float point error

We are still seeing this error:
http://support.microsoft.com/default.aspx?scid=kb;en-us;818899
on the latest version of sqlserver2k :
select @.@.version
--
Microsoft SQL Server 2000 - 8.00.760 (Intel X86) Dec 17 2002 14:22:05
Copyright (c) 1988-2003 Microsoft Corporation Enterprise Edition on Windows
NT 5.0 (Build 2195: Service Pack 4)
Does anyone know when MS will release an official patch. Suposedly, the
have a dll then can give you but nothing official. This is kindof a pretty
bad bug to leave open.> We are still seeing this error:
> http://support.microsoft.com/default.aspx?scid=kb;en-us;818899
Well, did you follow the instructions in the article, by contacting
Microsoft product support and obtaining the .807 hotfix (which is later than
the "latest version" of .760)? There is no fee for the call or the hotfix
itself if you demonstrate to them that you are affected by the issue the
hotfix fixes.
Otherwise, you can try to find later patches (e.g. see
http://www.microsoft.com/technet/security/bulletin/ms03-031.mspx which
updates you to .818).
Barring those two actions, you will have to wait for SP4.
> This is kindof a pretty bad bug to leave open.
Well, that really depends on how many users it has affected, doesn't it?
--
http://www.aspfaq.com/
(Reverse address to reply.)|||As you know, SQL Server QFEs are cumulative. The fix described in this
article is Build 2000.00.0807. The latest publicly available hot fix is
Build .0878, which would include a fix for this error.
http://support.microsoft.com/?kbid=838166
Also know that SP4 is currently in Beta testing and will include builds up
to .0972.
Sincerely,
Anthony Thomas
"Aaron [SQL Server MVP]" <ten.xoc@.dnartreb.noraa> wrote in message
news:OrJl5DOAFHA.824@.TK2MSFTNGP11.phx.gbl...
> We are still seeing this error:
> http://support.microsoft.com/default.aspx?scid=kb;en-us;818899
Well, did you follow the instructions in the article, by contacting
Microsoft product support and obtaining the .807 hotfix (which is later than
the "latest version" of .760)? There is no fee for the call or the hotfix
itself if you demonstrate to them that you are affected by the issue the
hotfix fixes.
Otherwise, you can try to find later patches (e.g. see
http://www.microsoft.com/technet/security/bulletin/ms03-031.mspx which
updates you to .818).
Barring those two actions, you will have to wait for SP4.
> This is kindof a pretty bad bug to leave open.
Well, that really depends on how many users it has affected, doesn't it?
--
http://www.aspfaq.com/
(Reverse address to reply.)|||> As you know, SQL Server QFEs are cumulative.
That is true until .977. For example, to install .993 on .760, you first
need to get to the .977 hotfix installer -- .993 won't install on < .977 (it
will complain about missing prerequisites).
> The fix described in this
> article is Build 2000.00.0807. The latest publicly available hot fix is
> Build .0878, which would include a fix for this error.
However, it is not always the best to just install the latest. He may
prefer to go through PSS, demonstrate that .807 is his issue, and fix that.
.878 might bring about other problems that he doesn't have time to fully
test. Sure, it might be more convenient to get a patch that is publicly
available, but I don't jump to the conclusion that it is absolutely the best
answer.
> Also know that SP4 is currently in Beta testing and will include builds up
> to .0972.
And then the numbering scheme jumps significantly (SP4 beta is .2026, yet
there are several .973+ hotfixes already available).
I'm still curious why they chose to break at .977 to use the new hotfix
installer, breaking the chain of cumulative hotfixes. It would have been a
much more logical break, IMHO, to wait for SP4 -- introduce the new hotfix
installer at a stable, fully tested service pack... <shrug>
A|||Yes, I found this out with .0859. The biggest reason I jumped on .0878, it
seemed more stable.
I also noticed the large build increase on deploying the SP4 Beta (2026?)
However, from the fix list off of the beta description, it only includes
fixes through .0972 and I've seen KB listing all the way up to .1000 +. So,
I'm not sure how a Build including only fixes to the 972 level could be
labeled 2026. That's got me stumped. Not to mention that it bombs the
replication if you've already applied 878. I've got that thread running in
the Beta newsgroup.
Now, I wouldn't want to slam PSS, because they've been very helpful,
especially in crash circumstances; however, they have been known to push hot
fixes a little too eagerly as well. I figure once it has gone public, at
least, it has somewhat stabilized. But yes, I realize, that even SP level
code bases can introduce new bugs.
As always, thanks for your insight.
Sincerely,
Anthony Thomas
"Aaron [SQL Server MVP]" <ten.xoc@.dnartreb.noraa> wrote in message
news:%23zfCiKbAFHA.3940@.TK2MSFTNGP09.phx.gbl...
