www.khoral.com
\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./.\./
================================================================
I...-241
Bug Description: putpixel: Pixmaps specified w/width > 10 won't display
Toolbox: ENVISION
Object: putpixel
Full bug descripton:
for putpixel, pixmaps specified w/ width > 10 won't display
######################################################################
Bug Report
Here's something really obscure: When running putpixel -gui, enter
10 or more as Width of Pixel Value Display & click Run. The application
acts like it is running but then the normal cursor returns. But this
is only true for a .ppm file.
Now try running the same thing on a .ppm file with 8 or 9 as the width,
& it will run this time.
================================================================
I...-244
Bug Description: putpalette: enter width & height on CLUI has no effect
Toolbox: ENVISION
Object: putpalette
Full bug descripton:
Entering width & height in putpalette -gui has no effect
######################################################################
Bug Report
No matter what you enter for width & height for startup of putpalette,
it always comes up at the same size.
================================================================
I...-256
Bug Description: Get bejillion error msgs when plotting a fcn w/ unknown variable
Toolbox: ENVISION
Object: editimage, xprism
Full bug descripton:
Statement of problem...
In my normal spaced-out state I attempted to plot a fcn of x when trying to
verify a bug report, & got lots & lots of error messages.
================================================================
I...-287
Bug Description: BOOTSTRAP examples - Rewrite as testsuites
Toolbox: BOOTSTRAP
Object: examples
Full bug descripton:
Statement of problem...
Because the testuistes in the BOOTSTRAP toolbox are non-interactive, it would
be better if they were made into testsuites instead.
================================================================
I...-294
Bug Description: DESIGN kwidgets/viewport - Labels overlap eachother
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Statement of problem...
214 - The viewport example program places the "horiz label 1" label over top
of the "vert label 3" label. Also, all of the horiz labels border box are
bolded. The README does not indicate this behavior.
================================================================
I...-295
Bug Description: DESIGN layout/04.layout - Shrinking vertical dimension makes buttons flash
Toolbox: DESIGN
Object: layout
Full bug descripton:
Statement of problem...
215 - If you shrink the vertical dimension down enough to where the left and
right buttons are just bars, you can get the right button to get into a
state where it is flashing between creating a full button and just a bar.
================================================================
I...-296
Bug Description: DESIGN layout/05.layout - Buttons don't line up
Toolbox: DESIGN
Object: layout
Full bug descripton:
Statement of problem...
304 layout/05.layout: The left and right buttons do not line up the
same under the top button and over the bottom button. It appears
that they should be the same.
================================================================
I...-297
Bug Description: DESIGN layout/06.layout - Shrinking vertical dimension messes up scrollbars
Toolbox: DESIGN
Object: layout
Full bug descripton:
Paula Barth 5/22/97
I verified this bug on kafka as stated below. Also, when the right scrollbar gets into this state, it flickers and won't allow you control of your mouse.
You can't resize the window or minimize it. All you can do is Ctrl-C the
program.
Statement of problem...
216 - If you shrink the vertical dimension down enough to where the left and
right scrollbars are just bars, you can get the right scrollbar to be
larger than the left and kinda overlapping the bottom scrollbar.
================================================================
I...-299
Bug Description: DESIGN xvlang/finderlist - double-clicking can produce ghost glyph
Toolbox: DESIGN
Object: xvlang
Full bug descripton:
Statement of problem...
Paula Barth 5/22/97
I verified this bug on fresca as stated below.
218 - The README states that if I double click on a item in the finderlist
I will get the yin-yang symbol until I place the glyph in the workspace.
The yin-yang symbol flashes but goes away. I can then click many other
items with a similar result. When I move the cursor over to the
workspace, it will be the first item I clicked glyph. But occaisonally
there will be a ghost glyph that I can not get rid of in the workspace
without refreshing the screen. Sometime the ghost glyph will appear while
I am clicking and then disappear and then re-appear when I go to place a
glyph.
================================================================
I...-305
Bug Description: DESIGN xvlang/workspace_restore - Warning pop-up looks wrong
Toolbox: DESIGN
Object: xvlang
Full bug descripton:
Statement of problem...
224 - The warning pop-up has the bottom lines of something cut off at the
top of the box. It also says: Warning: Warning: Blah
This is reminiscent of Lost in Space with Warning: Warning: Will
Robinson.
================================================================
I...-313
Bug Description: DESIGN xvutils/multilist - Garbage in the text widget
Toolbox: DESIGN
Object: xvutils
Full bug descripton:
Statement of problem...
505) The example runs, however, when a color is entered in the input
space, garbage is appended to the input, which is also added
to the value outputed to the tty. This is probably due to the
text widget changes.
================================================================
I...-314
Bug Description: DESIGN xvutils/warn - README doesn't discuss this example program
Toolbox: NS
Object: DESIGN
Full bug descripton:
Statement of problem...
105: xvutils/warn -- the README for xvutils does not have an entry for
warn. the example ran correctly.
================================================================
I...-316
Bug Description: DESIGN xvwidgets/08.textfile - README doesn't properly describe program
Toolbox: DESIGN
Object: xvwidgets
Full bug descripton:
Statement of problem...
507) The README does not actually display a file, it display the file
path and name, and allows the user to input text. Is this
what was supposed to happen?
================================================================
I...-317
Bug Description: Display of integer image with range > 65536 unclear error msg on some archs
Toolbox: ENVISION
Object: NS
Full bug descripton:
26-Feb-1997 by squeedle
This is what I did to try to reproduce this bug:
Ran editimage, opened Function Pane from Options & entered a range of 68000 & plotted
2000 points. The Function was just "w". I just get an error message saying:
Program: editimage
Library: xvforms
Routine: xvf_action_cb
Cancelling action Create Data.
NOTE : THE ABOVE WAS ON A DEC ALPHA.
On the Linux box it plotted just fine, even with 1000 points.
I changed this to Err/Warn/Prompt Unclear.
Original Rpt: From kscp1@pc512.fel.tno.nl Thu Feb 22 00:59:23 1996
If we try to visualize an integer image (long, for example) with
a range bigger 65536 we get undefined colors. This can result in
a segmentation violation.
The problem is that in .../library/xvisual/src/ImageUtil.c
function ImageInit care is taken for integer images that
its values are scaled up from the minimum, but that no check
is performed to the maximum.
The problem is how to get this maximum. It clearly is not useful to
have 2^32 (for 32 bit machines -- 2^64 for 64 bit machines) entries
in the lookup table. But the X lookup table (per color) is limited
to 2^16 entries, so a greyscale of 2^16 should do. Next problem: how
to map all those entries to this limited range? (Actually, we often
have as little as 6 or 8 bits used by the hardware. And
distinquishable by the human eye)
The method used for real images is to map the whole dynamic range of
the image to the dynamic range of the hardware (in 256 bins.) Should
this also be used for integer images? So this means that we also have
to change the behaviour of 16 bit (short) images!
Or should we pick a certain dynamic range and clip / wrap around the
other values?
If you have an opinion about which strategy to follow
I might find time to make a patch.
Notes : See email message Feb 24 "Forwarded mail" guiviz
for workspace reproducing bug
================================================================
I...-325
Bug Description: char vs pixel geometry on menuforms - 0.1 turns into 0.0960
Toolbox: DESIGN
Object: xvforms
Full bug descripton:
Statement of problem...
304 - 04.ellipse: in the width window (and the others), when I type in 0.1,
it changes to 0.0960899666 when I hit the carriage return.
================================================================
I...-326
Bug Description: ENVISION annotate/04.ellipse - Changing border width produces odd results
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
305 - 04.ellipse: first set border width to extra wide, then change it
to fine, and the border changes into a pair of horz and vert
tangent lines.
================================================================
I...-327
Bug Description: ENVISION annotate/04.ellipse - Wrong menu pops up
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
306 - 04.ellipse: pull down the border width menu and *slowly* move
your mouse down to the menu items. as you cross over the first
line, you suddenly get the border type menu instead. (same
thing happens between the border color and fill color menus).
Another problem, When the pulldown menus are set to be the same
size, and stacked on top of each other, as they are with the
border width and border type, it is rather difficult to make
a selection for the border width(top menu). This is because,
even if the pulldown menu is selected, if the mouse is over
the second or lower pulldown menu, the menu changes. You
actually have to move the mouse outside of the menu, and come
in from the left side to select the first option 'extra fine'
for the border width. This probably would not work if the
first entry of the second or lower menu was not so short 'solid'.
This would cause certain options to be unatainable.
================================================================
I...-328
Bug Description: ENVISION annotate/05.line - Moving a line doesn't produce expected results
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
307 - 05.line: make the line extra wide and a bright color. interactivly,
place one end of the line in the upper right corner. grab the other
and bring it to the upper left. now (still grabbing the upper left
end) drag the end to the bottom of the image. note (1) that it
leaves a trail of lines over the image, and (2) the right end of
the line also moves down (you would expect it to stay stationary).
================================================================
I...-331
Bug Description: ENVISION annotate/06.circle - Changing circle size leaves stray marks
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
308 - 06.circle: make the border extra wide. make the radius big, then small.
note that stray marks are left where the big circle was.
================================================================
I...-333
Bug Description: ENVISION annotate/07.rectangle - Changing border width leaves stray marks
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
309 - 07.rectangle: make border width extra wide, set X position to 0.
a partial outline of the rect's old position is left.
523) Again, garbage characters entered in the fill color space. Another
side note, when the rectangle is placed in edit mode and moved,
a significant amount of garbage is left on the screen when
the object is moved around.
================================================================
I...-336
Bug Description: ENVISION annotate/08.string - Can't select bold fonts
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
310 - 08.string: can't select any of the bold fonts.
Also affects annotate/09.stringvalue, annotate/11.date, annotate/12.timer,
(Issues 342,344,&346, respectively)
================================================================
I...-340
Bug Description: ENVISION annotate/09.stringvalue - %10.4g only displays 2 decimal places
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Paula Barth 5/22/97
I verified this bug on fresca as stated below. Also, if you set
the Value to '45.123456' and the Format to '%10.6g' the result
printed in the window has four decimal places and looks like: 45.1235
Yowzer! Where did the 4 go?
Statement of problem...
313 - 09.stringvalue: set format to %10.4g and value to 45.123 and it only
displays with two decimal places.
================================================================
I...-341
Bug Description: ENVISION annotate/09.stringvalue - Can't justify center left, center right
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
312 - 08.string: no way to justify center left, center right, center bottom,
or center top.
525) The justification does not seem to move the number, however, it can
be moved if the parent is placed in edit mode, and moved with
the mouse.
Also, the Foreground, Background, and Highlight colors all have
garbage characters as their input values untill a value is
specifically selected from the menu.
================================================================
I...-343
Bug Description: ENVISION annotate/09.stringvalue - Uses different coordinate system
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
311 - 08.string: seems to use a right hand coordinate system instead of
the left hand system you normally see in image processing (pos y dir
is down).
================================================================
I...-345
Bug Description: ENVISION annotate/11.date - XDefaultBackground instead of white
Toolbox: ENVISION
Object: annotate
Full bug descripton:
Statement of problem...
315 - 11.date: select 'white' back/fore ground color or hilit color,
and it says you have 'XtDefaultBackground' rather than white.
526) The background Color contains garbage charcters in it.
================================================================
I...-355
Bug Description: ENVISION autocolor/3.autocolor_timer - menu items disappearing & reappearin
Toolbox: ENVISION
Object: autocolor
Full bug descripton:
Statement of problem...
319 - 3.autocolor_timer: runs very slow. got confused about what's going
on with all the menu items disappearing and reappearing.
================================================================
I...-356
Bug Description: ENVISION image/12.updateimage - image did not update when file overwritten
Toolbox: ENVISION
Object: image
Full bug descripton:
Statement of problem...
104 image/12.updateimage -- README indicates that the image will
change when the image file is overwritten. This did not happen
when I overwrote the file.
