|Anonymous | Login | Signup for a new account||2018-12-10 15:28 CET|
|Main | My View | View Issues | Change Log | Roadmap | Docs|
|Viewing Issue Simple Details|
|ID||Category||Severity||Reproducibility||Date Submitted||Last Update|
|0005207||[JEDI VCL] 00 JVCL Components||major||always||2010-03-23 16:37||2011-06-10 16:09|
|Status||resolved||Product Version||Daily / GIT|
|Summary||0005207: stack overflow testing sample of issue 0005190|
i've just done an update with the current svn and rebuild the jvl and jvcl
for testing the resolution of issue 0005190 but ... when i open the project
that i've sended some days ago attacched on issue 0005190 and i try to run it, i get a "Stack Overflow" when the program do "CDS1.open" ... if i try at designtime to active the table then the grid populate without error ... doh!
What's happened ?
|Tags||No tags attached.|
other test that i've done :
- the stack overflow appear also in other program that worked previous of the last update
- deleted all the jvcl and jcl directory and re-downloaded all. Same problem
- done 2 revision downgrade. Same problem
- deleted all. Copied an old version. All worked well as before !
- upgraded only JvDBLookup.pas from the current SVN. All worked well ... but the component don't do automatic refresh as stated in the resolve of 5190. I need to call the ForceDisplay procedure that i've seen it's included in the current branch.
my conclusion: the problem is elsewhere in the upgrade of the jedi component and not in the new JvDBLookup.pas but it's a severe problem because i can't upgrade all the svn now ...
i've had the time to do some other test :
- Deleted all an re-downloaded to the current svn
- installed jcl and jvcl
- tested the program attached in 5190 (after commented out the ForceDisplay), result : stack overflow !
- downgraded only the JvDBLookup.pas to the revision 12723
- tested the program attached in 5190 (after commented out the ForceDisplay) result : all goes well, doesn't crash (but the bug of 5190 is not fixed in this version)
so my previous conclusion was not correct :-)
the is really in the JvDBLookup.pas
I do the same, no more stack overflow.
Thanks for your time
Fixed in SVN.
The "Locate" that is executed by the grid started a endless recursion of DataLink events.
|2010-03-23 16:37||Thor||New Issue|
|2010-03-24 09:09||Thor||Issue Monitored: Thor|
|2010-03-24 18:59||outchy||Relationship added||child of 0005190|
|2010-03-25 09:01||Thor||Note Added: 0017323|
|2010-03-28 16:53||nestor||Issue Monitored: nestor|
|2010-03-29 11:59||Thor||Note Added: 0017326|
|2010-03-29 15:41||nestor||Note Added: 0017329|
|2010-03-29 18:04||AHUser||Note Added: 0017330|
|2010-03-29 18:04||AHUser||Status||new => resolved|
|2010-03-29 18:04||AHUser||Fixed in Version||=> Daily / SVN|
|2010-03-29 18:04||AHUser||Resolution||open => fixed|
|2010-03-29 18:04||AHUser||Assigned To||=> AHUser|
|2010-03-30 09:21||Thor||Issue End Monitor: Thor|
|2011-06-10 16:09||obones||Fixed in Version||Daily / SVN => 3.40|
|Mantis 1.1.6[^] Copyright © 2000 - 2008 Mantis Group|