> As you know, SQL Server QFEs are cumulative.
That is true until .977. For example, to install .993 on .760, you first
need to get to the .977 hotfix installer -- .993 won't install on < .977 (it
will complain about missing prerequisites).
> The fix described in this
> article is Build 2000.00.0807. The latest publicly available hot fix is
> Build .0878, which would include a fix for this error.
However, it is not always the best to just install the latest. He may
prefer to go through PSS, demonstrate that .807 is his issue, and fix that.
.878 might bring about other problems that he doesn't have time to fully
test. Sure, it might be more convenient to get a patch that is publicly
available, but I don't jump to the conclusion that it is absolutely the best
answer.
> Also know that SP4 is currently in Beta testing and will include builds up
> to .0972.
And then the numbering scheme jumps significantly (SP4 beta is .2026, yet
there are several .973+ hotfixes already available).
I'm still curious why they chose to break at .977 to use the new hotfix
installer, breaking the chain of cumulative hotfixes. It would have been a
much more logical break, IMHO, to wait for SP4 -- introduce the new hotfix
installer at a stable, fully tested service pack... <shrug>
A|||> Yes, I found this out with .0859. The biggest reason I jumped on .0878,
it
> seemed more stable.
I've heard many similar complaints.
> I also noticed the large build increase on deploying the SP4 Beta (2026?)
> However, from the fix list off of the beta description, it only includes
> fixes through .0972 and I've seen KB listing all the way up to .1000 +.
I'm not sure how they're going to deal with this (other than slip the other
fixes in before release). BTW, can you share any articles that are 1000+?
The highest I can find is .993 (and not using the pitiful on-again off-again
search at support.microsoft.com).|||> I also noticed the large build increase on deploying the SP4 Beta (2026?)
> However, from the fix list off of the beta description, it only includes
> fixes through .0972 and I've seen KB listing all the way up to .1000 +.
So,
> I'm not sure how a Build including only fixes to the 972 level could be
> labeled 2026. That's got me stumped.
My guess is that the hotfixes that have been pushed since December 9th (when
2026 was forged) have been worked into both branches, and will be included
in SP4 when it goes live... what that means for the beta process, I'm not
sure... clearly we can't be testing everything right up to the point of
release, and we're already using a build that's behind on at least 6 unique
hotfixes so far.
A|||Yea, I agree; they have to cut the beta release off at some point. I've
asked about the build number with no response though. I think it was a goof
but until I hear otherwise, we'll have to take it as they put it.
Anthony Thomas
"Aaron [SQL Server MVP]" <ten.xoc@.dnartreb.noraa> wrote in message
news:%23c8UxPcAFHA.3264@.TK2MSFTNGP12.phx.gbl...
> I also noticed the large build increase on deploying the SP4 Beta (2026?)
> However, from the fix list off of the beta description, it only includes
> fixes through .0972 and I've seen KB listing all the way up to .1000 +.
So,
> I'm not sure how a Build including only fixes to the 972 level could be
> labeled 2026. That's got me stumped.
My guess is that the hotfixes that have been pushed since December 9th (when
2026 was forged) have been worked into both branches, and will be included
in SP4 when it goes live... what that means for the beta process, I'm not
sure... clearly we can't be testing everything right up to the point of
release, and we're already using a build that's behind on at least 6 unique
hotfixes so far.
A|||Yea, that's why I said .1000 +, I know I came across one article that listed
a .1193 or something like that and have been searching ever since but no
luck...yet.
I also came across a slip-steamed install for MSDE for the new, free MS ADS
(?) server. It's an SUS system for deploying patches to Server Systems. It
had a build of 8.00.0880 but can't find a build description or why it was
inserted into this installation. MS also slip-streamed SP3a, of course, and
MS03-031, individually, like 880 wouldn't install without it.
Anthony Thomas
"Aaron [SQL Server MVP]" <ten.xoc@.dnartreb.noraa> wrote in message
news:uCtTUMcAFHA.3088@.TK2MSFTNGP10.phx.gbl...
> Yes, I found this out with .0859. The biggest reason I jumped on .0878,
it
> seemed more stable.
I've heard many similar complaints.
> I also noticed the large build increase on deploying the SP4 Beta (2026?)
> However, from the fix list off of the beta description, it only includes
> fixes through .0972 and I've seen KB listing all the way up to .1000 +.
I'm not sure how they're going to deal with this (other than slip the other
fixes in before release). BTW, can you share any articles that are 1000+?
The highest I can find is .993 (and not using the pitiful on-again off-again
search at support.microsoft.com).