================================================================
I...-358
Bug Description: ENVISION image3d/1.putimage3d - image mostly blue
Toolbox: ENVISION
Object: image3d
Full bug descripton:
Statement of problem...
533) The image is mostly blue, not a very good example.
================================================================
I...-360
Bug Description: ENVISION pseudo/1.pseudo_display - Button overshoots the mouse position
Toolbox: ENVISION
Object: pseudo
Full bug descripton:
Statement of problem...
321 - 1.pseudo_display: button overshoots the mouse position when dragging
the RGB buttons.
See also threshold/1.threshold_display (Issue 366)
================================================================
I...-361
Bug Description: ENVISION pseudo/1.pseudo_display - Color cell does not come up correctly
Toolbox: ENVISION
Object: pseudo
Full bug descripton:
Statement of problem...
320 - 1.pseudo_display: color cell does not come up correctly. after
starting example prog, click 'show pseudo palette' to false, and
then back to true.
531) Psuedo palette does not get displayed properly until the 'show
psuedo palette' button is toggle from true to false, and back
again. (Or the window is resized --squeedle, 22 May 97)
This also affects 4.pseudo_update, 6.pseudo_readwrite, 7.pseudo_readonly
(Issues 362,364,365, respectively)
================================================================
I...-363
Bug Description: ENVISION pseudo/4.pseudo_update - Color cell object in lower left corner
Toolbox: ENVISION
Object: pseudo
Full bug descripton:
Statement of problem...
322 - 4.pseudo_update: color cell object is in *lower* left instead of
upper left as it should be.
================================================================
I...-371
Bug Description: May not plot sensibly when trying to plot over area that is undef
Toolbox: ENVISION
Object: xprism
Full bug descripton:
26-Feb 1997 squeedle
First of all, not all of these functions are actually *undefined* from -1 to 1 -
some are just asymptotic but the differential is calculable everywhere.
The following fcns still have problems:
acos(x), ln(x), log2(x), log1p(x), and sqrt(x) will not plot anything or give any kind of error message about it.
impulse(x) just draws a line over the x axis.
gamma(x) gives many error msgs saying "Invalid argument 3.25486e+78 to gamma"
Original Rpt:
xprism may crash on some architectures when attempting
to plot a function over an interval where that function
is undefined
list of functions which cause core dump when plotted over
undefined interval:
xprism; use function listed from -1 to 1:
- tan(x) - tanh(x) - acos(x) - asinh(x) - ln(x) - log2(x)
- loglp(x) - sqrt(x) - cbrt(x) - erf(x) - erfc(x) - step(x)
- impulse(x) - gamma(x)
================================================================
I...-372
Bug Description: Numeric Label X and Y fields can change become invalid
Toolbox: ENVISION
Object: xprism
Full bug descripton:
verified 26-Feb-1997 by squeedle
Notes : The Numeric Label Format X and Y fields in the Precise
Axis Attributes form can be changed to invalid values,
ie. %z etc. If %s is used, it causes an error window to
appear, but after that, you can enter %g and it does not
update plot. If you manage to get it back by entering
various values in the field %d, %h etc, and then go back
and use %c or %s, it gives a segmentation fault and the
program crashed.
================================================================
I...-385
Bug Description: Xprism fails on plotting constant data
Toolbox: ENVISION
Object: xprism
Full bug descripton:
verified 25 Feb 1997 by squeedle
xprism cannot plot a constant input. The output gets layed out incorrectly.
To repeat this ( I used Zen), use the constant glyph and use a non-integer
constant. Input this to 2D plot or interactive and the numbers that should be
along the X Y axis are displayed in the middle of the plot area. Rafael
Santos reported that on his Free BSD system it crashes with an arithmetic
exception.
Note: to reproduce quickly, cantata -wksp
~/work/guiviz/tb/examples/const_plot.wk
================================================================
I...-391
Bug Description: incorrect error msg for extractor
Toolbox: ENVISION
Object: extractor
Full bug descripton:
######################################################################
Bug Report
The extractor.sh script issues the following command:
+ extractor -i /research/vision/squeedle/grafix/scallopsmall.ras -o scallopsmall.kdf -priv y -update 1.00033333 -gui -alloc 2 -disp false - policy 3 -mult false -pres 1 -shape 4 -x 100 -y 350 -A /usr/tmp/extractor.ans
which produces the following error msg:
Error: The [-gui] option may not be combined with interactive prompting [-P],
answer file input [-a], or answer file output [-A]
I am not sure if there should be an error msg or not, but the one given
is certainly not correct.
================================================================
I...-392
Bug Description: incorrect error msg for getimage
Toolbox: ENVISION
Object: getimage
Full bug descripton:
######################################################################
Bug Report
when running getimage with the command:
getimage -a /usr/tmp/getimage.ans
recieved the following error msg:
'/research/release/kpro/khoros.ans' answer file not found
This error msg names the wrong file that cannot be found
================================================================
I...-395
Bug Description: kmath would not do float to unsigned int or unsigned long conversion
Toolbox: DATAMANIP
Object: kmath
Full bug descripton:
Steve Kubica 2-May-1997
This is actually an guiviz bug.
The conversion is being done okay. A -1 is being cast to
4294967295 (since it's unsigned). This value confuses the color
class, which wants to create a 3 by -2147483647 map in the
function ColorMapInitialize.
To reproduce, on mohawk :
% echo -1 > file; kasc2val -type 7 -i1 file -o file; putimage -i file
kmath would not do float to unsigned int or unsigned long conversion
######################################################################
Bug Report
when running cantata.sh on display_type.wk, got two errors
ERROR: Illegal NULL Parameter for putdata. This was because
the program was attempting to convert a float to both an unsigned int
and an unsigned long.
Program: putdata
Library: kmath
Routine: kdata_cast()
--
Khoral Research, Inc.
6001 Indian School Road N.E. URL http://www.khoral.com/
Suite 200 fax 505 881-3842 |/ ) |
Albuquerque NM, 87110 |\ \ |
================================================================
I...-399
Bug Description: Can't reduce the i-file button in GUISE.
Toolbox: DESIGN
Object: guise
Full bug descripton:
######################################################################
Bug Report
While using GUISE, and editing an input file variable, the size of the
button increases as the text gets larger, however, does not get smaller
when the text is reduced. In fact, you cannot reduce it at all, forcing
the user to delete the input string, and recreate it.
Chris
===============================================================
A. Christopher Moorman | Email: chris@khoral.com
Khoral Research, Inc. |
6001 Indian School Rd. NE | http: //www.khoral.com/
Suite 200 | phone: (505) 837-6500 Ext 505
Albuquerque, NM 87110 | FAX: (505) 881-3842
==============================================================
================================================================
I...-400
Bug Description: Cannot quit runwksp when in rootwindow
Toolbox: DESIGN
Object: runwksp
Full bug descripton:
######################################################################
Bug Report
If you runwksp -rootwindow, you must kill the process by hand, i.e. the
Quit button doesn't seem to work.
================================================================
I...-403
Bug Description: ENVISION plot_color/04.2Dcoords_map - Garbage chars in foreground color
Toolbox: ENVISION
Object: plot_color
Full bug descripton:
Statement of problem...
535) The Foreground Color entry in the 2D Plot Object Menuform contains
a bunch of garbage characters. The functionallity is correct,
this is just a cosmetic issue.
================================================================
I...-404
Bug Description: ENVISION plot_color/09.3Dcoords_map - Garbage chars in foreground color
Toolbox: ENVISION
Object: plot_color
Full bug descripton:
Statement of problem...
535) The Foreground Color entry in the 2D Plot Object Menuform contains
a bunch of garbage characters. The functionallity is correct,
this is just a cosmetic issue.
================================================================
I...-405
Bug Description: ENVISION plot_color/07.3Dfile_map - Garbage chars in foreground color
Toolbox: ENVISION
Object: plot_color
Full bug descripton:
Statement of problem...
535) The Foreground Color entry in the 2D Plot Object Menuform contains
a bunch of garbage characters. The functionallity is correct,
this is just a cosmetic issue.
================================================================
I...-447
Bug Description: kgenstruct does not support typedef'd arrays
Toolbox: BOOTSTRAP
Object: kgenstruct
Full bug descripton:
Statement of problem...
kgenstruct does not support typedef'd arrays.
For example, if you have this in your .x file
typedef int int_a[10];
it will assume that this is just an integer, not an integer array.
Typedef variable arguments also do not work (eg typedef int int_v<>;)
This is not an easy problem to fix, but it doesn't really affect anyone,
so there is no priority to fix it. It should be documented that this is
not working, though.
================================================================
I...-448
Bug Description: Start Daemon fails when echo statements are present in .cshrc
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
This bug is probably present all architectures, however, the problem was
noticed on the SGI's. If a echo statement is present in the .cshrc file,
the dist_start_daemon routine returns an error about the hostname, user name,
etc. being invalid and the daemon is unable to start. This was noticed when
the user selected the configure remote hosts options under workspace from
within cantata. When the user selected a host and tried to start the
daemon, the process fails.
================================================================
I...-449
Bug Description: Local hosts daemon for distrib. computing fails to start
Toolbox: BOOTSTRAP
Object: kutils
Full bug descripton:
Statement of problem...
This problem was also first noticed on a cluster of SGI workstations.
When the configure remote hosts option is selected from within cantata
under the workspace option, and the user trys to start a daemom on the
local host, it fails with the following error:
TOOLBOX: DESIGN
Program: Cantata
Routine: dist_start_daemon
Error: No such process
Host 'localhost' may not be valid, user 'fas' may not be valid, the
command 'rsh' may not be valid, or there is no .rhosts file allowing
remote access.
The user can, however, enter the actual name of the local machine in the
line below the machine list, and the daemon will start fine.
================================================================
I...-463
Bug Description: If someone does not agree to License, program should exit
Toolbox: BOOTSTRAP
Object: klicense
Full bug descripton:
Statement of problem...
I just ran craftsman for the first time on toaster.
Interesting sequence of events:
1. kregister pops up
2. I click on the options menu and say "Do not register"
3. It pops up the licence and asks me if I accept the terms.
4. I say 'No'
5. It pops up an information box saying "Your version of Khoros
is not REGISTERED or LICENCED...
6. Craftsman pops up:
Toolbox: DESIGN
Program: Craftsman
Routine: khoros_imprint
Library: kutils
Error: No such file or directory
Your version of Khoros is not REGISTERD or LICENCED...
(yeah...yeah...)
7. I hit OK.
8. It pops up the licence and asks me if I accept the terms.
(uhhh...enough already)
9. I say 'No'.
10. It pops up an information box saying "Your version of Khoros
is not REGISTERED or LICENCED...
(yeah, I heard)
I thought I'd mail off since I'm not sure that anyone bothers to be
argumentative with kregister...I think it could have exited somewhere
around step 4 or 5...
================================================================
I...-467
Bug Description: Image/panicon display clips tall images
Toolbox: DESIGN
Object: xvisual
Full bug descripton:
Paula Barth 5/22/97
Verified this bug on fresca. The image, after magnifying the height by
1.5 will be 768 pixels tall. Except putimage only displays 750 pixels.
--------------
Statement of problem...
Image/panicon display clips tall images without poping up pan icon.
If an image is "tall enough", it kicks in a pan icon and displays with
an image height of around 750.
If an image is "tall", but not "tall enough", it still displays only
with a height of about 750, but doesn't bring up a pan icon. The net
effect is that the image display is "clipped" at 750.
To view this subtle bug, try the following :
% kresample -i image:ball -o /tmp/tall -hmag 1.5
% putimage -i /tmp/tall
/tmp/tall has a height of 768, but is clipped at 750 in the display (visually
it appears clipped, and the position widget also concurs).
================================================================
I...-468
Bug Description: igamut corrupts data
Toolbox: IMAGE
Object: igamut
Full bug descripton:
Statement of problem...
Running the kpro version,
run
$ cantata -wksp ~ele/transfer/ifft.wk
Click on (i.e., run) editimage. Click 'Options'. Click
'Colormaps'. Select 'Greyscale', then 'Close'. Select 'Files'. In
'Output Image', enter '/usr/tmp/junk.kdf' and hit carriage return.
