This is part of The Pile, a partial archive of some open source mailing lists and newsgroups.
Subject: [ANNOUNCE] PostgreSQL v7.0.1 Released From: The Hermit Hacker <scrappy@hub.org> Date: Fri, 2 Jun 2000 13:49:19 -0300 (ADT) 06-01-2000 - PostgreSQL v7.0.1 Released The PostgreSQL Global Development Group is proud to announce the release of PostgreSQL v7.0.1. This is essentially acleanup of v7.0. A dump/restore is not required if you're moving from v7.0. If you're migrating from a release earlier than v7.0 a dump/restore will be necessary. The release is available on ftp.postgresql.org, at: ftp.postgresql.org/pub/source/v7.0.1 As well as our various mirror affiliates ... Changes Fix many CLUSTER failures (Tom) Allow ALTER TABLE RENAME works on indexes (Tom) Fix plpgsql to handle datetime->timestamp and timespan->interval (Bruce) New configure --with-setproctitle switch to use setproctitle() (Marc, Bruce) Fix the off by one errors in ResultSet from 6.5.3, and more. jdbc ResultSet fixes (Joseph Shraibman) optimizer tunings (Tom) Fix create user for pgaccess Fix for UNLISTEN failure IRIX fixes (David Kaelbling) QNX fixes (Andreas Kardos) Reduce COPY IN lock level (Tom) Change libpqeasy to use PQconnectdb() style parameters (Bruce) Fix pg_dump to handle OID indexes (Tom) Fix small memory leak (Tom) Solaris fix for createdb/dropdb Fix for non-blocking connections (Alfred Perlstein) Fix improper recovery after RENAME TABLE failures (Tom) Copy pg_ident.conf.sample into /lib directory in install (Bruce) === Subject: [HACKERS] 7.0.1 Problems. From: Lamar Owen <lamar.owen@wgcr.org> Date: Sun, 04 Jun 2000 00:49:01 -0400 Well, in attempting to build RPMs for 7.0.1, I have unearthed a 26 megabyte problem. 26 megabytes is the difference in the size of an unpacked 7.0.1 source tree versus 7.0. 7.0.1 is 26 MB _smaller_. What's missing? A diff of a du on each tree shows that the unpacked docs are missing in 7.0.1. This causes a make all in the doc tree to fail (which was reported earlier this week on-list) -- which make all is required by the RPMs for the documentation build. However, that is only 7 MB of the 26 missing MB. So, I do a diff of find -print on each tree..... I also do a wc. The wc results are revealing -- 7.0's tree has 4,172 files. 7.0.1's tree has 2,562 files. Hmmm.... The PostScript docs are missing in 7.0.1, guys. Except for internals.ps, that is. Hmmm... Where is the rest of the space? Well, there are some *.o and *.so files laying around in the 7.0 tree. No, I haven't done a build here -- but, there is a complete build here -- or, it _looks_ like a complete build. 7.0.1 doesn't have any *.o's -- which is good! So, the *.o and *.so's (refint.so and autoinc.so) account for the majority of the missing 26MB. But, the docs are still missing...and I see no note about docs in HISTORY, so I'm assuming that them being missing is not intentional. I may have to make a patch from 7.0 to 7.0.1 for the non-doc portion, and patch a 7.0 tree (minus the *.o's), and generate a tree to put into any 7.0.1 RPMs I might generate. Until then (or 7.0.1-and-a-half is released), there will be no new RPMs, as I really don't have any intention of building doc-less RPMs.... :-(. Sorry I didn't catch this sooner, Marc. I should have followed through on my hunch that I should build RPMs during the prerelease period, after you announced here that the 7.0.1 release candidate was available -- regrettably, I did not do that build. RPM-building will show problems other builds won't. === Subject: Re: [HACKERS] 7.0.1 Problems. From: The Hermit Hacker <scrappy@hub.org> Date: Sun, 4 Jun 2000 23:50:41 -0300 (ADT) On Sun, 4 Jun 2000, Lamar Owen wrote: > On Sun, 04 Jun 2000, The Hermit Hacker wrote: > > On Sun, 4 Jun 2000, Alfred Perlstein wrote: > > > * Lamar Owen <lamar.owen@wgcr.org> [000604 13:57] wrote: > > > > Unless there have been changes, I figured I'd take what I needed from > > > > the 7.0 tarball, unless a 7.0.1 re-release or 7.0.2 release is planned > > > > soon. I just want to see where the direction is going to be before doing > > > > that. > > > > I wouldn't, shipping outdated docs is a nice way of shooting your > > > users in the feet. > > > actually, were there much changes to the docs themselves since the release > > of v7.0? the tar files I can find on the site are dated May 8th ... > > The major changes deal with the changed CVSROOT. There are other changes -- > a diff -uNr between postgresql-7.0/doc/src and postgresql-7.0.1/doc/src is 100K > or so, with the majority of it being $Header differences. > > I will wait to package 7.0.1 RPMs until a direction is set by Steering on this > issue, or a week passes. If a week passes without a set direction, I'm going > to