This sounds like bug 1643. Can you please add yourself to the CC: list in that bug, run the test patch that is in that bug, and add your output of the test patch to that bug?
bluespot wrote:
- when I open a OOo Base file with a MySQL connection, the database connection parameters are empty.
This is an OpenOffice.or 2.0.3 bug. OpenOffice.org 2.0.3, IMHO, has many regressions and new bugs.
<rant>Supposedly, Sun's engineers are only fixing bugs in these 2.0.x releases but my brief look at that code when I upgraded from OOo 2.0.2 to OOo 2.0.3 is that they are making sometimes huge and unnecessary changes. For example, in OOo 2.0.3, they refactored their custom memory manager and made it very difficult to turn it off (I use Mac OS X's malloc() and free() functions as I usually put my trust in the OS over specific applications custom code). Also, my impression is that Base is still not that stable is really in the development stage so use it with that in mind.</rant>
Patrick
Last edited by pluby on Thu Aug 17, 2006 9:42 am; edited 1 time in total
This sounds like bug 1643. Can you please add yourself to the CC: list in that bug, run the test patch that is in that bug, and add your output of the test patch to that bug?
Actually, this is an OpenOffice.org bug. As described in this bug, this same errant dialog occurs on Linux.
Posted: Thu Aug 17, 2006 11:34 am Post subject: Re: NeoOffice Aqua Beta 2 (Intel) regressions
Thank you very much for your prompt answers, I will have a look next Monday (when I'll be back at work...). I'm rather shocked to learn that there are so many regressions from 2.0.2 to 2.0.3 I do think that I will have to come back to Alpha if these bugs can't be solved (I saw that the concerned OOo bug is dated from May 2006 )
doctype wrote:
bluespot wrote:
- when I open a OOo Base file with a MySQL connection, the database connection parameters are empty.
No such problem here, MySQL-based Database Files "migrated" perfectly from Alpha 4 to Aqua Beta.
Pretty strange. I even attempted to update the database parameters firstly via MySQL Java driver but there is no field to put the server name (only the database name !!!). Then I could update via the JDBC classname, but the changes aren't taken in account in a subsequent re-opening of the Base file......
(rather pretty bothering as I manage statistics every day and I'm actually stuck because of this bug) *BUT* i want to applause the NeoOffice team for the great Aquafication work !
I've downloaded the latest patch as indicated by Patrick, but unfortunately it didn't solve the network bug
Therefore, it does confirm the infamous OOo bug.
Any hope to get it fixed ? As said before, it's a severe show-stopper for me as I'm regularly working with files on a AFP server : I'll have to reinstall NeoOffice Alpha.
I've downloaded the latest patch as indicated by Patrick, but unfortunately it didn't solve the network bug
Therefore, it does confirm the infamous OOo bug.
Any hope to get it fixed ? As said before, it's a severe show-stopper for me as I'm regularly working with files on a AFP server : I'll have to reinstall NeoOffice Alpha.
I finally traced down the source of this bug and, what a surprise, it was due code added in OOo 2.0.3 by the OOo X11 team. Their code attempts to fix the network file locking bugs that I fixed nearly 2 years ago. What amazes me is not that their code is buggy, but that I have donated well-tested code to them in nearly 18 months ago and they chose to ignore my code and reinvent the wheel.
Anyway, please try out the test patch listed at the end of bug 1663.
I finally traced down the source of this bug and, what a surprise, it was due code added in OOo 2.0.3 by the OOo X11 team. Their code attempts to fix the network file locking bugs that I fixed nearly 2 years ago. What amazes me is not that their code is buggy, but that I have donated well-tested code to them in nearly 18 months ago and they chose to ignore my code and reinvent the wheel.
They said it was too old since it was for OOo 1.1.4 and they were working on OOo 2.0.
I finally traced down the source of this bug and, what a surprise, it was due code added in OOo 2.0.3 by the OOo X11 team. Their code attempts to fix the network file locking bugs that I fixed nearly 2 years ago. What amazes me is not that their code is buggy, but that I have donated well-tested code to them in nearly 18 months ago and they chose to ignore my code and reinvent the wheel.
They said it was too old since it was for OOo 1.1.4 and they were working on OOo 2.0.
Yup. I remember. Of course, it took me less than a day to merge my OOo 1.1.x changes into the OOo 2.0.x code. Why? Because the OOo 2.0.x file handling code did not change much between OOo 1.1.x and OOo 2.0.x.
In any case, their fix in OOo 2.0.3 doesn't work and is probably worse than no fix in that their fix creates a bunch of empty files before it throws up the cryptic "Can't create backup copy" error.
I remember seeing something in IssueZilla "recently" in a bug about network volumes where one engineer said something like "oh, maybe that's what Patrick's code did"
It would be pure comedy if it didn't adversely affect users so much.
Smokey _________________ "[...] whether the duck drinks hot chocolate or coffee is irrelevant." -- ovvldc and sardisson in the NeoWiki
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Mon Aug 21, 2006 10:01 pm Post subject:
This is yet another reason why as of late I'm coming around to the opinion that we may need to split off from following OOo proper and performing our own code review of new code we incorporate into our source base.
For non Sun-supported platforms, the level of Quality Assurance being performed by OOo as an organization does not meet my personal standards. If freezing out OOo code is the way we prevent regressions due to subpar engineering from creeping into our product, so be it.
Joined: Nov 21, 2005 Posts: 1285 Location: Witless Protection Program
Posted: Mon Aug 21, 2006 10:56 pm Post subject:
OPENSTEP wrote:
This is yet another reason why as of late I'm coming around to the opinion that we may need to split off from following OOo proper and performing our own code review of new code we incorporate into our source base.
For non Sun-supported platforms, the level of Quality Assurance being performed by OOo as an organization does not meet my personal standards. If freezing out OOo code is the way we prevent regressions due to subpar engineering from creeping into our product, so be it.
ed
I have noticed more code problems, and checking problems, with the Mac code. To few people trying to do to many huge jobs??
<rant - but true?>To many cooks, not enough quality checking of the code? Plus, It does not seem clear if work is focused on Aqua, X11, both or some mixed combination. Lack of ... focus??<rant>
If it was my code, I would be very concerned. I don't have the skill to check the code, but I would be very concerned about the quality of recent Mac code updates.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Mon Aug 21, 2006 11:15 pm Post subject:
As Patrick ranted above, it is not a problem isolated to the Mac team; some of the new instability is even coming from within Hamburg. The switch to a forced six month release schedule was not accompanied by a similar increase in the quality controls. Any time you try to reduce your standard delivery schedule by 75% something's gotta give.
To be fair, I suspect the schedule was forced upon them by marketing, but still...it's yet another thing that both users and forks that demand stability need to be aware of. "Point" updates coming out of Hamburg are no longer just bugfixes but include non-trival amounts of new, untested code.
Anyway, please try out the test patch listed at the end of bug 1663.
The test patch is working pretty fine, thank you very much ! And by the way, it does correct the erractic behaviour I got with the Base files : the database connection parameters are now correctly retrieved.
Besides, it's a real shame this story about network bug ... I nearly felt from my chair !!!
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
I do think that I will have to come back to Alpha if these bugs can't be solved (I saw that the concerned OOo bug is dated from May 2006 
)
I don't have the skill to check the code, but I would be very concerned about the quality of recent Mac code updates.