Then close. Now run 'Compress Colors' which uses the junk.kdf image
as input. Then run 'Display Image'. The result is a hosed image. Go
into 'Options' and change the normalization to see the nature of the
hosing better.
This is an addendum to the bug report I sent in yesterday on igamut.
Yesterday's report is attached to this letter.
Igamut is working correctly except for the fact that it does not check
to see that there are three layers of data. You should get an error
message if there are not 3 layers of data. If you feed it a
256x256x1, instead of a 256x256x3, if will not complain (but it
should) and will give you garbage output.
In the description of the but below, if you map the data through its
map before feeding it to igamut, it will work fine.
================================================================
I...-477
Bug Description: changing from cont-run to single-run mode in Cantata fails first time
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
The manual says that in single-interation mode, any changes to
the parameters will not fire the glyphs.
This is not happening only in the first time when you change from
continous mode to single mode.
to see this problem do the following:
1. get any workspace with more than 2 glyphs
2. set the continuous mode (default) and Run
3. change any parameter in the middel glyph - it will run
4. set to single mode
5. change any parameter in the middle glyph - it will RUN ->BUG
6. repeat again 5 and it's ok
================================================================
I...-478
Bug Description: synchronization problems in cantata
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Paula Barth 5/21/97
Re-verified on naduah. Same behavior.
Statement of problem...
1) %cantata -wksp ~mirage/testtb/examples/workspaces/sync.wk
2) Run the workspace. Everything works fine the first time through.
3) Click the Stop button to stop the scheduler. (This is not needed to
demonstrate the bug, but it makes it easier to see what's happening...)
4) Change the input file to some other image
5) Re-run the workspace. Note that the Compare Objects glyph runs before
the bottom path finishes executing, thus resulting in an error.
================================================================
I...-479
Bug Description: Autocolor should set COLOR MODEL attribute
Toolbox: ENVISION
Object: putimage
Full bug descripton:
Statement of problem...
Run the following command:
% cantata -wksp ~mirage/testtb/examples/workspaces/autocolor.wk
After autocolor, putimage display the image colorful, even knowing
that the image has a Map but none color model.
If you map the colormap into value, the image has 3 elements but
none color model as before, than putimage display the image as
it was 3 independe elements, showing just the first.
The solution I recommend is to change autocolor to set attribute
colormodel: RGB
================================================================
I...-480
Bug Description: khistops/test1.sh failed on jolt
Toolbox: DATAMANIP
Object: khistops
Full bug descripton:
Statement of problem...
datamanip/testsuite/khistops/test1.sh failed on jolt.
The output of the test is as follows:
running test1.sh for khistops ... process (16802)
FAILED: Test 1 --> equalize - unsigned byte input, no mask, no map, whole
Passed: Test 2 --> equalize - signed byte input, no mask, no map, whole
Passed: Test 3 --> stretch - unsigned short input, no mask, no map, rows
FAILED: Test 4 --> same as test#1, but for large data set handling
================================================================
I...-485
Bug Description: Large fonts causes text fields to overlap
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Statement of problem...
Using large fonts causes text fields to overlap.
In windows that contain multiple text fields,
using a large font can cause text fields to run
into each other. For example, in the cantata "Variables"
window the results of evaluating an expression
overlaps the "Result of Evaluation:" label.
A related problem can be seen in multi-line text
widgets. Some characters, especially "y", lose
their descenders, so that "glyph" looks like "gluph"
================================================================
I...-491
Bug Description: kconfigure says it found binaries, but it didn't!
Toolbox: BOOTSTRAP
Object: kconfigure
Full bug descripton:
Original Rpt:
When kconfigure is run for you after running
kbuild, it says:
bootstrap binaries located in [...]
dataserv binaries located in [...]
design binaries located in [...]
ok so far . . .
datamanip binaries located in [...]
[similar for all standard toolboxes
I know this is checking where they would be
installed, but I haven't even installed
DATAMANIP et al, so this is a little misleading
perhaps.
================================================================
I...-503
Bug Description: cantata should wait to create tmpfiles till a connection is made
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
cantata generates glyph output tmpfiles when you open the form.
it should wait until you run the glyph or make a connection to
the output.
Tom Konerth adds:
I agree firmly, especially with the user defined glyph. You open the pane and it shows some bogus file in var/tmp. And when you open the browser, it is in /var/tmp also. Which makes you do a bunch of clicking to just get to your home
area. Also, with the bogus tmp file, if you connect it up it shows that data is ready, if you don't remember to set it to a valid file the wksp fails first time.
================================================================
I...-507
Bug Description: Expression glyphs do not handle stream connections
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
The Expression glyphs do not seem to handle stream connections
properly. I have two connected in series and the first one starts running
and the second one never starts.
================================================================
I...-518
Bug Description: UNIX Domain Socket needs to be added to Preferences
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
UNIX Domain Socket is a transport type in the transport list but not
in the Preferences/Workspace menuform.
================================================================
I...-519
Bug Description: When entering an out-of-range integer, the error statement says a double
Toolbox: IMAGE
Object: igamut
Full bug descripton:
Statement of problem...
When entering an out-of-range integer, the error statement says a double
argument is out of range. Example
igamut -i ~ele/data/e43p2oct90_512x768.rgb.kdf -o foo -n 64 -p 40 -r 0.5
================================================================
I...-523
Bug Description: can't read craftsman info message
Toolbox: DESIGN
Object: craftsman
Full bug descripton:
Statement of problem...
In craftsman, create a library in some toolbox. When it asks you if
you want the toolbox to depend on this library, say yes. Then it pops
an info window called "Information for craftsman," but the window is
too small and you can's scroll down to see what's getting cut off.
================================================================
I...-536
Bug Description: When copying a lib the tools fail to create the X dependency
Toolbox: DESIGN
Object: Craftsman
Full bug descripton:
Statement of problem...
When you copy a library which has an X dependency, it fails to
remember that when adding the dependency to the new toolbox.
To recreate, copy a library from design (such as xvlang) to a
new toolbox. It will do an AddLibrary as if it were a C
library.
================================================================
I...-538
Bug Description: procedure window does not resize
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
In cantata, if you open a procedure glyph and turn it into a
top-level window, its workspace does not resize when you
resize the window.
================================================================
I...-540
Bug Description: deleting a glyph on the procedure window doesn't work correctly
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
In cantata, if you open a procedure glyph and then turn it
into a top-level window, deleting a glyph on the procedure
window doesn't always work correctly.
To reproduce:
- start cantata
- open workspaces:procedures:Procedure2
- open a procedure
- click on the up-arrow in the upper right corner
to create a new top-level window
- select a glyph in the main window and delete it with the trashcan tool.
the glyph is deleted
- select a glyph in the procedure window and try to delete it.
nothing happens. If a glyph is selected in the main window, it
will be deleted instead.
================================================================
I...-547
Bug Description: make gui doesn't update usage.c file with changes to uis
Toolbox: BOOTSTRAP
Object: kcodegen
Full bug descripton:
Statement of problem...
If you change a *.form file or one of the included files of a *.form file
and do a make gui, it doesn't update the compiled in version of the
form in usage.c file. Only a make gw will do that. It's probably designed
that way, but it's inconsistent with the separation we've made between
graphical X code generation and clui code generations.
================================================================
I...-548
Bug Description: kbugreport is annoying and sends mail without subjects
Toolbox: BOOTSTRAP
Object: kbugreport
Full bug descripton:
Statement of problem...
Kbugreport asks way too many questions about things that
are not necessarily related to the actual bug.. This is
rather annoying.. However, this message is really about
the fact that it sends a bug to khoros-bugs, and I get
an auto reply saying that "your bug report concerning
(No subject specified) will be addressed soon" or some such
nonsense.. The annoying part is the "No subject specified" which
of course, wasn't one of the millions of questions that was
asked of me.. :)
You either need to set the subject, or take that subject line
out of the automated reply. Personally, I'd like to see
the short description go into the subject, but that's up to
you..
================================================================
I...-551
Bug Description: glyph connections disappear and can't be redrawn but still exist
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
Tom Konerth 5-1-97
I verified this on Lizard, although the report is from a SGI. So I imagine it is hardware/OS independent.
I would like to report a bug with Cantata. I haven't checked the faq to
see if someone already mentioned about it. I have compiled the whole
Khoros distribution under Irix 6.2, but I'm using it today from a
workstation with Iris 5.2 or 5.3.
When a bunch of interconnected glyphs are selected with the selection
rectangle, and if I move this selection beyond the limits of the window,
and then come back, the interconnections have disappeared, and never
come back, even if I force a redraw. But the interconnections still
work, although they do not appear any more.
================================================================
I...-552
Bug Description: "Close" and "CLOSE" inconsistant (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
E-mail : ??
Display Depth :
######################################################################
Bug Report
In Xprism, the "close" button on some windows is "Close" and on others
is "CLOSE". This is an entirely cosmetic bug.
================================================================
I...-553
Bug Description: "Close" Button in wrong place on help (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
E-mail : ??
Display Depth :
######################################################################
Bug Report
The on-line help box has the "close" button on the bottom of the
box, instead of in the upper right hand corner (where it is on all
the other windows). This is purely cosmetic.
================================================================
I...-560
Bug Description: kfile_read does not return errors correctly
Toolbox: BOOTSTRAP
Object: kutils(kfile_read)
Full bug descripton:
Statement of problem...
I'm not sure how to reproduce this locally, but error conditions
happening in the read routine are not passed up via the
Khoros transports. (i.e. an error happening to read is not
passed back to the user when they call kread) This is a major
problem for coders expecting our k-calls to work like the unix
call. As for reproducing, umm.. I have a case at lbl with
an NFS problem they are having. The trick is to open a file,
and then for some reason not having permission to read the file.
It's not extremely critical to reproduce this exact error.. This
is a design flaw in how the transport works.. It assumes that
a -1 returned from the transport read routine means EOF, however,
this isn't necessarily the case.
================================================================
I...-563
Bug Description: Extra Wide--> Extra Fine doesn't redraw correctly (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
Plot:
Files:
plot2d:small-exp
Options:
Axis2D:
Open Menuform:
Axis Line Width:
Set the X Line Width to something, then to some "thinner" thing
leaves a wider line on top than it's supposed to be, until the "Refresh"
button is pushed.
For Example:
Set the X Line Width "Extra Wide", then "Extra Fine" leaves a wide
line on top (where it should now be "Extra Fine")
Set the Y Line Width to something, then to some "thinner" thing
leaves a wider line on the left, and a split (and wide) line on the
right.
For Example:
Set the Y Line Width "Extra Wide", then "Extra Fine" leaves a wide
line on left and split line on the right.
================================================================
I...-564
Bug Description: Tic Justification makes no chnages (XPrism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Plot:
Files:
plot2d:small-exp
Options:
Axis2D:
Open Menuform:
General:
Turned Grid on for X and Y axes
Precise:
Tic Justification:
Changed it for X and for Y from inside to outside and back
What Happened:
No apparent change. I just don't see any difference, even when I enlarged
the window so that it filled my whole screen. I wasn't sure what difference
this would accomplish, but it either isn't happening, or it's very subtle.
================================================================
I...-565
Bug Description: Browser Buttons are in an unexpected place
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Called up the Alias Browser (this is also true for the Directory Browser)
What Happened:
The "Ok", "Cancel", and "Help" buttons are all at the bottom of the
window instead of being in the upper right hand corner.
Also, "Cancel" is used instead of "Close" or "CLOSE".
================================================================
I...-568
Bug Description: Help page displaying "p","g","q" oddly (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Called up "Help" from the main Xprism window
What Happened:
Although most of the p's, g's, and q's seem to be ok, some
of them are not displayed entirely. For example, the first
subsection "Plot" looks like this:
Displays the Plot subform, which is used to create and display plots.
^ ^ ^
------
These three letters are cut
off on the bottoms.
