CRASH: 5.0aB2 Regular Crashing when Mouseovering NA27 w. GRAMCORD with McReynolds Int.

Continuation of thread http://community.logos.com/forums/t/60098.aspx (whose title I can't update).

So there is something about the contents of one layout that seems to trigger the crashes, but another, simpler one does not. The crash-prone layout has a fair number of tabs open (see Screenshot 1) and the second is much simpler.

See the other thread for crash logs.

Thanks
Richard

Screenshot 1 (crashing)

Screenshot 2 (no crashes so far)

Find more posts tagged with

Comments

Jon

Crosspost from your other thread:

Well, it looks like all the crashes are the same, and maybe related to a mouse-over popup of some sort. Maybe try reloading some of the things you were looking at and put your mouse on different links etc, see if you can trigger it deliberately...

Richard Lyall

https://community.logos.com/discussion/comment/427743#Comment_427743

OK thanks will do.

UPDATE: Just triggered it twice there when mouseovering NA27 - log attached. Yep this is looking quite reproducible now ... it's NA27. I've highlighted the rough area that I was mouseovering when crash occurred.

5417.Logos Log Files.zip

Please note that the "conjunction, coordinating, copulative" popup is "always on top" and is superimposed on all other open Windows (outside L5) - I haven't dismissed the crashed L5 yet.

Richard Lyall

https://community.logos.com/discussion/comment/427747#Comment_427747

Yep this is 100% reproducible now. If I scrub the mouse over NA27, within a few seconds it will crash.

Will try again with Inline off.

UPDATE: Yes, still happens with Inline off.

Jon

https://community.logos.com/discussion/comment/427754#Comment_427754

I guess if it's just limited to that resource you could try hiding and unhiding the resource (and restart Logos after each step) to redownload it, or use a different GNT in the meantime and wait for a Logos person to weigh in.

Richard Lyall

https://community.logos.com/discussion/comment/427779#Comment_427779

Yes have switched in the meantime ... how can you tell it's a mouseover by the way?

Richard Lyall

https://community.logos.com/discussion/comment/427754#Comment_427754

UPDATE: have tried a number of other Greek NTs but have not yet triggered the crash. The one that I'm viewing when crashing occurs is:

LLS:1.0.317
2009-10-26T18:11:47Z
NA27INT.logos4

I can't reproduce the crash with any other GNT, whether Inline is off or on, and whether Visual Filters are active or not.

Jon

https://community.logos.com/discussion/comment/427786#Comment_427786

how can you tell it's a mouseover by the way?

Looking at the stack trace in the error logs and having a guess [:)] it goes from the function that actually caused the crash down through the stack (each one having been called by the one on the next line). Even without knowing how the code works the function names give you the gist of what's going on...

If you have a look about a third of the way down the stack trace in LogosError.log and work up you can see this:

OnMouseOverSomething —> RaiseActionEventAt —> RaiseDisplayAction .... ShowInformation —> ShowInformationStatusBar

I think the actual crash happens when it tries to load the info to show in the popup (I presume the parsing info popup) and the object its trying to extract the info from isn't there for some reason...

Brisa Davis

https://community.logos.com/discussion/comment/427798#Comment_427798

I am able to reproduce this crash, and will report it to Development.

Brisa Davis

https://community.logos.com/discussion/comment/427840#Comment_427840

This issue appears to be resolved in 5.0a Release Candidate 1. If you continue to see this issue after updating, please let us know.

(Note: At the time that I am posting this, we have not yet shipped Release Candidate 1. No need to panic, you haven't missed a release [:P])

Richard Lyall

https://community.logos.com/discussion/comment/428061#Comment_428061

Great thanks!

Richard Lyall

https://community.logos.com/discussion/comment/428068#Comment_428068

Seems good so far in 5.0a RC1 - thanks! Unable to reproduce original crash now.