Re: [SQL] Function Parameters in GROUP BY clause cause errors
How amazing is that? I call it a night and come back to find that a bug has been identified and patched while I sleep.
When will it appear in the binaries (I see that the release version is still 8.1.3)? I thought about trying to compile from source, but after looking at the mingw page was too intimidated by figuring out what to download to try compiling source.
Best Regards,
Robert Davidson
-----Original Message-----
From: Tom Lane [mailto:tgl@sss.pgh.pa.us]
Sent: Wednesday, March 22, 2006 8:27 PM
To: Davidson, Robert
Cc: pgsql-sql@postgresql.org; pgsql-bugs@postgresql.org
Subject: Re: [SQL] Function Parameters in GROUP BY clause cause errors
I wrote:
Hmm, this seems like a plpgsql deficiency. It feels it can generate a
separate parameter symbol ($n) for each occurrence of each variable it
passes into a SQL query. But for this query to be legal, the two
instances of IntervalMinutes have to be represented by the *same*
parameter symbol (notice they are not in the regurgitated query).
It would be more efficient anyway to not generate multiple parameters
for the same value, so we oughta fix this.
Patch applied to HEAD and 8.1 branches.
regards, tom lane
On Thu, Mar 23, 2006 at 05:30:28AM -0800, Davidson, Robert wrote:
How amazing is that? I call it a night and come back to find that a bug has been identified and patched while I sleep.
When will it appear in the binaries (I see that the release version is still 8.1.3)? I thought about trying to compile from source, but after looking at the mingw page was too intimidated by figuring out what to download to try compiling source.
We generally don't worry about point releases too much unless there's a
pretty serious bug, so it'll probably take a while.
Unfortunately I don't know of anyone building nightly binaries. This is
something we should think about doing, especially for windows...
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461
Jim C. Nasby wrote:
On Thu, Mar 23, 2006 at 05:30:28AM -0800, Davidson, Robert wrote:
How amazing is that? I call it a night and come back to find that a bug has been identified and patched while I sleep.
When will it appear in the binaries (I see that the release version is still 8.1.3)? I thought about trying to compile from source, but after looking at the mingw page was too intimidated by figuring out what to download to try compiling source.
We generally don't worry about point releases too much unless there's a
pretty serious bug, so it'll probably take a while.Unfortunately I don't know of anyone building nightly binaries. This is
something we should think about doing, especially for windows...
What for? The only binaries we should be letting out are for declared
releases. If the bug is serious enough to warrant cutting a generally
distributed binary for, it's worth making a release for, IMNSHO. The
Windows packaging team usually follow up a release with a package very
quickly indeed.
cheers
andrew
On Fri, Mar 24, 2006 at 09:32:27AM -0500, Andrew Dunstan wrote:
Jim C. Nasby wrote:
On Thu, Mar 23, 2006 at 05:30:28AM -0800, Davidson, Robert wrote:
How amazing is that? I call it a night and come back to find that a bug
has been identified and patched while I sleep.When will it appear in the binaries (I see that the release version is
still 8.1.3)? I thought about trying to compile from source, but after
looking at the mingw page was too intimidated by figuring out what to
download to try compiling source.We generally don't worry about point releases too much unless there's a
pretty serious bug, so it'll probably take a while.Unfortunately I don't know of anyone building nightly binaries. This is
something we should think about doing, especially for windows...What for? The only binaries we should be letting out are for declared
releases. If the bug is serious enough to warrant cutting a generally
distributed binary for, it's worth making a release for, IMNSHO. The
Windows packaging team usually follow up a release with a package very
quickly indeed.
In this case, my guess is that this bugfix isn't important enough to
push a release. If there was a nightly build, the bug reporter could use
that for development until an official release came out.
The only reason I even suggested this is because setting up a windows
build environment is substantially more complicated than building on
unix. If he was on unix, he could easily just cvs up on the appropriate
branch and build it himself.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461
I now have two minor bugs for which patches exist.
I am game to write up some better instructions for installing mingw to compile postgresql on Windows if there are any notes on which packages need to be installed. I have successfully compiled and installed on Linux.
If there were a place to download patches, this would allow moderately adventurous Windows users to roll their own and avoid the overhead of nightly Windows builds.
Best Regards,
Robert Davidson
-----Original Message-----
From: Jim C. Nasby [mailto:jnasby@pervasive.com]
Sent: Friday, March 24, 2006 6:45 AM
To: Andrew Dunstan
Cc: Davidson, Robert; Tom Lane; pgsql-hackers@postgresql.org
Subject: Re: [HACKERS] Nightly builds
On Fri, Mar 24, 2006 at 09:32:27AM -0500, Andrew Dunstan wrote:
Jim C. Nasby wrote:
On Thu, Mar 23, 2006 at 05:30:28AM -0800, Davidson, Robert wrote:
How amazing is that? I call it a night and come back to find that a bug
has been identified and patched while I sleep.When will it appear in the binaries (I see that the release version is
still 8.1.3)? I thought about trying to compile from source, but after
looking at the mingw page was too intimidated by figuring out what to
download to try compiling source.We generally don't worry about point releases too much unless there's a
pretty serious bug, so it'll probably take a while.Unfortunately I don't know of anyone building nightly binaries. This is
something we should think about doing, especially for windows...What for? The only binaries we should be letting out are for declared
releases. If the bug is serious enough to warrant cutting a generally
distributed binary for, it's worth making a release for, IMNSHO. The
Windows packaging team usually follow up a release with a package very
quickly indeed.