When I minimize the window, and re-open it, the letters that I was
looking at are fine, but others (like "ggy" in "buggy" at the bottom)
are still having this trouble.
================================================================
I...-571
Bug Description: 2D Axis Object Menuform (can't put pi in range) (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Plot:
Functions:
f(x) = sin(x)
Options:
Range:
changed X range from (-3.14??? to 3.14??) to (-pi to pi)
clicked "Change Range"
What Happened:
Got an error box:
Library: kforms
Routine: kvf_call_expr_handler
^'-pi' is not a proper double
I clicked "ok" and got:
Library: xvforms
Routine: kvf_action_cb
Cancelling action CHANGE RANGE
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
occured IRIX fresca
occured AIX shasta
================================================================
I...-572
Bug Description: For large # of Areas Across -- seg fault on taishan (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Layout:
set number of Areas Across to 200
Close
Plot:
Functions:
f(x) = x
PLOT FUNCTION
.
.
.
.
PLOT FUNCTION
[created 70 plots]
Close
What Happened:
After about 70 plots, they stopped getting smaller, and only one plot was
displayed
at a time (full size). I can't tell which menuform it is. Changing the
background color
on a specific menuform doesn't seem to have any effect.
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured* Solaris tamale
different** Linux taishan
occured IRIX fresca
occured AIX shasta
*Got an error box:
Library: xvwidgets
Routine: xvw_vset_attributes
xvobject is NULL. The object argument for xvobject's set attributes function
needs to be
an active (non-null) object.
**Segmentation Fault
================================================================
I...-573
Bug Description: cantata -gui (Miniature not in proper font) (Cantata)
Toolbox: DESIGN
Object: cantata
Full bug descripton:
######################################################################
Bug Report
What I Did:
on command line:
cantata -gui
What Happened:
The GUI window comes up with "Miniature" in a regular font instead
of grey-ed out. It just looks different than everything else on
the windown.
================================================================
I...-574
Bug Description: "Band Number" vs "Frame Number"(Animate)
Toolbox: ENVISION
Object: animate
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I Did:
File:
Open: Workspaces:5DHead
clicked Animate glyph (to run it)
placed animation window
Options: Attributes
set "Show Band Number" to False
What Happened:
In the "Animation" Window, the line which displays the "Frame Number"
disappeared.
It would be more clear if the wording on these two things agreed
(and I find "Frame Number" more intuitively obvious).
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
untested IRIX fresca
occured AIX shasta
================================================================
I...-578
Bug Description: Comment Box has no wrap/scroll bars(Cantata)
Toolbox: DESIGN
Object: cantata
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I Did:
File:
Open: Workspaces:5DHead
click on "Comment Icon"
What Happened:
The text in the Comment Icon isn't wrapped at the end of the
line, and there are no scroll bars, and the "End" key (on the
keyboard) doesn't put me to the end of the line.
Also, the text box for the Comment Box doesn't enlarge, when I
resize the window (even if I close it and re-open it).
The problem is that I can't see the rest of the text, unless I
move the cursor (using -> key), so that the text scrolls by.
This is very irritating, and makes the information hard to read.
Also, the text overwrites the bounds of the box by a few pixels
(on the right hand side), and these artifacts are left when the
line is edited).
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
untested IRIX fresca
occured AIX shasta
================================================================
I...-579
Bug Description: No display of workspace name (Cantata)
Toolbox: DESIGN
Object: cantata
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I Did:
File:
Open: Workspaces:5DHead
What Happened:
No where on the Cantata screen does it say what workspace I'm using.
It would be more usable if I could tell
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
untested IRIX fresca
occured AIX shasta
================================================================
I...-586
Bug Description: New Connection Type didn't work in a Procedure (Cantata)
Toolbox: DESIGN
Object: cantata
Full bug descripton:
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I Did:
opened workspaces:5D-Head
rubberbanded Extract,Autocolor,Expand
clicked the "Create Procedure" Icon
placed the Procedure
clicked the white triangle on the Procedure glyph
Options:Preferences:Glyphs:
Set Connection Type to Direct Line
click [Apply Changes]
Close
What Happened:
In the Procedure, the Connection Type did NOT change,
even though it did change in the Main Workspace. I tried
changing the Connection Type to all of the possibilities, and
it didn't change it for the Procedure. Whatever the preference
was when I opened the procedure, that's what it stayed.
What OS/machine:
test status OS machine
---------------------------------------
occured OSF naduah
occured Solaris tamale
occured Linux taishan
untested IRIX fresca
untested AIX shasta
================================================================
I...-595
Bug Description: 2D Axis Displayed Range (0..0) unexpected (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I did:
Plot:
Files:
Plot Dimension set to "2D"
selected file "plot2d:small-exp"
Options:
selected "Axis2D(1)"
selected "Open Menuform" (on the right)
selected "Actual Displayed Range" for X to be 0 to 0
clicked "Change Range"
What Happened:
Rather than displaying the entire graph as a single horizontal
line (which I expected), the graph's range became 0 to 0.1 (which
had been the previous right hand boundary), but the values in the
2D Axis Object Menuform remained 0 to 0.
================================================================
I...-596
Bug Description: "Remove Area" doesn't prompt the user y/n (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Display Depth :
######################################################################
Bug Report
What I Did:
opened a file of data
clicked "Remove Area"
What Happened:
The Area just disappeared. Unlike "Clear", I wasn't prompted
whether or not I wanted to have that area removed.
An "undo" button would help, too.
================================================================
I...-598
Bug Description: Line Marker -> Line Plot (incorrect redraw) (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Plot:
load a file (in this case, dps)
Options:
Select "Plot2D"
Open Menuform (on the right)
set "Plot Type" = "Line Marker"
set "Plot Type" = "Line Plot"
What Happened:
There are boxes left on the screen (or piece of boxes) where the
redraw didn't cover them. Pushing "refresh" on the main panel
redraws this correctly
================================================================
I...-600
Bug Description: Tic Justification doesn't redraw correctly (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Plot:
Files:
set "Plot Dimension" = "2D Plot"
use Alias Browser to access "plot2d:small-exp"
Options:
select "Axis2D(1)"
select "Open Menuform" (right hand side)
Range:
change "Actual Displayed Range" to X: 0 to 3, Y: 0 to 100
General:
Turned on the Grid for X and Y
change X and Y axes labels
Precise:
Tic Justification ~ "Outside" for X and Y
What Happened:
The way the tics are justified depends not only on how they've
been set, but also how they were set before (see the table below).
This looks to be a simple re-draw error.
Tic Justification change Effect
Inside -> Outside correct
Inside->Center correct
Center->Inside leaves the tics centered
Center->Outside correct
Outside->Inside shows tics inside AND outside
Outside->Center correct
================================================================
I...-603
Bug Description: Colormap submenu incomplete when first opened (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
xprism &
Colormaps
What Happened:
The Colormap submenu comes up, but there are no lists of things
under "Autocolor Procedures" or "Colormap Operations".
Note: If something has been put into an area, and that area
is "Clear"-ed or removed, then the Colormap submenu
exists.
Everything works just fine - the colormap lists are displayed as soon as something
is plotted, but then does not disappear. I think this is just an inconsistency if
it is even a bug at all. --squeedle 22 May 97
================================================================
I...-606
Bug Description: Extra Line Drawn (Xprism)
Toolbox: ENVISION
Object: xprism
Full bug descripton:
######################################################################
Bug Report
What I Did:
Plot:
Functions:
f(x) = sin(x)
set X from -pi to pi
PLOT FUNCTION
Close
What Happened:
There is an extra, incorrect line drawn in the lower left quandrant of the
axis. It goes from about (-3, sin(-3)) to (0,0). The line is grey (the
background is black, axis yellow, box green, foreground color magenta).
Pressing "REFRESH" does not fix the problem
What OS/machine:
test status OS machine
---------------------------------------
no OSF piglet
no Solaris tamale
no Linux taishan
no IRIX fresca
occurs AIX shasta
================================================================
I...-609
Bug Description: not able to open classed files
Toolbox: ENVISION
Object: spectrum
Full bug descripton:
######################################################################
Bug Report
when i try to open a classed:file i get always get a dialog box about an error
log. when i read this log
it always says:
SUMMARY:
Of the 0 lines in the Predetermined Classification File,
there were 0 conflicts, 0 UTM coordinates out of bounds.
occasionally when i attemot to open a classed:file, either through the command
line, or through the files
subform, i get another, much more ambigous message, which i cannot recreate
right now. it says something to
the effect of "can only open 0 of the 3 required lines". however, the files
menu still lists this
classed:file as the classification file. i am really not sure what is
happening.
================================================================
I...-611
Bug Description: Composer auto shut-down/write files (Cantata)
Toolbox: DESIGN
Object: NS
Full bug descripton:
#
# Please fill out all blank fields. Only two fields can be longer than
# one line: Work-Around and Bug Report.
#
# For a description of what to put in the Severity, Priority and
# Responsibility fields, please see ~kpro/qualifier/how_to/reporting_bugs.
#
# If you have any questions, see Shannon or Paula.
#
Priority :
Display Depth :
######################################################################
Bug Report
What I Did:
File:
Open: Workspaces:5DHead
clicked the "Comment Icon"
Options: Composer:
Composer:
File Options:
Copy Selected File
~/temp
What Happened:
Composer copied the file, but I couldn't find it where I expected it
to be. Probably composer wrote it to a different directory (but where?).
Then Composer automatically closed itself.
Because it takes a while to load Composer, it is irritating when it
automatically closes itself down, because I have to re-open it before
using it again.
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
untested IRIX fresca
occured AIX shasta
================================================================
I...-612
Bug Description: [Command Icon] glyph not under "Command Icon"
Toolbox: DESIGN
Object: cantata
Full bug descripton:
######################################################################
Bug Report
What I Did:
Glyphs:Program Utilities:Command Icons
What Happened:
I was looking for the [Command Icon] glyph, and it seemed intuitive that
it would be under "Glyphs:Program Utilities:Command Icons", but instead
it was under "Glyphs:Program Utilities:General"
================================================================
I...-617
Bug Description: DESIGN layout/06.layout scrollbar gets changed permanently
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Paula Barth 5/22/97
I verified this bug on fresca.
Although the slider doesn't appear to get small again, as soon as it
is clicked on, or as soon as any other scrollbar is moved, the slider
becomes the correct size. It appears that a refresh isn't being done
when the change from Vertical to Horizontal is being made. I don't
consider this to be a large problem, so I'm makeing it 3a/Low.
######################################################################
Bug Report
If you put 06.layout/example into edit mode & change the scrollbar from
Vertical to Horizontal (or vice versa) then the slider button fills the
scrollbar permanently, that is, if you switch it back it doesn't get
small again.
================================================================
I...-620
Bug Description: DESIGN console is not displaying one thingie per line
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
It appears that you wanted each message to display on a separate line but
it doesn't.
================================================================
I...-623
Bug Description: Strange lifetime for selection buffer
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: OSF1
Compiler: cc Native Fortran
######################################################################
Bug Report
Apparently if your buffer is full with something selected before running
an application, this will be replaced by whatever you select in a Khoros
application, but when the application is finished, two things happen:
the selected text from the application is inserted before the text
from before the application; on the second paste the application
selected text goes away, leaving the original buffer.
================================================================
I...-625
Bug Description: Update README for xvutils/save example
Toolbox: DESIGN
Object: examples
Full bug descripton:
######################################################################
Bug Report
There is no "Ok" button or warn dialog for this example. The description
just needs to be updated is all.
================================================================
I...-626
Bug Description: envision/annotate/05.line should not move line out of window
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: OSF1
Compiler: cc Native Fortran
######################################################################
Bug Report
You should not be able to move the line annotation outside of the display
area. I was able to drag the line to the far right & I couldn't get it
back again.
================================================================
I...-627
Bug Description: annotate/14.image_annotate/ rectangles eat into border
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: OSF1
Compiler: cc Native Fortran
######################################################################
Bug Report
annotate/14.IMAge_annotate/example - in this example, load image:ball
& use locations. click clear & the upper left corner will show that
the border has had a small chunk taken out of it.
