Posted: Thu Dec 16, 2004 10:17 am Post subject: New Neo/J 1.1 Alpha 2 patch available
Ed and I have fixed many of the bugs that were reported against NeoOffice/J 1.1 Alpha 2 "Patch-8" and so I have posted "Patch-9".
This patch includes the following bug fixes:
1. Random crashing - I fixed this bug by removing the Java "-Xrs" startup option that was added in an earlier patch.
2. Better memory usage - I fixed a massive Java 1.3.1 memory leak. Java 1.3.1 was leaking a massive amount of memory every time a new window was created which would eventually cause Neo/J to eventually consume all available memory.
3. Smaller PDF files - I fixed a couple of bugs in my "save as PDF" code.
4. Safer drag-and-drop behavior - I fixed the bug that was causing Neo/J to delete data when dragging the mouse over another application.
In addition, Fridrich Strba has enhanced Neo/J's WordPerfect support by adding experimental support for WordPerfect for Mac 3.5e files.
You can download "Patch-9" from the following URL:
Joined: Jun 20, 2003 Posts: 104 Location: Birmingham, Alabama; and Amelia (Terni), Italy
Posted: Thu Dec 16, 2004 11:20 am Post subject: Patch 9 more stable
I can say that after minimal testing, patch 9 increases NOJ's stability by (so far) eliminating the crash when hiding NOJ, leaving it for a while, then reactivating it from the dock. That one improvement eliminates at least 50% of the crashes I experienced, perhaps 75%. Good work! Cheers, _________________ Gib Henry
My PDF issues are completely gone! I am very happy now. But I got overconfident and tried messing some more. I don't know what exactly triggered it, but the app crashed.
Roark else beat me to reporting it, though. Details are in bug 263. I attached a crashlog which seems to be pretty much identical to the one that was pasted.
Roark else beat me to reporting it, though. Details are in bug 263. I attached a crashlog which seems to be pretty much identical to the one that was pasted.
Can you open Console.app and see if there are any lines that say "pure virtual method called"?
Could you try out the test patch that I justed added to the comments in bug 209?
Downloaded it and starting up now. No guarantees that I can give it a serious workout right now, because it is getting pretty late here.
EDIT: It is now an hour later and I am giving up on getting ExtendedPDF to work (see other thread).
But on the bright side, switching between various apps and NeoOffice/J by means of the dock icon has been perfectly wellbehaved (though there was a half-second hitch in one case - like it needed to swallow and only then could start talking). So Kudo's again to you, Patrick.
I experienced the same crash again (as Bug 263, which has reverted to Bug 209). It's still hanging in the background, with the spinning beach ball. This is with patch 9 and the "testpatch1" from Bug report 263 & 209.
I had it open and in the background for an hour or so, doing other things in the foreground. Clicked on it and used the scroll wheel, and it just hung, beach ball and all. I am using a different computer now (iBook G4 at work), and both this and the frankenmac at home have USB Overdrive installed. I have USBOD set to page up/down with a scroll up/down (less jumpy than the default). I wonder if the bug is related to the scrolling ability that you rigged up?
Otherwise, I feel NeoJ is getting really really close, and I'm really looking forward to a stable version. I'm getting by fine with Mellel and OOo, but NeoJ is the dream of the near future. You've done amazing work, Guys!
I experienced the same crash again (as Bug 263, which has reverted to Bug 209). It's still hanging in the background, with the spinning beach ball. This is with patch 9 and the "testpatch1" from Bug report 263 & 209.
Without "testpatch1", didn't it crash? If so, let me know as that would indicate that a deadlock is now occurring which, although not good, is at least a change.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Thu Dec 16, 2004 8:06 pm Post subject:
Note : For the folks still reporting the bugs during activate (209), I built a new debug library that will spit out both the appleevent as well as any symbols that are being bound lazily...I think it's one of those bindings that's causing the crash but in my searches I can't track down what the symbol would be. Hopefully traces from this might help track things down. Thus far the only two I see are GetEventClass and GetEventKind, but those are system library functions which should always exist.
Library is available for download via bittorrent :
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Thu Dec 16, 2004 8:09 pm Post subject:
Of note on the above torrent:
- includes patch 9 fixes
- also includes Patrick's latest code for ignoring unhandled appleevents
- install only after patch 9 is installed (doesn't affect vcl.jar)
I have not further experienced the quit or rapp crashes with the removal of -Xrs, but for folks that are still having them hopefully this library can at least help us track down what symbol is causing the failure in dyld.
I experienced the same crash again (as Bug 263, which has reverted to Bug 209). It's still hanging in the background, with the spinning beach ball. This is with patch 9 and the "testpatch1" from Bug report 263 & 209.
Without "testpatch1", didn't it crash? If so, let me know as that would indicate that a deadlock is now occurring which, although not good, is at least a change.
Patrick
Patrick,
I may have lost track, but I think the crashes happened with and w/out "testpatch1." I applied "testpatch1" after posting Bug 263. On the other hand, I had a patch8 version running on a third computer (G4 450Mhz Powermac at work) for a day or so (running in the background with a big text file opened) and it was fine, except for a screen redraw issue that I've seen before. I could bring it to the front and scroll (also with USB Overdrive) normally. Scrolling got rid of the blank screen. I just quit it normally. Maybe something introduced between patch8 and 9?
Note to Ed: I had no luck getting the torrent to work. I'm a newbie, though I've downloaded YDL with bit torrent before. Perhaps there are no "seeders" awake?
This crashing bug has been around since Ed first reported it in early November. My guess is that a Panther update has caused a bug in Java 1.3.1. Why do I suspect this? Because I have a very long list of Java 1.3.1 crashing bugs that appeared in Panther that did not exist in Jaguar.
Not much we can do about the fact that there is a JVM crashing bug other than to keep trying various hacks. It's tedious, but it has worked so far.
Anyway, getting back on topic, I think "testpatch1" has no effect good or bad. So, I have uploaded http://www.planamesa.com/test/testpatch2.tar.gz. This patch isn't ideal, but it prevents all Apple events from ever reaching the JVM. This has the side effect of disable opening documents from the Finder, but it Neo/J does not crash with this patch, then I can add that feature back.
All times are GMT - 7 Hours Goto page 1, 2, 3, 4, 5, 6, 7Next
Page 1 of 7
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum