Can you try rebooting and let me know if this help?
Patrick[/quote]
Hi Patrick,
Thanks for patch 12. I will install it now, but I wanted to give an update on patch 11. The freeze I experienced (and repeatably with the frankenmac) was with patch 11. I eventually ran Yasu (a nifty utility program) to delete system cache in addition to all its other functions (optimize, repair permissions, etc.). Threw out the NeoJ folder in my Library, let NeoJ build a new one. It's been running fine on the iBook G4 and the frankenmac at home. I can't get the latter to crash with the old chore (opening the 5.7 MB file of Shakespeare and letting it spellcheck the entire thing!).
The whole restart issue changes everything--makes me think we may have been chasing our tail in circles on this one. =) Also, even though I know well enough to quit the program before I run the patches, some sort of warning would be good for the novices out there, if it can be arranged easily.
I'll let you know how patch 12 is going. Thanks again!
Patch12 seems to be working like a champ. And as I pull myself back for a bit, perspective-wise, I really have to say you guys have made serious progress here. It launches fast, relaunches REALLY fast, loads ridiculously large docs in reasonable time and with decent stability. The memory footprint is MUCH smaller and the Wordperfect filter, even though I haven't used it much, seems to work great. And I haven't seen any of those whited-out dialogue boxes with the ASCII files.
As always (or at least for the last few days), I hope the patch sticks, but even if another glitch emerges (I hope not---fingers and toes crossed!), this is a super program that has become much better in the past few days.
This is just bug 266 (http://bugzilla.neooffice.org/bug.php?op=show&bugid=266&pos=9). My patch checking logic is pretty moronic in that it only checks if there is a file on the download server that is newer than the last time you checked. It doesn't actually look whether you installed the patch or not. It just notifies you.
Maybe one of these days I'll enhance it to actually check if you have the patch installed.
I think Patrick's reasoning in that bug is pretty solid, although I'll admit the behavior was a bit strange when I noticed it
pluby, bug 266 wrote:
The patch update checking script only tests if there is a patch that is newer than the last time that you ran Neo/J.
I keep meaning to run Neo/J while not connected, after a new patch has been released, and then run it again later to make sure it will still notify me of a new patch--I haven't been clear for a while on whether the patch logic (as publicly explained) will handle that case properly. This could be an issue with people on dial-up (like I am after I leave campus for the day )....
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Mon Dec 20, 2004 11:08 pm Post subject:
While it may be moronic, I suspect it will just lead users to checking if they have "Block Pop-up Windows" enabled. Seriously, I think it's much better to have some form of update checking in there then the alternative of having no update checking at all. I think it's great to be able to get potential users of patches informed of new ones, particularly those who aren't the kind of folks that drop by the testing forums on a frequent basis.
The only simple thing I can think of would be to see if there's a way to automatically have the patch installers go through and "touch" all of the ".lastpatchcheck" files on the local system as the last step of the patch install, touching them to the date of the dimage if possible. That should prevent the subsequent informative patch page from opening post-patch-install if I'm not a moron.
Of course, doing that opens up bad scenarios. Suppose patch "14" is available (or even a beta!). A user gets their hands on an oudated patch 12, for instance, installs it, and those files get touched. The update page then fails to display even though a newer patch is available. I think it's probably better to err on the side of mis-notifications of potential newer versions myself.
Let me know if you'd like me to investigate that further. I personally had no problem with what was there and thought it was the better solution myself to avoid "false negatives".
In any case, thanks for doing the automatic upgrade notification for us Patrick. It's even helped me out by trying to keep track of which of my 5 test partitions I've accidentally forgotten to install patches onto
Sorry, I guess I'm confused. Why not just check to see if there is a patch # > the latest patch installed? So if you've only installed patch 5, but patch 7 is available, well.. 5 < 7 so it'll load the patch page, regardless of dates?
While it may be moronic, I suspect it will just lead users to checking if they have "Block Pop-up Windows" enabled. Seriously, I think it's much better to have some form of update checking in there then the alternative of having no update checking at all. I think it's great to be able to get potential users of patches informed of new ones, particularly those who aren't the kind of folks that drop by the testing forums on a frequent basis.
The only simple thing I can think of would be to see if there's a way to automatically have the patch installers go through and "touch" all of the ".lastpatchcheck" files on the local system as the last step of the patch install, touching them to the date of the dimage if possible. That should prevent the subsequent informative patch page from opening post-patch-install if I'm not a moron.
Of course, doing that opens up bad scenarios. Suppose patch "14" is available (or even a beta!). A user gets their hands on an oudated patch 12, for instance, installs it, and those files get touched. The update page then fails to display even though a newer patch is available. I think it's probably better to err on the side of mis-notifications of potential newer versions myself.
Let me know if you'd like me to investigate that further. I personally had no problem with what was there and thought it was the better solution myself to avoid "false negatives".
In any case, thanks for doing the automatic upgrade notification for us Patrick. It's even helped me out by trying to keep track of which of my 5 test partitions I've accidentally forgotten to install patches onto
Joined: Sep 18, 2003 Posts: 434 Location: London, UK
Posted: Tue Dec 21, 2004 5:17 am Post subject:
While we're on the subject of patches, can I make a suggestion about the current naming scheme for them. With the number of users/testers increasing, and of the number of patches themselves increasing in frequency, it is getting to a stage where the use of a simple number as the nomenclature for the patch is making it difficult to keep track of which patch is associated with which version of NO/J and, especially so, of which testpatch is associated with which patch. This is most apparent when reading through the bugzilla reports when you haven't managed to do so for a few days or longer. Can I suggest that the patches are re-titled to include the date of release as well as or in preference to a simple number title so e.g. patch 12 could become "2004_12_20 patch 12" or simply "2004_12_20 patch". If you have two or more patches/testpatches released on the same date, you can simply append a B to the second, a C to the third, etc. (Fwiw, the Year-Month-Day date format is preferable as it means the files are guaranteed to appear in the correct order in the Finder or whatever dialogues they occur in, whereas all other date formats aren't).
No doubt that would bugger up your patch checking algorithm quite a bit right now!
Cheers,
Jonathan _________________ PBG4, 1.5GHz, SuperDrive, 1GB RAM, 128MB VRAM, 5400rpm 80GB HD, MacOS X 10.4.5
How does Apple number their beta/alpha releases & updates? The #.#.# seems pretty clear, but I've never seen how they show test versions.
If Neo/J followed that, patches would just increase the least significant (ie, the third) # in the sequence...
Finally, don't these test patches include all the previous ones? That is, I don't think patch12 requires patch11 or patch10 to be installed first... if you just downloaded Alpha 2, you should be able to install the most recent patch (12 at the moment) and be instantly up-to-date.
W
JKT wrote:
While we're on the subject of patches, can I make a suggestion about the current naming scheme for them. With the number of users/testers increasing, and of the number of patches themselves increasing in frequency, it is getting to a stage where the use of a simple number as the nomenclature for the patch is making it difficult to keep track of which patch is associated with which version of NO/J and, especially so, of which testpatch is associated with which patch. This is most apparent when reading through the bugzilla reports when you haven't managed to do so for a few days or longer. Can I suggest that the patches are re-titled to include the date of release as well as or in preference to a simple number title so e.g. patch 12 could become "2004_12_20 patch 12" or simply "2004_12_20 patch". If you have two or more patches/testpatches released on the same date, you can simply append a B to the second, a C to the third, etc. (Fwiw, the Year-Month-Day date format is preferable as it means the files are guaranteed to appear in the correct order in the Finder or whatever dialogues they occur in, whereas all other date formats aren't).
No doubt that would bugger up your patch checking algorithm quite a bit right now!
Yet another crash with patch12. Can't seem to reproduce it, and I'm afraid I can't help decipher the crash log. I wasn't paying close attention, but I think it triggered when I quit from the keyboard. I posted it as "patch12crash.txt" to Bugzilla.