================================================================
I...-631
Bug Description: xvobjects/console doesn't clear before printing new msg
Toolbox: DESIGN
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
xvobjects/console/example
What happened:
The console printed a new message every second, but it just strung them
all together in one long text line. From the README, I expected it to
overwrite each second.
================================================================
I...-633
Bug Description: cannot change x and y offsets in putdata
Toolbox: DESIGN
Object: cntata
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
run the command cantata -tbmenus true -commandbar true -console false -wksp workspaces:SimpleArith
from one of the displayed images, select options.
from the options dialog, attempt to change the x and
y offsets. Nothing happens.
If this is not a bug, then if I cannot change something, I
should not be able to see it or it should be grayed out
and no input possible.
================================================================
I...-634
Bug Description: plot/06.plot3d_viff & plot/08.plot3d_colormap shows up white
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
ENVISION example plot/06.plot3d_viff shows up completely white except for
a small corner.
================================================================
I...-638
Bug Description: plot_color/01.2Dfile_fg problems
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: OSF1
Compiler: cc Native Fortran
######################################################################
Bug Report
In ENVISION plot_color/01.2Dfile_fg the beginning & end lines extend outside
the axes. Also if the color of the plot is changed, the color remains the
same for the part outside the axes. Also the "Y Axis" label overwrites a
Y axis value label.
================================================================
I...-643
Bug Description: xvutilities/error can't select text by double clicking
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
xvutilities/error/example
What happened:
even though every other error box let me select text by double clicking
on a word (to select the word) or triple clicking on part of a line to
select the line, this particular one didn't
================================================================
I...-646
Bug Description: xvutilities/query has text bleed-thru on the top of label1
Toolbox: DESIGN
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
xvutilities/query/example
What happened:
The second pane (the one with the labels) has the text "Please personalize
these labels", but half of it is hidden behind the text fill-in line of
label1, and the graphic is partly hidden "behind" the "label1" text.
It is possible to see these things more clearly, by enlarging the window.
================================================================
I...-647
Bug Description: xvutilities/README has extra paragraph
Toolbox: DESIGN
Object: examples
Full bug descripton:
######################################################################
Bug Report
The xvutilities/README file has the following paragraph at the end:
The warn dialog box is displayed until you click on
"Ok"; the example prints your selection to the tty.
This is repeated three times.
Either this paragraph has been separated from the command it was
describing, or it is incorrect. the xvutilities/save/example
doesn't have a warning box, or an "Ok" button to click.
================================================================
I...-648
Bug Description: xvwidgets/05 movement and value are the same values
Toolbox: DESIGN
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
xvwidgets/05.scrollbars1/example
What happened:
As I moved the scrollbars back and forth, two things were reported
to the commandline: "value" and "movement". These two things had
exactly the same data in them, though. Perhaps something else was
meant by "movement"
================================================================
I...-650
Bug Description: annotate/06 circle becomes elipse in some cases
Toolbox: ENVISION
Object: examples
Full bug descripton:
Paula Barth 5/22/97
I verified this on fresca. Some additional comments:
1) The circle remains a circle, until you change the Border Width to
"Extra Wide". Then it becomes an ellipse.
2) If you change the Border Width back to "Extra Fine", the annotation
remains elliptical, but grows larger.
######################################################################
Bug Report
What I did:
annotate/06.circle/example
set Border Color to white
set Radius to 0.5
set X Center to 0.1
set Y Center to 0.1
set Border Width to "Extra Wide"
What happened:
The circle is now an ellipse.
This problem mainly seems to occur for large circles
around the edges of the window.
================================================================
I...-651
Bug Description: annotate/07 [CR] in Height box moves rectangle
Toolbox: ENVISION
Object: examples
Full bug descripton:
Paula Barth 5/22/97
I verified this on fresca. Every time you hit with the cursor
in the Height field, the Y position increments about 0.002. Very strange...
######################################################################
Bug Report
What I did:
annotate/07.rectangle/example
shift-clicked on the "parent" window (to put it into edit mode)
resized the rectangle using the mouse in the "parent" window
put the cursor in the "Height" text area on the "Rectangle Object Menuform"
hit return
What happened:
even though I wasn't changing ANY values, every time I hit return while I
was in the Height text box, the rectangle moved up and to the left a little
bit.
================================================================
I...-653
Bug Description: annotate/13 Border Color "blue" but really black
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
annotate/13.polyline/example
What happened:
the Polyline Object Menuform says
Border Color is blue
but the actual image, clearly has black lines
================================================================
I...-655
Bug Description: color/autocolor/2 & color/autocolor/3 has no image
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
color/autocolor/2.autocolor_readwrite/example
What happened:
There's no image displayed in the box, so changing the coloring schemes
has no effect.
================================================================
I...-656
Bug Description: color/printmapval/1 seg faults
Toolbox: ENVISION
Object: examples
Full bug descripton:
Paula Barth 5/22/97
I tested this on fresca. The example program ran fine. The labels are
showing up just fine.
HOWEVER, when I testing this on kafka, I got a segmentation fault on
start-up!
######################################################################
Bug Report
What I did:
color/printmapval/1.printmapval_display/example
What happened:
the position object has three grids, the top one is labeled "Red",
and it appears that the bottom two should be labeled "Green" and "Blue"
respectively, but the place where the words appear is blacked out.
Probably, just changing the background color for those two pieces of
text (to non-black) would fix this problem.
================================================================
I...-660
Bug Description: color/pseudo/7 can't enlarge "pseudo" window
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
color/pseudo/7.pseudo_readwrite/example
What happened:
I can't resize the "pseudo" window, so I can't see the palette
very clearly.
================================================================
I...-661
Bug Description: color/pseudo/7 no image
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
color/pseudo/7.pseudo_radwrite/example
What happened:
No image was loaded, so no change in color has any effect.
================================================================
I...-663
Bug Description: image/05 didn't enlarge the whole picture
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
image/05.enlarge/example
What happened:
Although the ball appears to have been enlarged correctly, the floor
under the ball does not appear in this picture. It appears to have been
cropped, which wasn't listed in the README as one of the things that this
program was supposed to do.
================================================================
I...-665
Bug Description: image/11 doesn't update screen correctly
Toolbox: ENVISION
Object: examples
Full bug descripton:
Paula Barth 5/22/97
Verified this on fresca and kafka. Here's what I found:
1) The white box in the pan icon doesn't show up unless you minimize
the window and then re-open it.
2) The pan icon doens't appear to respond to the mouse. If you try to move
the white box nothing happnes. But, if you minimize the window
then re-open the window, the pan icon is now where you moved it.
3) Also, once you have moved the pan icon, the wander mechanism doesn't
wander very far. Before trying to move the icon, things were wandering
large distances...
######################################################################
Bug Report
What I did:
image/11.USA_wander/example
What happened:
After the image loaded, I clicked on Florida (on the pan icon), but the
image never loaded that part of the map.
Once I minimized and re-opened the window, it did, however, show Florida
Also, I'm not sure whether the black box in the pan icon is actually moving
as the focus of the image "wanders". If it is moving, it is moving only
a tiny amount, and I can't detect it.
================================================================
I...-667
Bug Description: imgicon/1 doesn't recover well from being made smaller
Toolbox: ENVISION
Object: examples
Full bug descripton:
Paula Barth 5/22/97
I verified this bug on fresca. It is exactly as LaMaia describes it.
######################################################################
Bug Report
What I did:
imgicon/1.imigcon_display/example
made the window a little bit smaller
What happened:
The window bounced back to it's default size, but the image inside
of it has been somewhat corrupted. Even refreshing it, or minimizing
it (then returning it to normal) doesn't repair the image.
However, stretching the window to be larger than default, and allowing
it to bounce back to it's normal dimensions DOES restore the image.
================================================================
I...-668
Bug Description: miscellaneous/01 shrinking window shrinks plot
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
miscellaneous/01.image_plot2d/example
made window somewhat smaller
What happened:
Although the window sprang back to normal size, the 2d-plot
on top of the image shrank. The image remained normal size.
================================================================
I...-669
Bug Description: miscellaneous/01 plot disappears when new image loaded
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
What I did:
miscellaneous/ 01.image_plot2d/example
opened image:VW
What happened:
the 2D plot disappeared entirely, when I opened a new image
================================================================
I...-670
Bug Description: Spectrum's Legend subform needs cosmetic work
Toolbox: ENVISION
Object: spectrum
Full bug descripton:
Operating System Version: SunOS
Compiler: cc f2c
Perl Version:
X Version:
######################################################################
Bug Report
The Legend subform of spectrum needs cosmetic work.
1) The RGB scrollbars come out in red, white, and blue rather than
red, green, and blue. patriotic but incorrect.
2) The layout of the RGB scrollbars isn't so great; the top part
of the red scrollbar is truncated
2) The legend entries look ok, but the size of the colorboxes will
shrink the first time a cluster is added to that class.
================================================================
I...-671
Bug Description: Default values for doubles lose precision in .pane file.
Toolbox: DESIGN
Object: xvforms
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: OSF1
Compiler: cc Native Fortran
######################################################################
Bug Report
When you define a double precision variable for a gui
using Guise and give it a default value with more than
6 digits of precision, it loses precision by the time
it is written to the pane file. If you edit the pane
file by hand, it appears to work ok, i.e. no precision
is lost when the default value is read.
To reproduce:
1. create a kroutine
2. edit the pane file with Guise
3. add a double precision value to your pane
4. make it optional and specify a default value
of 1.23456789012
5. save everything and generate code
6. look at the entry in the pane file, it will
be 1.23456000 (or something similar)
================================================================
I...-674
Bug Description: plot_color/README
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: IRIX
Compiler: cc Native Fortran
######################################################################
Bug Report
The README says:
NOTE: None of these examples use tacking. Thus, if the window is interactively
resized using the window manager, the plot will not grow with the window.
Nonetheless, each of the plots DOES grow when the window is resized, and
they certainly shrink when the window is made smaller.
================================================================
I...-678
Bug Description: Mosaic & Netscape selections problems
Toolbox: DESIGN
Object: composer
Full bug descripton:
######################################################################
Bug Report
In Composer the menu behaves strangely if you are attempting to select
Mosaic from the menu, and neither Mosaic nor Netscape are being executed.
================================================================
I...-680
Bug Description: Problem with putmapval map display resizing
Toolbox: ENVISION
Object: examples
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: V3.2
Compiler: cc Native Fortran
######################################################################
Bug Report
In running putmapval, I accidentally entered a Height of 110. After waiting
a long time it finally displayed properly, but then when I changed the
height to 11 the window never resized properly.
================================================================
I...-682
Bug Description: While deleting bogus object in craftsman no busy window displayed
Toolbox: DESIGN
Object: craftsman
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: Linux
Compiler: gcc f2c
######################################################################
Bug Report
I had a bogus object that craftsman asked if it could delete. I said yes and
it went about doing it. But no busy window brought up.
================================================================
I...-684
Bug Description: klint'ing from craftsman does not know all the objects
Toolbox: DESIGN
Object: craftsman
Full bug descripton:
Advanced Khoros Bronze Release 2.2.0.0 Bug Reporting Form
Operating System Version: Linux
Compiler: gcc f2c
######################################################################
Bug Report
After creating a xvroutine object and pane object I did a klint toolbox and
got the message in the console window:
klint -tb KK------------------------------------------------
object sdf does not exist in KK
You get the some error if you select any object and do a klint object.
Running klint -tb kk from the command line works fine.
