Re: [Firebird-devel] Coalesce of the damned

2015-10-28 Thread Claudio Valderrama C.
> -Original Message- > From: Arno Brinkman [mailto:fbsupp...@abvisie.nl] > Sent: MiƩrcoles, 28 de Octubre de 2015 21:27 > > I've to dig deep in my memory (ooh boy), but i think the > stream flag NULL is > not treated proper in this case. > > Thinking i've fixed something simular or eve

Re: [Firebird-devel] Coalesce of the damned

2015-10-28 Thread Arno Brinkman
Hi Claudio, Simplified test case: SELECT dt.* FROM RDB$DATABASE LEFT JOIN (SELECT COALESCE(RDB$SECURITY_CLASS, 'NO WAY') AS COALESCE_RET, 'YES THIS TO' AS CONST_RET FROM RDB$DATABASE WHERE 1=1) dt ON (1 = 0); Result: COALESCE_RET CONST_RET --

Re: [Firebird-devel] Coalesce of the damned

2015-10-28 Thread Adriano dos Santos Fernandes
Em 28/10/2015 20:31, Claudio Valderrama C. escreveu: >> -Original Message- >> From: Adriano dos Santos Fernandes [mailto:adrian...@gmail.com] >> Sent: Martes, 27 de Octubre de 2015 6:40 >> >> The problem is that we can't go to evaluate blr_coalesce for a "null >> stream". > > Yes, I thoug

Re: [Firebird-devel] Coalesce of the damned

2015-10-28 Thread Claudio Valderrama C.
> -Original Message- > From: Adriano dos Santos Fernandes [mailto:adrian...@gmail.com] > Sent: Martes, 27 de Octubre de 2015 6:40 > > The problem is that we can't go to evaluate blr_coalesce for a "null > stream". Yes, I thought so but there are two cases with a null stream: - coalesce

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Mark Rotteveel
Yes, it should, but you are assuming that it is so. If there is one thing I have learned when I was a tester: verify your assumptions. And as the old and new API are different, it might be holding on to something that delays the unload. Mark - Reply message - Van: "Vlad Khorsun" Aan:

[Firebird-devel] [FB-Tracker] Created: (CORE-4981) Server crashes and DB becomes corrupt when updating against tables that have been ALTERED

2015-10-28 Thread Tim Kelly (JIRA)
Server crashes and DB becomes corrupt when updating against tables that have been ALTERED - Key: CORE-4981 URL: http://tracker.firebirdsql.org/browse/CORE-4981 Pr

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Vlad Khorsun
28.10.2015 13:42, Mark Rotteveel wrote: > I notice that you use the new API. What happens if you use the old API? Speaking about *dll unload* - i think nothing should be changed. Regards, Vlad -- Firebird-Devel maili

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Mark Rotteveel
On Wed, 28 Oct 2015 12:58:00 +0200, Vlad Khorsun wrote: > 28.10.2015 10:36, Arno Brinkman wrote: >> Hi, >> >> Latest snapshot build Firebird-3.0.0.32134-0_Win32.zip using the >> fbclient.dll >> >> Still 500ms for unloading the fbclient.dll :-( > >I can't reproduce it. How did you measure unl

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Vlad Khorsun
28.10.2015 10:36, Arno Brinkman wrote: Hi, Latest snapshot build Firebird-3.0.0.32134-0_Win32.zip using the fbclient.dll Still 500ms for unloading the fbclient.dll :-( I can't reproduce it. How did you measure unload delay ? Regards, Vlad PS I've attached source code of my tiny test prog

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Alex Peshkoff
On 10/28/2015 11:36 AM, Arno Brinkman wrote: > Hi, > > Latest snapshot build Firebird-3.0.0.32134-0_Win32.zip using the > fbclient.dll > > Still 500ms for unloading the fbclient.dll :-( Arno, FYI - I've checked on linux, and yes - there is some platform-specific code when unloading library. --

Re: [Firebird-devel] Firebird 3.0 B2 (and lastsnapshot)fbclient.dll(Win32) slow unload?

2015-10-28 Thread Arno Brinkman
Hi, Latest snapshot build Firebird-3.0.0.32134-0_Win32.zip using the fbclient.dll Still 500ms for unloading the fbclient.dll :-( Kind Regards, Arno Brinkman ABVisie a...@abvisie.nl http://www.abvisie.nl -Oorspronkelijk bericht- From: Alex Peshkoff Sent: Tuesday, October 27, 2015