Yet another crash with patch12. Can't seem to reproduce it, and I'm afraid I can't help decipher the crash log. I wasn't paying close attention, but I think it triggered when I quit from the keyboard. I posted it as "patch12crash.txt" to Bugzilla.
Waldo, part of the issue JKT was talking about is that a number of the "released" Apple Installer patches (Patch 9, etc.) had semi-private .tgz testpatches (testpatch5.tgz, etc.) for reporters to try and see if Patrick and Ed had fixed things before spinning a "released" patch. While trying to fix bug 209 and a couple of related ones, there were a number of these testpatches (some mentioned only in Bugzilla, some mentioned in the threads as well) for several different released patches. The number and publicity of these recently is a unique situation in the 8 months I've been around here...
I lost track of the specifics, too, but I wasn't experiencing any of the crashes so I basically ignored the testpatches.
What about NeoJ-ver-Patch-num-testpatch-num.ext when/if there are testpatches again? E.g, NeoOfficeJ-1.1_Beta-Patch-1-testpatch-2.tgz
The existing method for numbering released patches (NeoOfficeJ-1.1_Alpha_2-Patch-1.dmg) seems pretty clear; it's just the recent flurry of testpatches that have messed things up
Oh, no worries. It's just a relief to find that it didn't leave a log. For about half an hour there I was preparing to commit ritual suicide upon seeing bug 209 potentially open again or at least burn a figure of that silly java mascot in effigy
The spinning beachball is quite frustrating though. I did notice a deadlock with regards to the display change notifications at one point during my rapp debugging and attached to get traces but hadn't done anything about those yet. I wouldn't think that expose or the screen savers should be sending display manager notifications, though, so it may be unrelated.
ed
Patrick and Ed,
First, congrats on the Beta, it looks great. However, I have a bit of bad news. I'm afraid that it is crashing on the iBook G4, consistently now, and I am beginning to wonder if it is what Ed remarked on in the quoted text. The freeze (with spinning beach ball) seems to happen consistently after the screen saver (a custom one built using the "Build Slide Show" app in the System folder) is invoked. Since it is a freeze and not a crash, maybe knowing this can help us pin it down. I would be happy to test patches on this iBook to see if we can make it go away (since Console won't catch it, right?). NeoJ is behaving well on my frankenmac, which has a different screen saver. No crashes there so far.
Does this make sense? Is it possible that when the screen saver is invoked it is crashing NeoJ due to "display manager notifications"? I'll try testing it now with no screen saver enabled.
First, congrats on the Beta, it looks great. However, I have a bit of bad news. I'm afraid that it is crashing on the iBook G4, consistently now, and I am beginning to wonder if it is what Ed remarked on in the quoted text. The freeze (with spinning beach ball) seems to happen consistently after the screen saver (a custom one built using the "Build Slide Show" app in the System folder) is invoked. Since it is a freeze and not a crash, maybe knowing this can help us pin it down. I would be happy to test patches on this iBook to see if we can make it go away (since Console won't catch it, right?). NeoJ is behaving well on my frankenmac, which has a different screen saver. No crashes there so far.
Does this make sense? Is it possible that when the screen saver is invoked it is crashing NeoJ due to "display manager notifications"? I'll try testing it now with no screen saver enabled.
I don't think this is a crash. Instead, I think it is a deadlock. Most likely, Neo/J is blocked waiting for a lock to be unset so that it can continue.
The simplest way to debug this would be to attach to the frozen Neo/J process with the gdb debugger and then print a stack trace using the "bt" command with gdb.
Do you happen to have gdb or the Apple developer tools installed on your machine?
Does this make sense? Is it possible that when the screen saver is invoked it is crashing NeoJ due to "display manager notifications"? I'll try testing it now with no screen saver enabled.
I don't know if it will help, but I found another possible deadlock condition in the code. I have no idea whether this works or not. But if you would like to see if my code works, I have put a test patch here:
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