In this case, my guess is that this bugfix isn't important enough to
push a release. If there was a nightly build, the bug reporter could use
that for development until an official release came out.
The only reason I even suggested this is because setting up a windows
build environment is substantially more complicated than building on
unix. If he was on unix, he could easily just cvs up on the appropriate
branch and build it himself.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461
Import Notes
Resolved by subject fallback
On Fri, Mar 24, 2006 at 06:49:43AM -0800, Davidson, Robert wrote:
I now have two minor bugs for which patches exist.
I am game to write up some better instructions for installing mingw to compile postgresql on Windows if there are any notes on which packages need to be installed. I have successfully compiled and installed on Linux.
If there were a place to download patches, this would allow moderately adventurous Windows users to roll their own and avoid the overhead of nightly Windows builds.
Much better to just use CVS and grab the tip of the appropriate branch.
AFAIK the popular windows CVS programs should be able to talk to the
anonomous CVS server.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461
Davidson, Robert wrote:
I now have two minor bugs for which patches exist.
I am game to write up some better instructions for installing mingw to compile postgresql on Windows if there are any notes on which packages need to be installed. I have successfully compiled and installed on Linux.
If there were a place to download patches, this would allow moderately adventurous Windows users to roll their own and avoid the overhead of nightly Windows builds.
If the patches have been applied to the stable branch, you should not
need to download them yourself; just check out the relevant branch and
build it. e.g. to check out the 8.1 stable branch, do:
cvs -d :pserver:anoncvs@anoncvs.postgresql.org:/projects/cvsroot co
-P -r REL8_1_STABLE pgsql
And if they haven't been applied, you should wait till they have been,
since only then will you know that they have been properly reviewed.
In short, you shouldn't need to apply patches directly, unless you are
testing to assist the patch author.
cheers
andrew
Andrew Dunstan <andrew@dunslane.net> writes:
Davidson, Robert wrote:
If there were a place to download patches, this would allow moderately adventurous Windows users to roll their own and avoid the overhead of nightly Windows builds.
... you shouldn't need to apply patches directly, unless you are
testing to assist the patch author.
In any case, patches that are done but not yet applied would be
available from the pgsql-patches archives. Assuming you've got a copy
of "patch" or equivalent, you could cherry-pick unapplied stuff from
there.
This is all old hat to longtime Unix users, but I think it's a vanishing
skill among newer ones because of the prevalence of package-based
distributions. And of course Windows and Mac users are completely
unaccustomed to the idea of rolling their own builds. Documentation
wouldn't be a bad idea.
regards, tom lane
I am game to write up some better instructions for installing
mingw to compile postgresql on Windows if there are any notes
on which packages need to be installed. I have successfully
compiled and installed on Linux.
I would be very grateful for this.
It would also be very helpful to have a list of recommended software
versions and from where to download them.
I for one am still struggling with Kerberos on Windows, and it would
be a great help for me to know where to get libraries and headers from.
Yours,
Laurenz Albe
Import Notes
Resolved by subject fallback
I am game to write up some better instructions for
installing mingw to
compile postgresql on Windows if there are any notes on
which packages
need to be installed. I have successfully compiled and installed on
Linux.I would be very grateful for this.
It would also be very helpful to have a list of recommended
software versions and from where to download them.I for one am still struggling with Kerberos on Windows, and
it would be a great help for me to know where to get
libraries and headers from.
The pginstaller README might help you - it contains the steps used to
build it for the installer...
http://cvs.pgfoundry.org/cgi-bin/cvsweb.cgi/pginstaller/pginst/README?re
v=1.18&content-type=text/x-cvsweb-markup
right at the bottom.
//Magnus
Import Notes
Resolved by subject fallback
I for one am still struggling with Kerberos on Windows, and
it would be a great help for me to know where to get
libraries and headers from.The pginstaller README might help you - it contains the steps used to
build it for the installer...
One of my problems is that it needs VC++ to build, and I don't have
that.
I tried the precompiled binaries, but the headers declare all functions
as __stdcall which breaks my mingw build.
If I change the headers to omit that __stdcall, libpq.dll builds but
crashes.
Any help would be welcome.
Yours,
Laurenz Albe
Import Notes
Resolved by subject fallback
I for one am still struggling with Kerberos on Windows,
and it would
be a great help for me to know where to get libraries and headers
from.The pginstaller README might help you - it contains the
steps used to
build it for the installer...
One of my problems is that it needs VC++ to build, and I
don't have that.
I would expect it to build fine with the free download they have these
days, but I haven't tried it.
//Magnus
Import Notes
Resolved by subject fallback