================================================================
I...-685
Bug Description: output is not labeled
Toolbox: DESIGN
Object: cantata
Full bug descripton:
######################################################################
Bug Report
in a workspace that produces multiple outputs, such as complex_arithmatic,
it would be mucho mas better if the outputs were labeled... so that you
were not just staring at a bunch of pictures wondering what they were or
which procedure generated which image.
also, it would be nice if there was a way to close all of the output images
at once... more than 2 can get tedious
================================================================
I...-687
Bug Description: those icky faces don;t go away when you reset
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
when the user is running in non-continunous mode-
if the user does something to cause those icky faces, and then
resets, they are still there. it is unclear which parts of cantata
are part of the workspace, and which are part of the environment.
if the user resets, the error condition should go away... since it
does not the user might conclude that the program itself is in error
and shut down. if the user does try to "fix" the error, the only way
to test this is to run the process, which can take a long long time.
in continuous mode, this problem is "solved" because the process is
always updating itself.
cantata should have a way to determine what types of input files are
acceptable... many processes take so long to run, that they should not
be the only available trouble-shooting method.
================================================================
I...-688
Bug Description: redraws problems, especially w/ grid=1 and redraw after move
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
the workspace area often has problems redrawing, especially when the grid is on
and preferneces are set to redraw connection after glyph movement. also, when
other windows are placed in front of the cantata window, it loses most of ht
:auxillary: information hidden behind them. the glyohs themselvs never get
corrupted.
pressing reset does not help.
pressing "redraw workspace" wipes the screen completely.
================================================================
I...-690
Bug Description: cant' delete glyphs inside a windowed procedure
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
opened up infile_var.wk
created a procedure inside a while loop
added and then tried to delete a compress colors glyoh
from the embedded procedure. was not able to delete the
glyph while the procedure was a its own window.
[report text here]
================================================================
I...-691
Bug Description: cannot reconnect a glyph to output inside a procedure
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
in the infile_var.wk created an embedded proceedure indie a while loop.
added a few other glyphs that take input from the absolute value glyph.
this is the glyoh that slso feeds the output. i then broke the connection
between the abs value glyoh and the little external output icon. i could not
reinstate the connection untill i deleted all the other connections that were
coming out of this glyph. after i reconnected the glyph to the external output
icon, i was able to reconnect the glyoh to all the other icons i created.
see /research/sysadmin/sarahb/bronze_qualifier/for a copy of the files i created.
infile_var2.wk and infile_var3.wk
================================================================
I...-692
Bug Description: info glyphs dissapear too fast
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
when running a workspace
occasionally info. icons (those big blue i things) appear beneath
a glyph but disappear within seconds. sometimes they just go away
sometiumes they are replaced by an icky face. i would really like
to know what these glyphs are saying
================================================================
I...-699
Bug Description: seg fault with puticon
Toolbox: ENVISION
Object: puticon
Full bug descripton:
Paula Barth 5/19/97
I verified this behavior running & displaying on naduah. It actually goes
so far as to dump a core... It doens't seem to happen on any other
architecture.
######################################################################
Bug Report
with command line
puticon -i /research/devel/mirage/testimages/data/images/florida.kdf -priv y
-update 2.0003938
an outline of an image does appear but the process then seg faults
Note: MY & DA May 19, 1997
- it only happens on naduah's display
- it doesn't have to do with [-priv]
- it doesn't have to do with [-update]
- it only happens with image:florida
- it doesn't happen with putdata -imgicon
================================================================
I...-701
Bug Description: editimage gui "out of order"
Toolbox: ENVISION
Object: editimage
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
the gui subform (superform??) is in order
file
colormap
clip
the files subform is in order
file
clip
colormap
================================================================
I...-702
Bug Description: incorrect behavior with editimage gui
Toolbox: ENVISION
Object: editimage
Full bug descripton:
Khoros Product : Khoros
# #####################################################################
Bug Report
used gui to run following:
+ editimage -i image:khoros -gui
$ENVISIONBIN/editimage -i /research/release/kpro/sampledata/data/images/
ankle_complex.xv -cmap /research/release/kpro/sampledata/data/images/myelin.kdf
-clip /research/release/kpro/sampledata/data/images/ball.xv -priv 0
-alloc 1 -x -1 -y -1 -update 1
(formatted for your reading ease)
thats input file image:ankle_MRI
input colormap image:fat
input clipgate image:ball
get a weird looking picture (of course)
just run edit image with out the gui and use the files subform to
execute the same commands
different weird picture
also try this using
inout image image:mandril
colormap image:gull
also neither of these colorschemes seems to be at all related
to the colormap image (or to the parent for that matter)
================================================================
I...-709
Bug Description: using answerfile to give invalid command doesn't give error
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Khoros Product : Khoros
######################################################################
Bug Report
cantata -wksp workspaces:Loops
gives error that workspace does not exist and runs empty cantata
cantata -wksp workspaces:Loops -A /usr/tmp/test.ans
does not give any error that workspace does not exist and then runs cantata
cantata -a /usr/tmp/test.ans
does not give any error that workspace does not exist and then runs cantata
================================================================
I...-711
Bug Description: cantata cannot open a file
Toolbox: DESIGN
Object: cantata
Full bug descripton:
######################################################################
Bug Report
>From the top level of mirage, I ran cantata and
then attempted to open a workspace file. When
I selected open from the file pulldown menu,
I always got the following error dialog.
Library: xvforms
Routine: xvf_action_cb
Cancelling action Open...
John Sturtevant
================================================================
I...-712
Bug Description: Incorrect behavior for adjacent pull down menus
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
######################################################################
Bug Report
run the example program envision/examples/annotate/13.polyline.
in it you will find two pull down menus for line type and line
width that occur one on top of the other. click in the top one
and slowly pull down to one of the selections. You will instead
have the lower menu popup. Extremely annoying. this problem is
due to the fact that there is approcimately a 3-4 pixel separation
between the main menu button and the pull down portion. Then the
pointer enters that separation, it is entering the main button
for the lower menu and thus it pops up. If you pull down really
fast, this does not happen.
this also happens on the sun.
================================================================
I...-716
Bug Description: A glyph sometimes core dumps running inside a procedure
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Subject: Re: cantata dumps on creating procedure under SunOS 4.1
Newsgroups: comp.soft-sys.khoros
References: <5lfq1m$e0a$1@crawfish.nvl.army.mil>
Organization: Medical College of Wisconsin - Milwaukee, WI
Distribution: usa
Operating System Version: 4.0
Bug Description:
Timothy Williams (williams@nvl.army.mil) wrote:
: Anybody else having a problem with cantata dumping core when creating
: a procedure or restoring a workspace with a procedure in it? (Even
: very simple procedures!) I know I've created procedures under Khoros
: 2.0.2, but I don't have that version installed anymore to test that.
Tim,
Not exactly, but in the last couple of days I have come across a
cantata bug which may have something to do with what you are
describing. I haven't noticed this behavior previously, and the
workspace in which it occurs does contain a procedure. Sometimes this
workspace runs without difficulty, but on other occasions several
types of problem can occur. One problem is that, after clicking on a
glyph to start it, the center of the glyph turns red, but the
program represented by the glyph does not begin running immediately.
After some delay it does run, but after it completes there is again
a delay before cantata acknowledges completion. (I observed this
pattern by using top to show which programs are actually running.)
Clicking the stop button in cantata (temporarily) corrects this behavior.
The second problem is actual core dumps, which have in fact occurred
within the procedure.
I have a shell script which does the same thing as this workspace, and
it runs without difficulty, so I would think this is a cantata issue.
I haven't had any difficulty creating or loading a procedure, so this
may very well be an unrelated matter, but your post gave me an excuse
to bring it up.
I have only tried this under Linux, but did get the same result on two
machines running slightly different versions of Linux.
Good luck,
Matthew Fleming
For what its worth, this problem has been observed with Khoros 2.1 on
Caldera Network Desktop and Red Hat Linux 4.1. Both installations used
X11R6 and Athena. The CND system includes gcc v. 2.7.0, the Red Hat has
something like gcc 2.7.2.1. Here is the workspace:
# Khoros Visual Programming Workspace
#
# cantata workspace file (/home2/k22/dl2.wk) was created
# on Fri May 16 08:41:55 1997
# by user k21
#
WorkspaceBegin: 2.1
VariablesBegin:
# No variables currently declared. #
VariablesEnd:
# Glyph 'User defined'
Glyph:DESIGN:user_def::localhost:1:22:22:: -o
/home2/k22/new_cogimgs/c104.vif
# Glyph 'Remove Segments'
Glyph:DATAMANIP:krmseg::localhost:3:122:22:: -loc
[k21@noname k22]$ cat dl2.wk
# Khoros Visual Programming Workspace
#
# cantata workspace file (/home2/k22/dl2.wk) was created
# on Fri May 16 08:41:55 1997
# by user k21
#
WorkspaceBegin: 2.1
VariablesBegin:
# No variables currently declared. #
VariablesEnd:
# Glyph 'User defined'
Glyph:DESIGN:user_def::localhost:1:22:22:: -o
/home2/k22/new_cogimgs/c104.vif
# Glyph 'Remove Segments'
Glyph:DATAMANIP:krmseg::localhost:3:122:22:: -loc
# Glyph 'RGB to XYZ'
Glyph:COLOR:cpRGBNTSCtoXYZ::localhost:5:222:22::
# Glyph 'Set Attribute'
Glyph:DATAMANIP:ksetdattr::localhost:7:322:22:: -cspace 8
# Glyph 'Extract'
Glyph:DATAMANIP:kextract::localhost:9:422:22:: -tsize '1' -esize '1'
-woff '0' -hoff '0' -doff '0' -toff '0' -eoff '0' -subpos 0
# Glyph 'Map Data'
Glyph:DATAMANIP:kmapdata::localhost:11:502:22::
# Glyph 'Unmask Data'
Glyph:DATAMANIP:kunmask::localhost:13:202:142:: -real '0' -imag '0'
-replace
# Procedure 'lesional segmentation'
ProcedureBegin:lesional segmentation:(unknown):localhost:15:42:102:
VariablesBegin:
# No variables currently declared. #
VariablesEnd:
# Glyph 'Subtract From'
Glyph:DATAMANIP:ksubfrom::localhost:1:662:122:: -real '0'
# Glyph 'Insert Segments'
Glyph:DATAMANIP:kinsertseg::localhost:3:722:182:: -val
# Glyph 'Copy from Value'
Glyph:DATAMANIP:kcpfromval::localhost:5:542:202:: -mask
# Glyph 'Close Holes'
Glyph:MMACH:vclohole::localhost:7:462:142::
# Glyph '3x3 Str.El.'
Glyph:MMACH:vstrflat::localhost:9:422:222:: -s00 '1' -s10 '1' -s20 '1'
-s01 '1' -s11 '1' -s21 '1' -s02 '1' -s12 '1' -s22 '1'
# Glyph 'Normalize'
Glyph:DATAMANIP:knormal::localhost:11:382:142:: -lval '0' -uval '255'
-type 3 -w -h
# Glyph 'Thresh Inside'
Glyph:DATAMANIP:kthreshin::localhost:13:302:142:: -uc '1' -lc '3'
-fval '255' -tval '0'
# Glyph 'Labeling'
Glyph:RETRO:vlabel::localhost:15:222:142:: -w '0' -m 1 -c 1 -s '1'
-merge 0 -f '0.07'
# Glyph 'K-Means'
Glyph:DATAMANIP:kkmeans::localhost:17:122:142:: -map -n '50000' -k '2'
# ExtPort 'External Input (1)'
ExtPort:in:External Input (1):i1:19:565:147:
# ExtPort 'External Output (2)'
ExtPort:out:External Output (2):o1:21:859:207:
# ExtPort 'External Input (3)'
ExtPort:in:External Input (3):i1_2:23:25:167:
# Connections 'Subtract From'
NodeConnection:data:1:o:3:i1
# Connections 'Insert Segments'
NodeConnection:data:3:o:21:
# Connections 'Copy from Value'
NodeConnection:data:5:o:3:i2
# Connections 'Close Holes'
NodeConnection:data:7:o:5:i
# Connections '3x3 Str.El.'
NodeConnection:data:9:o:7:str
# Connections 'Normalize'
NodeConnection:data:11:o:7:i
# Connections 'Thresh Inside'
NodeConnection:data:13:o:11:i
# Connections 'Labeling'
NodeConnection:data:15:o:13:i
# Connections 'K-Means'
NodeConnection:data:17:o1:15:i1
# Connections 'External Input (1)'
NodeConnection:data:19::1:i1
# Connections 'External Output (2)'
# Connections 'External Input (3)'
NodeConnection:data:23::17:i1
AnnotationsBegin
AnnotationsEnd
ProcedureEnd:lesional segmentation
# Glyph 'Detect lines'
Glyph:DERMATOSCOPY:dlines::localhost:17:342:162:: -s '2.000' -l
'0.000' -u '1.000' -bd 1 -correct 1 -junctions 1
# Glyph 'Convert Type'
Glyph:DATAMANIP:kconvert::localhost:19:82:242:: -type 3 -scale '1'
-roff '0' -ioff '0' -val
# Glyph 'Convert Type'
Glyph:DATAMANIP:kconvert::localhost:21:442:162:: -type 3 -scale '1'
-roff '0' -ioff '0' -val
# Glyph 'Maximum'
Glyph:DATAMANIP:kmaximum::localhost:23:222:242::
# Glyph 'Display Image'
Glyph:ENVISION:putimage::localhost:25:142:322:: -xoffset '0' -yoffset
'0' -complex 5 -normaltype 1 -normalmethod 1 -redcol 1 -greencol 2
-bluecol 3 -redfn 'M0' -greenfn 'M1' -bluefn 'M2'
# Glyph 'Normalize'
Glyph:DATAMANIP:knormal::localhost:27:282:102:: -lval '0' -uval '255'
-type 3 -w -h
# Glyph 'Edit Image'
Glyph:ENVISION:editimage::localhost:44:282:322:: -priv 0 -alloc 1 -x
'-1' -y '-1' -update '1' -i image:khoros
# Glyph 'Edit Image'
Glyph:ENVISION:editimage::localhost:46:422:322:: -priv 0 -alloc 1 -x
'-1' -y '-1' -update '1' -i image:khoros
# Connections 'User defined'
NodeConnection:data:1:o:3:i
# Connections 'Remove Segments'
NodeConnection:data:3:o:5:i
NodeConnection:data:3:o:15:i1_2
NodeConnection:data:3:o:25:i
# Connections 'RGB to XYZ'
NodeConnection:data:5:o:7:i
# Connections 'Set Attribute'
NodeConnection:data:7:o:9:i
# Connections 'Extract'
NodeConnection:data:9:o:11:i
# Connections 'Map Data'
NodeConnection:data:11:o:15:i1
NodeConnection:data:11:o:19:i
# Connections 'Unmask Data'
NodeConnection:data:13:o:27:i
# Connections 'lesional segmentation'
NodeConnection:data:15:o1:13:i
# Connections 'Detect lines'
NodeConnection:data:17:o1:21:i
# Connections 'Convert Type'
NodeConnection:data:19:o:23:i2
# Connections 'Convert Type'
NodeConnection:data:21:o:23:i1
# Connections 'Maximum'
# Connections 'Display Image'
# Connections 'Normalize'
NodeConnection:data:27:o:17:i
# Connections 'Edit Image'
# Connections 'Edit Image'
AnnotationsBegin
AnnotationsEnd
WorkspaceEnd: 2.1
================================================================
I...-720
Bug Description: Info window needs a scrollbar
Toolbox: DESIGN
Object: kwidgets
Full bug descripton:
######################################################################
Bug Report
Whenever I create a kroutine with an associated library in craftsman,
it puts up a nice, informative, long message telling the user what
files were changed (toolbox.def, include file, makefiles, etc).
This message is so long that it scrolls off the window and most of it
can't be seen. I happen to know that you can click in the window and
use the down arrow key to see the rest of the message, but this isn't
intuitive. A scrollbar would be much better.
================================================================
I...-722
Bug Description: First Entry in "File Operations" behaves oddly (Composer)
Toolbox: DESIGN
Object: composer
Full bug descripton:
######################################################################
Bug Report
What I Did:
created a toolbox called "Training" (Lab 1 of Software Development Course)
created an object called "testk" (Lab 2 of Software Development Course)
pulled down menu from "File Operations"
What Happened:
When List Files is set to UIS
The first entry (Guise in this case) behaves oddly:
it can be selected as well as the 2nd entry (Edit, in this case)
when the cursor is right on top of "Guise", it is NOT selected
the mouse has to be moved up or left
When List Files is set to SOURCE
The first entry is BLANK, which looks odd,
When List Files is set to CONFIG
The first two entries are blank
When List Files is set to INFO
the first line is blank
it is possible to have 3 selections (including blank) selected, even if
they are a few spaces apart
When List Files is set to APP-DEFAULTS
it's possible to select ALL the options on the pulldown menu
When List Files is set to PIXMAPS
it's possible to select ALL the options on the pulldown menu
When List Files is set to MISC
it's possible to select ALL the options on the pulldown menu
When List Files is set to ALL
it's possible to select ALL the options on the pulldown menu
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
didn't Solaris tamale
didn't Linux taishan
didn't IRIX fresca
didn't AIX shasta
All of the platforms had places where there was a blank selection from
the pull down menu, but only OSF had the other problems.
================================================================
I...-724
Bug Description: composer seg faults don't reach command line when spawned by Craftsman
Toolbox: DESIGN
Object: composer
Full bug descripton:
######################################################################
Bug Report
What I Did:
created a toolbox called "Training" (Lab 1 of Software Development Course)
created an object called "testk" (Lab 2 of Software Development Course:step 1)
craftsman
selected "training"
selected "testk"
Object Operations:edit (composer)
placed composer window
created a seg fault in composer (already reported)
What Happened:
Although composer seg faulted, that data didn't go to the command line. I had
to run
composer independently in order to find that information out.
What OS/machine:
test status OS machine
---------------------------------------
occured OSF piglet
occured Solaris tamale
occured Linux taishan
occured IRIX fresca
occured AIX shasta
================================================================
I...-725
Bug Description: information text behaves strangely
Toolbox: TRAINING
Object: testlk
Full bug descripton:
######################################################################
Bug Report
What I did:
craftsman &
created a toolbox called "training" (Software Development Course lab 1)
selected the toolbox "training"
Object Operations:Create Object
on the "Create a New Object" subform, filled out:
Object Name = testlk
Category = Training
Subcategory = LKroutine
Short Description of Object = this is testlk
Install in Cantata? = Yes
Create Library Routine? = Yes
Library Name = testlib
clicked [Create KROUTINE]
What happened:
got the following dialog box:
----------------------------------------------------------------------------
Operation: generate software object
Toolbox: TRAINING
Object: testlk
You specified that the kroutine should have an associated library routine,
in library testlib.
This library does not exist in toolbox `TRAINING'.
Do you want to create the library in the toolbox?
----------------------------------------------------------------------------
clicked [Create]
got the following dialog box:
----------------------------------------------------------------------------
Do you want to make the TRAINING toolbox
depend on the testlib library?
----------------------------------------------------------------------------
clicked [Yes]
craftsman worked on this for a while, and then I got the following:
----------------------------------------------------------------------------
Since the `testlib' library has been added to
toolbox `TRAINING',
the following has been done:
1. An AddLibrary(testlib, ...) line has been
added to
$TRAINING/repos/config/imake_config/toolbox.def
2. An include statement for the library:
#include
has been added to
$TRAINING/include/training.h
3. All Makefiles in the TRAINING toolbox were
rebuilt.
This assumes it's a C, C++, or Fortran library and
no previously defined libraries depend on it.
If it's an X library or the ordering needs to be
changed, you must edit these files manually and
regenerate Makefiles.
----------------------------------------------------------------------------
The Problem:
The information box (above) allowed me to use the up and down arrows
to move through the text, but it didn't let me
double-click to select a word
triple-click to select a line
quadruple-click to select all the text.
When I resized the window, the area where the text was didn't resize.
Where this occured:
piglet OSF
tamale Solaris
taishan Linux
fresca IRIX
shasta AIX
================================================================
I...-730
Bug Description: Procedure glyph does not reflect state
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
This is not architecture dependent. If a procedure is running but has
been started while the procedure window is open, this state is not
reflected on the Procedure glyph when its window is closed.
================================================================
I...-731
Bug Description: Can't see menu options that are off the screen
Toolbox: NS
Object: NS
Full bug descripton:
######################################################################
Bug Report
If you have a menu that is so large it goes off the top and bottom of your
screen, there is no way to select the hidden options.
Where can I get a menu that is so large it behaves in such a way you ask?
Well, try the following:
1) % craftsman
2) From the Options menu, select Preferences
3) Set (Sub)Catagories from to "all toolboxes"
4) Close the Preferences subform
5) From the Object Operations menu, select Create Object
6) When the Create Object subform pops up, click on the Subcategory
button. It puts up a huge menu, half of which you can't see.
================================================================
I...-734
Bug Description: There is a bug in the histogram glyph in starting value.
Toolbox: DATAMANIP
Object: khistops
Full bug descripton:
######################################################################
Bug Report
There is a bug in the histogram glyph in
starting value.
Please see the workspace below.
the first case, starting value is 0
the second case starting value is 1 (there is a bug)
This bug only happens if bin width is different from 1. In the
case shown, bin width is 2.
--roberto
===========
# Khoros Visual Programming Workspace
#
# cantata workspace file (histogram-bug.wk) was created
# on Wed May 21 17:24:14 1997
# by user lotufo
#
WorkspaceBegin: 2.1
VariablesBegin:
# No variables currently declared. #
VariablesEnd:
# Glyph 'Piecewise Linear'
Glyph:RETRO:vgpwl::localhost:1:23:43:: -r '100' -s '1' -c '256' -l '0' -b '1' -m '256' -xp '256' -yp '1' -xr '256' -yr '0' -xf '0' -yf '0' -xw '0' -yw '0' -t 'float'
# Glyph 'Convert Type'
Glyph:DATAMANIP:kconvert::localhost:3:103:43:: -type 3 -scale '1' -roff '0' -ioff '0' -val
# Glyph 'Histogram'
Glyph:DATAMANIP:khisto::localhost:6:183:43:: -bins '256' -binwidth '4' -min '0.0000000000' -w -h
# Glyph 'Supported Formats'
Glyph:DATAMANIP:kformats::localhost:9:263:43:: -ascii
# Glyph 'File Viewer'
Glyph:DESIGN:khelp::localhost:12:343:43::
# Glyph 'File Viewer'
Glyph:DESIGN:khelp::localhost:15:342:122::
# Glyph 'Supported Formats'
Glyph:DATAMANIP:kformats::localhost:17:262:122:: -ascii
# Glyph 'Histogram'
Glyph:DATAMANIP:khisto::localhost:19:182:122:: -bins '256' -binwidth '4' -min '1' -w -h
# Connections 'Piecewise Linear'
NodeConnection:data:1:o:3:i
# Connections 'Convert Type'
NodeConnection:data:3:o:6:i
NodeConnection:data:3:o:19:i
# Connections 'Histogram'
NodeConnection:data:6:o1:9:i
# Connections 'Supported Formats'
NodeConnection:data:9:o:12:i
# Connections 'File Viewer'
# Connections 'File Viewer'
# Connections 'Supported Formats'
NodeConnection:data:17:o:15:i
# Connections 'Histogram'
NodeConnection:data:19:o1:17:i
AnnotationsBegin
AnnotationsEnd
WorkspaceEnd: 2.1
================================================================
I...-737
Bug Description: FMM when quitting Animate
Toolbox: ENVISION
Object: animate
Full bug descripton:
Statement of problem...
Ran animate under purify, & got FMM every time I quit Animate, even without
doing anything first.
================================================================
I...-739
Bug Description: problem loading Feedback work space using CC compiler
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
It seems on poopsie that when you try to load the Feedback workspace there
is an infinite loop in the expression parser. I tested it on dew, piglet,
kafka and zen without any problems. Both kpro and mirage have the bug. To
test just start up cantata on poopsie and do a load workspace of Feedback.
