Thread: Promoting the enhanced branch
As any of you who watch pgsql-committers will have seen, I have now moved the enahnced branch to CVS tip. Specifically, I have: 1) Created a REL-08_01_STABLE branch from CVS tip 2) Moved the REL-07_03_ENHANCED code to CVS tip 3) Bumped the version number to 08.02.0001 4) Standardised filenames and directory names (but not library names) to reflect that we now only have an ODBC 3.51, Unicode build. 5) Defaulted the ODBC version number to 0x351 in all places (previously it had different values in different files) 6) updated the Windows installer as appropriate, and introduced a new version specific element to the path to allow concurrent installation with 08.01.xxxx. 7) Updated all libnames to psqlodbc35w.(so|lib|dylib|whatever). 8) Updated the release notes. The code builds as expected on Linux (with unixODBC) and Windows XP with VC++ 6.0. Please take a few minutes to checkout the code and ensure everything looks as it should if you are able. Thanks, Dave.
Hi Dave. Addition of some information. > The code builds as expected on Linux (with unixODBC) and Windows XP with > VC++ 6.0. VC6 and VC7 are supported in Windows. Moreover, Only VC7 supports a Delay Load. iODBC and unixODBC are supported in Linux. > > Please take a few minutes to checkout the code and ensure everything > looks as it should if you are able. Preparing environment, it can see soon. My environment is now confused.:-( Regards, Hiroshi Saito
> -----Original Message----- > From: Hiroshi Saito [mailto:z-saito@guitar.ocn.ne.jp] > Sent: 10 April 2006 11:15 > To: Dave Page; Hiroshi Inoue > Cc: pgsql-odbc@postgresql.org; Ludek Finstrle > Subject: Re: [ODBC] Promoting the enhanced branch > > Hi Dave. > > Addition of some information. > > > The code builds as expected on Linux (with unixODBC) and Windows XP > > with > > VC++ 6.0. > > VC6 and VC7 are supported in Windows. > Moreover, Only VC7 supports a Delay Load. > iODBC and unixODBC are supported in Linux. Yes, well I don't have iODBC so didn't test that :-p As for the VC++ version issue - we currently use VC++ 6.0 for all PostgreSQL/pgAdmin related builds (that aren't mingw based), otherwise we need to ship VC7 runtimes as well (the VC6 ones are basically everywhere these days). I've been considering biting the bullet and doing this anyway, just haven't done so yet. If I do though, I'll probably start building with VC8 - of I've gotta distribute runtimes, they might as well be the latest! > > > > Please take a few minutes to checkout the code and ensure > everything > > looks as it should if you are able. > > Preparing environment, it can see soon. > My environment is now confused.:-( :-( Give it some coffee - that might help. Regards, Dave.
Am Samstag, 8. April 2006 21:01 schrieb Dave Page: > 7) Updated all libnames to psqlodbc35w.(so|lib|dylib|whatever). What is that supposed to accomplish? If we're dropping the Unicode/ANSI dichotomy, I'd rather go back to plain psqlodbc.so for the Unix builds. I'm not sure about the exact implications on Windows. But in general I think this will just annoy people if they have to change their configurations for no apparent reason. -- Peter Eisentraut http://developer.postgresql.org/~petere/
> -----Original Message----- > From: Peter Eisentraut [mailto:peter_e@gmx.net] > Sent: 10 April 2006 12:50 > To: pgsql-odbc@postgresql.org > Cc: Dave Page > Subject: Re: [ODBC] Promoting the enhanced branch > > Am Samstag, 8. April 2006 21:01 schrieb Dave Page: > > 7) Updated all libnames to psqlodbc35w.(so|lib|dylib|whatever). > > What is that supposed to accomplish? If we're dropping the > Unicode/ANSI dichotomy, I'd rather go back to plain > psqlodbc.so for the Unix builds. I'm not sure about the > exact implications on Windows. But in general I think this > will just annoy people if they have to change their > configurations for no apparent reason. Consistency with the actual ODBC version supported - it was psqlodbc30w.so (http://cvs.pgfoundry.org/cgi-bin/cvsweb.cgi/psqlodbc/psqlodbc/Makefile. am?rev=1.11.2.4&content-type=text/x-cvsweb-markup&only_with_tag=REL-07_0 3_ENHANCED) which is clearly wrong if the default build is for ODBC 3.51. I have no objection if you want to change it to psqlodbc.so though. On Windows it's a non-issue, though personally I would prefer it to be psqlodbc.dll. I'd also like to drop the naming variations in psqlodbc.h, but I'm wary of being too hasty to do that. Regards, Dave.
"Dave Page" <dpage@vale-housing.co.uk> writes:
> I have no objection if you want to change it to psqlodbc.so though.
+1 ... that's how the Red Hat RPMs have always installed it ...
regards, tom lane
I had the chance to grab this and build it on Fedora Core 4 and Solaris. Seems to work well.
Is there a planned release date for 08.02.0001?
Thanks,
Alan
--
"Dealing with failure is easy: Work hard to improve. Success is also easy to handle: You've solved the wrong problem. Work hard to improve."
-- Alan J. Perlis
Is there a planned release date for 08.02.0001?
Thanks,
Alan
On 4/8/06, Dave Page <dpage@vale-housing.co.uk> wrote:
As any of you who watch pgsql-committers will have seen, I have now
moved the enahnced branch to CVS tip. Specifically, I have:
1) Created a REL-08_01_STABLE branch from CVS tip
2) Moved the REL-07_03_ENHANCED code to CVS tip
3) Bumped the version number to 08.02.0001
4) Standardised filenames and directory names (but not library names) to
reflect that we now only have an ODBC 3.51, Unicode build.
