Fri 10 Nov 2006 02:17:37 PM UTC, comment #12:
Thinking more about this, I'm closing it, the summary refers to a segfault so it is closed...
|
Tue 07 Nov 2006 08:09:29 AM UTC, comment #11:
A testcase for this can be simply done using Ming's actionscript compiler, there's already a stub in CVS (testsuite/actionscript.all/MovieClip.as)
If you feel like installing Ming and extending that testcase
that'd be nice, but I woudlnt' consider it high priority.
|
Mon 06 Nov 2006 09:42:50 PM UTC, comment #10:
Do you need me to make a testcase? That would require mc.LoadMovie() to be implemented (I guess it isn't). Otherwise I don't think createEmptyMovieClip() can be of much use ;)
|
Mon 06 Nov 2006 07:24:28 PM UTC, comment #9:
Well with my last commit this doesn't abort anymore.
Doesn't work, but doesn't abort, so goes back to postponed/normal mode.
|
Mon 06 Nov 2006 03:13:08 PM UTC, comment #8:
Committed an hugly patch for the exposed bug, which is now
also testable by running testsuite/actionscript.all/MovieClip.swf
nonetheless I get another other assertions failing while playing that movie, so I'm going after each one , adding testcases for them
and committing fixes.
Just also committed a patch for Function.apply().
I'm afraid too many ActionScript builtin-functions do currently
abort() on invalid calls.. anyway, let's give it a try.
|
Mon 06 Nov 2006 11:09:59 AM UTC, comment #7:
Udo, added you in Cc as this bug is related to bogus createEmptyMovieClip implementation. Since you used createEmptyMovieClip in the last testcase you sent I'm going to try using it to also test this..
|
Sun 05 Nov 2006 10:26:21 PM UTC, comment #6:
It turns out that this bug is new !
It was introduced by Vitaly's patch adding support for createEmptyMovieClip.
With his patch the sprite_definition can now legally exist
with a NULL m_movie_def member, which breaks some implicit
assertions.
It would be cleaner to subclass sprite_definition
with an ActionScript-defined sprite_definition kind
|
Sun 05 Nov 2006 10:12:48 PM UTC, comment #5:
It turns out that this actually triggers a segmentation fault:
gnash \
http://foo.keybit.net/~simm/swfview.swf?timeout=600&map=maps/tww.swf
segfault in 0.7.2 branch
|
Sun 26 Feb 2006 09:54:50 PM UTC, comment #4:
I get a similar error with another MIng generates file.
error message is:
gnash: jpeg.cpp:258: void jpeg::jpeg_error_exit(jpeg_common_struct*): Assertion `0' failed.
and Ming file is a simple slide show with 3 Jpeg images (800x600) with crossfade transition and a mp3 backgound file.
File is about 4.5 Mb attached to this comment
Franco Spinelli
-unavailable-
|
Thu 26 Jan 2006 12:03:51 PM UTC, comment #3:
This is a complex application, with lots of movie loading.
I think Gnash is far from ready to play it.
I set status to 'postponed' and assigned the item to myself,
as I do have access to sources.
|
Sat 07 Jan 2006 03:13:01 AM UTC, comment #2:
Hmm, it works (doesn't crash) on the demos on the Ming site (http://www16.brinkster.com/gazb/ming/index.html).
|
Wed 04 Jan 2006 06:03:11 PM UTC, comment #1:
Can you send me a test case ? Your mapview page states the code isn't released yet, and it'll save me having to learn Ming enough to generate a test case first.
|
Tue 03 Jan 2006 09:58:42 PM UTC, original submission:
http://foo.keybit.net/~strk/mapview/index.html
Those have all been created with Ming (http://ming.sourceforge.net/) and with every one, I get the error:
gnash: action.cpp:1237: void gnash::as_global_assetpropflags(const gnash::fn_call&): Assertion `obj != __null' failed.
Aborted
A black window opens up and nearly instantly it closes with that error message in the shell.
This is from the latest CVS.
|