Here is the dbx trace back of where cantata hangs:
[1] 0xa199c(0x0, 0xee6dc264, 0x400, 0xa199c, 0xee600000, 0x0), at 0xa199b
[2] _p956static(0x0, 0xee6dc264, 0x400, 0x0, 0xffffffff, 0xbc648), at 0x23d6c
[3] 0xa1950(0x0, 0xee6dc264, 0x400, 0x0, 0xffffffff, 0xbc648), at 0xa194f
[4] _libc_read(0x0, 0xee6dc264, 0x400, 0x0, 0xffffffff, 0xbc648), at 0xee654c6
c
[5] __filbuf(0xbc508, 0x0, 0x0, 0x0, 0x0, 0x400), at 0xee6a9834
[6] kexpr_look(), at 0xeef4eae0
[7] kexpr_lex(), at 0xeef4cce0
[8] kexpr_parse(), at 0xeef46218
=>[9] kexpr_eval_expression(id = 3312616, mystr = 0x9067c8 "LEVEL = 0", error =
0xefffb10c ""), line 637 in "/poopsie/compiles10/kpro/bootstrap/sol2.5-CC/object
s/library/kexpr/src/utilities.c"
[10] kexpr_evaluate_generic(id = 3312616, string = 0xefffb50c "LEVEL = 0", typ
e = 8192, value = (nil), error = 0xefffb10c ""), line 755 in "/poopsie/compiles1
0/kpro/bootstrap/sol2.5-CC/objects/library/kexpr/src/routines.c"
[11] WorkspaceRestore(widget = 0x3328b8, file = 0xbc5c00, xoffset = 0, yoffset
= 0), line 326 in "/poopsie/compiles10/kpro/design/sol2.5-CC/objects/library/xv
lang/src/WorkRestore.c"
[12] WorkspaceActions(workspace = 0x328be8, attribute = 0x16a720 "workspaceRes
tore", calldata = 0xefffe714), line 1124 in "/poopsie/compiles10/kpro/design/sol
2.5-CC/objects/library/xvlang/src/WorkUtil.c"
[13] xvw_set_arg(object = 0x328be8, name = 0xef775010 "workspaceRestore", list
= 0xefffe9f0, indx = 0xefffe790, maxargs = 50, args = 0xefffe808), line 186 in
"/poopsie/compiles10/kpro/design/sol2.5-CC/objects/library/xvwidgets/src/attribu
tes.c"
[14] xvw_arguments(object = 0x328be8, list = 0xefffe9f0, args = 0xefffe808, ma
xargs = 50, routine = &`attributes.c`xvw_set_arg(_xvobject*, char*, void**, int*
, int, Arg*) at 0xef2aaa38, numargs = 0xefffe99c), line 515 in "/poopsie/compile
s10/kpro/design/sol2.5-CC/objects/library/xvwidgets/src/attributes.c"
[15] xvw_vset_attributes(object = 0x328be8, list = 0xefffea68), line 597 in "/
poopsie/compiles10/kpro/design/sol2.5-CC/objects/library/xvwidgets/src/attribute
s.c"
[16] xvw_set_attributes(object = 0x328be8, ...), line 718 in "/poopsie/compile
s10/kpro/design/sol2.5-CC/objects/library/xvwidgets/src/attributes.c"
[17] CommandbarCallbacks(object = 0x2ab930, client_data = 0xef774e60, call_dat
a = (nil)), line 599 in "/poopsie/compiles10/kpro/design/sol2.5-CC/objects/libra
ry/xvlang/src/CommandBar.c"
[18] xvw_callback_intermediary(widget = 0x2c13c8, client_data = 0x2c4518, call
_data = (nil)), line 100 in "/poopsie/compiles10/kpro/design/sol2.5-CC/objects/l
ibrary/xvwidgets/src/callback.c"
[19] NotifyHandler(widget = 0x2c13c8, event = 0xeffff338, params = (nil), num_
params = 4003437036), line 649 in "/poopsie/compiles10/kpro/design/sol2.5-CC/obj
ects/library/kwidgets/src/Button.c"
[20] HandleActions(0x2c13c8, 0xeffff338, 0x17b068, 0x0, 0x780b4c, 0xee9f99e4),
at 0xee9d8d14
[21] HandleSimpleState(0x1, 0x0, 0x1, 0x0, 0x0, 0x780b4c), at 0xee9d99c0
[22] _XtTranslateEvent(0x2c13c8, 0xeffff338, 0x8000, 0x5, 0x0, 0x2c13f8), at 0
xee9da4a4
[23] DispatchEvent(0xeffff338, 0x2c13c8, 0x8, 0x2c49d8, 0xef370a48, 0x1f), at
0xee99f654
[24] DecideToDispatch(0xeffff338, 0x10567c, 0x8, 0x2c13c8, 0x2c13c8, 0x0), at
0xee9a03b4
[25] XtDispatchEvent(0xeffff338, 0x0, 0xf3d10, 0x0, 0x1, 0xf2948), at 0xee9a04
e8
[26] XtAppProcessEvent(0xf2948, 0x0, 0xf2b14, 0xf2954, 0xf2b10, 0x1), at 0xee9
b12f0
[27] process_event(block = 1), line 84 in "/poopsie/compiles10/kpro/design/sol
2.5-CC/objects/library/xvwidgets/src/dispatch.c"
[28] xvw_process_event(), line 139 in "/poopsie/compiles10/kpro/design/sol2.5-
CC/objects/library/xvwidgets/src/dispatch.c"
[29] xvf_run_form(), line 71 in "/poopsie/compiles10/kpro/design/sol2.5-CC/obj
ects/library/xvforms/src/runform.c"
[30] main(argc = 1, argv = 0xeffff634), line 152 in "cantata.c"
================================================================
I...-744
Bug Description: ABR when quitting Spectrum
Toolbox: ENVISION
Object: Spectrum
Full bug descripton:
Statement of problem...
To reproduce, just run Spectrum under Purify (kafka) and quit.
This was thought to have been fixed, but it wasn't.
================================================================
I...-745
Bug Description: cantata cores dumps in run mode after selecting the clear button
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
JHS 5/28/97
I was able to get cantata to core dump once after many tries at reproduction
of this bug. For me it cored when I clicked on the black triangle of
a glyph. this would suggest that perhaps some memory is not getting
cleaned up correctly or possibly referenced after being freed.
-------------
cantatta cores dump if you put a workspace in
the run mode and then hit the clear workspace button.
It is not always, but if you try harder you can get it.
Try hiting the clear button in the middle of showing up
displays, file viewers etc.
I could not get a precise recipie to make this reproductible
but 4 in 10 times, it cored dump.
================================================================
I...-746
Bug Description: No shading in plot3d. No interactivity on Interactive 2d/3d plot
Toolbox: ENVISION
Object: xprism
Full bug descripton:
Statement of problem...
plot3d does not display any shading effects.
It is nearly unusable.
this is true. Chosing a shading option produced no real difference
in the appearance of the 3d plot.
Also, using the interactive version, changing the
options buttons under Plot:Plot Types (2d,polymarker,discrete, etc.)
they have no effect on the plot already shown.
I was unable to duplicate this part. Perhaps he is referring to
the non-interactive version where nothing happens if you change the
plot type on the initial dialog from which you ran the plot, but you
can change the plot from the options button on the plot display.
JHS
================================================================
I...-747
Bug Description: Mismatch between displayed object and stored object
Toolbox: ENVISION
Object: extractor
Full bug descripton:
Statement of problem...
When I extract a ROI with "Outside Shape" Extraction Policy
(any ROI Shape), at first the ROI seems to be extracted
correctly, at least it is displayed correctly (ROI Presentation
is "Image"). But the output file contains a different image,
namely one in which the extracted ROI appears in the topleft
corner of the mask segment, thus displaying a different image:
e.g. suppose i want to remove the "a"s from the following image:
+-------+ +-------+ +-------+
|bbbbbbb| |bbbbbbb| | bbbb|
|bbaaabb| What i get displayed is |bb bb| ,but the saved file is | aabb|
|bbaaabb| |bb bb| |bbaaabb|
|bbbbbbb| |bbbbbbb| |bbbbbbb|
+-------+ +-------+ +-------+
Somehow x-offset and y-offset information gets lost in saving the
image to file.
With "Inside Shape" and "Shape Outline" extraction policy this does
not happen. The output image is cropped to the selecion, and what I
get displayed and saved
+---+ +---+
to file is |aaa| and not |bbb| which would be the same bug.
|aaa| |bba|
+---+ +---+
I had a look at the source code of extractor, and as far as I could
understand, storing the extracted object to file is simply a
kpds_copy_object(roi_object, out_object);
command. So the whole thing is quite mysterious to me.
Maybe it's a bug of kpds_copy_object()?
Reproduce Bug By:
% extractor -i image:gull -policy 3 -o test.roi
extract an area (not in topleft corner or you won't notice the bug)
% putimage -i test.roi
see the roi has migrated to topleft corner.
Verified. use extractor to do a region of interest taht the the outside
shape and save to a file. then display that file. when I cut out the
center of an image, and saved the file, when I redisplayed it, the cutout
part was up in the upper left hand corner
================================================================
I...-748
Bug Description: Missing accuary in ifiltdesign when using gauss filter
Toolbox: IMAGE
Object: kimage_proc
Full bug descripton:
Statement of problem...
Missing accuary in ifiltdesign when using gauss filter is
from the use of the constant
#define M_LN2 0.69314718055994530942
with only the first 7 or 8 digits. The literal constant
should be changed to a more exact value as given in the define.
( Line 464 in
khoros2/image/objects/library/kimage_proc/src/lifiltdesign.c )
in the same directory is SQRT2 defined in a similary bad way.
================================================================
I...-749
Bug Description: cannot relocate "External In/Output" glyphs
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
In cantata, create a simple workspace of three glyphs linked together
(eg. a 2D Gaussian image connected to an "expand" glyph connected to a
"display image" glyph).
Now select the center glyph (eg. the "expand" glyph) and encapsulate
it in a proceedure. Open up the proceedure glyph.
At this point you may click and drag to move the "expand" glyph, but
the "External Input (2)" and "External Output (3)" glyphs may not be
moved relative to each other.
You may select all the glyphs, and drag them around together, but the
I/O glyphs are immovable relative to each other.
================================================================
I...-755
Bug Description: stream connection type does not work with needing multiple connections
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
If I have a file-based connection from one output to two inputs,
cantata allows me to change it to a stream-based connection by clicking on
the connection and selecting stream from the popup. This is not a viable
connection, however, since the glyphs accepting input consume the data
when they read it in, so only one glyph can receive a particular piece of
data. If the output of a glyph is already set to stream and I try to make
connections to multiple inputs cantata deletes all but the latest
connection.
================================================================
I...-756
Bug Description: The .f files produced by ghostwriter won't compile
Toolbox: BOOTSTRAP
Object: ghostwriter
Full bug descripton:
Statement of problem...
The .f files produced by ghostwriter won't compile. THere are two
problems. First of all,
----
C
C Khoros: $Id: known_bugs_info.html,v 1.1.2.2 1997/06/28 22:23:25 mirage Exp $
C
character*80 rcsid
parameter ( rcsid = "Khoros: $Id: known_bugs_info.html,v 1.1.2.2 1997/06/28 22:23:25 mirage Exp $" )
----
this is *fortran*..you have to have SEVEN !@$!@$ blanks, ok? ;-)
you can't have code starting in columns 1-6, ever.
the second problem is that this is *fortran* ;-) and you can't
have these two declarations just laying around outside of
code blocks (unlike C). the error msg i get when compiling
is something about a "missing end statement."
================================================================
I...-759
Bug Description: dip-wk:c04s06contrast-log.wk Display Image crashes
Toolbox: DESIGN
Object: cantata
Full bug descripton:
Statement of problem...
The Display Image gets a segmentation violation when trying to run
on the output from Insert segments.