5) Defaulted the ODBC version number to 0x351 in all places (previously
it had different values in different files)
6) updated the Windows installer as appropriate, and introduced a new
version specific element to the path to allow concurrent installation
with 08.01.xxxx.
7) Updated all libnames to psqlodbc35w.(so|lib|dylib|whatever).
8) Updated the release notes.
The code builds as expected on Linux (with unixODBC) and Windows XP with
VC++ 6.0.
Please take a few minutes to checkout the code and ensure everything
looks as it should if you are able.
Thanks, Dave.
--
"Dealing with failure is easy: Work hard to improve. Success is also easy to handle: You've solved the wrong problem. Work hard to improve."
-- Alan J. Perlis
-----Original Message----- From: "Alan Potter"<alanjpotter@gmail.com> Sent: 12/04/06 19:17:36 To: "pgsql-odbc@postgresql.org"<pgsql-odbc@postgresql.org> Subject: Re: [ODBC] Promoting the enhanced branch > I had the chance to grab this and build it on Fedora Core 4 and Solaris. > Seems to work well. Good, thanks for the feedback. > Is there a planned release date for 08.02.0001? 08.02.0001 will never be a formal release - look out for 08.02.0100 (only versions ending in 00 will be formal from now on),but I have no idea when yet. Regards, Dave -----Unmodified Original Message----- I had the chance to grab this and build it on Fedora Core 4 and Solaris. Seems to work well. Is there a planned release date for 08.02.0001? Thanks, Alan On 4/8/06, Dave Page <dpage@vale-housing.co.uk> wrote: > > As any of you who watch pgsql-committers will have seen, I have now > moved the enahnced branch to CVS tip. Specifically, I have: > > 1) Created a REL-08_01_STABLE branch from CVS tip > 2) Moved the REL-07_03_ENHANCED code to CVS tip > 3) Bumped the version number to 08.02.0001 > 4) Standardised filenames and directory names (but not library names) to > reflect that we now only have an ODBC 3.51, Unicode build. > 5) Defaulted the ODBC version number to 0x351 in all places (previously > it had different values in different files) > 6) updated the Windows installer as appropriate, and introduced a new > version specific element to the path to allow concurrent installation > with 08.01.xxxx. > 7) Updated all libnames to psqlodbc35w.(so|lib|dylib|whatever). > 8) Updated the release notes. > > The code builds as expected on Linux (with unixODBC) and Windows XP with > VC++ 6.0. > > Please take a few minutes to checkout the code and ensure everything > looks as it should if you are able. > > Thanks, Dave. > -- "Dealing with failure is easy: Work hard to improve. Success is also easy to handle: You've solved the wrong problem. Work hard to improve." -- Alan J. Perlis
Hi Dave. From: "Dave Page" > doing this anyway, just haven't done so yet. If I do though, I'll > probably start building with VC8 - of I've gotta distribute runtimes, > they might as well be the latest! Done.:-) It is built and a test is favorable. http://developer.pgadmin.org/~hiroshi/psqlODBC This build is VS2005Express and PSDK-x86 (VC8) However, nmake spec. is not created yet. openssl-0.7.8a postgresql-8.1.3 psqlodbc-08.02.0001(CVS) Regards, Hiroshi Saito
> -----Original Message----- > From: Hiroshi Saito [mailto:z-saito@guitar.ocn.ne.jp] > Sent: 14 April 2006 04:41 > To: Dave Page; Hiroshi Inoue > Cc: pgsql-odbc@postgresql.org > Subject: Re: [ODBC] Promoting the enhanced branch > > Hi Dave. > > From: "Dave Page" > > doing this anyway, just haven't done so yet. If I do though, I'll > > probably start building with VC8 - of I've gotta distribute > runtimes, > > they might as well be the latest! > > Done.:-) > It is built and a test is favorable. > http://developer.pgadmin.org/~hiroshi/psqlODBC > > This build is VS2005Express and PSDK-x86 > (VC8) > However, nmake spec. is not created yet. > > openssl-0.7.8a > postgresql-8.1.3 > psqlodbc-08.02.0001(CVS) Hi Hiroshi, It works well in VS2K5 Team Edition, however, the paths to PostgreSQL and OpenSSL are non-standard. We should really use the default paths that each install to for maximum compatibility with all installations. The attached project file works correctly with correctly with default binary installations of OpenSSL 0.9.8 (from http://www.openssl.org/related/binaries.html), and pgInstaller 8.1.3. Any objections to this being applied? Regards, Dave.
Attachment
> -----Original Message----- > From: pgsql-odbc-owner@postgresql.org > [mailto:pgsql-odbc-owner@postgresql.org] On Behalf Of Dave Page > Sent: 14 April 2006 21:15 > To: Hiroshi Saito; Hiroshi Inoue > Cc: pgsql-odbc@postgresql.org > Subject: Re: [ODBC] Promoting the enhanced branch > > > > The attached project file works correctly with correctly with > default binary installations of OpenSSL 0.9.8 (from > http://www.openssl.org/related/binaries.html), and pgInstaller 8.1.3. > Any objections to this being applied? Forgot to mention - it also cleans up some of the target paths which are not quite right. Regards, Dave.