T.R | Title | User | Personal Name | Date | Lines |
---|
2505.1 | | QUARK::LIONEL | Free advice is worth every cent | Thu Mar 22 1990 20:56 | 4 |
| Well, seeing as I was the last to "touch" DECburger, I'll look at
it when I have a chance. Is this the C, FORTRAN or Ada version?
Steve
|
2505.2 | C burger | SKYLRK::JOLLEY | This space left intentionally blank | Fri Mar 23 1990 14:16 | 8 |
|
Hi Steve,
We're working with just C for now.
Thanks,
Jolley
|
2505.3 | | QUARK::LIONEL | Free advice is worth every cent | Fri Mar 23 1990 14:34 | 3 |
| Ok, I can reproduce it on T5.4-4ET. I'll look into it.
Steve
|
2505.4 | Related to 2511 ?! | GYPSC::BADE | | Mon Mar 26 1990 09:06 | 5 |
| Please have a look to the XtUnmanage/XtManage -problem (note 2511).
Although this becomes nicely visible with the DECburger, we have run
into the problem with our (much larger) application.
Dirk Bade
|
2505.5 | | QUARK::LIONEL | Free advice is worth every cent | Mon Mar 26 1990 12:50 | 5 |
| That's out of my control - it's a problem in the toolkit. I can only try
to see if DECBURGER is doing something wrong, or determine that the toolkit
is doing something wrong.
Steve
|
2505.6 | | QUARK::LIONEL | Free advice is worth every cent | Tue May 01 1990 17:14 | 25 |
| I believe I have found the cause for this error, though I don't yet know
whether it is a programming error on my part or a bug in the toolkit.
To "fix" the problem, remove the following lines in routine ACTIVATE_PROC:
... in the loop that collects the burger toggles:
CALL XT$FREE(TEXT)
... in the case for one of the fries buttons:
CALL XT$FREE(CURRENT_FRIES)
(This is in FORTRAN, the Ada and C code have similar sequences.)
What seems to have changed is that the call to GET_VALUES to get the widget's
label is now returning a pointer to the actual widget label compound string,
rather than a pointer to a copy of the label string. The toolkit is incredibly
inconsistent about such things, and the documentation in this regard is
nil.
I have inquired separately about the correct action. Once I find the answer,
I will ask to have the examples updated for a future version of VMS.
Steve
|
2505.7 | | QUARK::LIONEL | Free advice is worth every cent | Wed May 02 1990 16:18 | 9 |
| I have made fixes in the Ada, C and FORTRAN versions of DECBURGER to
correct this error. It's a bit more complicated than I indicated in .6,
though removing the lines I noted there will allow the program to run -
it will just leak a bit of memory.
If you want the new version in any of the above languages, send me mail.
They'll likely show up in some future VMS release (after V5.4).
Steve
|
2505.8 | Do fries go with that shake? | QUARK::LIONEL | Free advice is worth every cent | Wed Jun 06 1990 16:39 | 7 |
| I am told that the final VMS V5.4 kit will have the latest versions of
the DECBURGER (and HELLOWORLD) examples with all known bugs fixed. I also
added some comments and a change history, so look at it to see what changed.
Note that these are not in T5.4-4GE.
Steve
|
2505.9 | | QUARK::LIONEL | Free advice is worth every cent | Tue Jul 17 1990 17:01 | 6 |
| I can confirm that the fix for this problem is in T5.4-4HW, and thus very
likely to be in V5.4. See the new comments and change history in both
the HELLOWORLD and DECBURGER example programs in Ada, C and FORTRAN for
additional changes.
Steve
|