package 7.0.1 RPMs with the 7.0 PostScript docs, unless I get my own > jade/DocBook system up and running building the docs here first. (RedHat 6.2 > ships with a jade/DocBook SGML toolset -- I just have to learn how to use it in > this context.) > > Your call. okay, since I've obviously screwed up naming conventions on ppl (.7.0.1 vs -7.0.1) and missed the docs, I'm just going to do a 're-release' with these problems fixed and call it v7.0.2 ... I'm not going to bother announcing this, cause there are no changes other then cleaning up the packaging, but don't want to confuse ppl by just changing the contents of hte existing tar files ... If anyone feels like looking it over and telling me if I'm missing anythign else, the 'generation script' is in ftp://ftp.postgresql.org/www/supported-bin/mk-release (as is mk-snapshot) ... please feel free to suggest any changes ... === Subject: Re: [HACKERS] 7.0.1 Problems. From: Lamar Owen <lamar.owen@wgcr.org> Date: Sun, 4 Jun 2000 22:15:40 -0400 On Sun, 04 Jun 2000, The Hermit Hacker wrote: > these problems fixed and call it v7.0.2 ... I'm not going to bother > announcing this, cause there are no changes other then cleaning up the > packaging, but don't want to confuse ppl by just changing the contents of > hte existing tar files ... Well, I hate to say it, but, in keeping with the spirit of 6.4.1->6.4.2, a short note to announce and general might be in order. > If anyone feels like looking it over and telling me if I'm missing > anythign else, the 'generation script' is in > ftp://ftp.postgresql.org/www/supported-bin/mk-release (as is > mk-snapshot) ... please feel free to suggest any changes ... Add a copy of the *.ps.gz docs after the copy of the tarred html docs. I _think_ that's all that's missing.... tar one up, and I'll try an RPM build tomorrow morning. I won't guarantee that an RPM build will catch all problems, but it will catch most. Not a bad script, BTW. === Subject: Re: [HACKERS] 7.0.1 Problems. From: Lamar Owen <lamar.owen@wgcr.org> Date: Sun, 4 Jun 2000 22:40:33 -0400 On Sun, 04 Jun 2000, Lamar Owen wrote: > On Sun, 04 Jun 2000, The Hermit Hacker wrote: > > these problems fixed and call it v7.0.2 ... I'm not going to bother > > announcing this, cause there are no changes other then cleaning up the > > packaging, but don't want to confuse ppl by just changing the contents of > > hte existing tar files ... > Well, I hate to say it, but, in keeping with the spirit of 6.4.1->6.4.2, a > short note to announce and general might be in order. Well, after writing that, and then re-reading it, I do want to say that I understand that this is different from the 6.4.1 mispackage -- as it was a package of the then CURRENT branch versus the stable branch, and was thus a more serious problem than the present one. I was not intending to make this one to be bigger than it is -- it is indeed a minor packaging error, not of the magnitude of 6.4.1. HOWEVER, the spirit of 6.4.1-6.4.2 is to at least make a brief note available. And I'm glad the 6.4.1 problem has not been repeated. My apologies. As an RPM packager, I have it a little easier than you do, Marc -- I can just release a minor update to the same version. Problems with 7.0-1? No problem -- here's 7.0-2. Save version of the program -- different release of the package. === Subject: Re: [HACKERS] 7.0.1 Problems. From: Lamar Owen <lamar.owen@wgcr.org> Date: Sun, 4 Jun 2000 23:11:14 -0400 On Sun, 04 Jun 2000, The Hermit Hacker wrote: > On Sun, 4 Jun 2000, Lamar Owen wrote: > > Add a copy of the *.ps.gz docs after the copy of the tarred html docs. I > > _think_ that's all that's missing.... tar one up, and I'll try an RPM build > > tomorrow morning. I won't guarantee that an RPM build will catch all problems, > > but it will catch most. Not a bad script, BTW. > I believe it was decided already that the .ps.gz files would be available > through the web, but not as part of the tar files themselves ... same with > the other formats (A1?) that thomas had worked on ... Ok, then I guess I'll drop them from the RPM, unless they are requested by popular demand, in which case I can build either a separate package for them, or I can incorporate them as separate source files.... It'll depend upon response to RPMs without the PostScript files. That saves a little space, too! Of course, I'll include a pointer to them in my README.rpm. It is now coming back to me about the discussion on that issue -- but, I then found them in the 7.0 tarball, and misunderstood that they were still going to be included. Just a minor change, no biggie. Ok, when 7.0.2 is ready, I'll run a trial build.... === Subject: Re: [HACKERS] 7.0.1 Problems. From: Bruce Momjian <pgman@candle.pha.pa.us> Date: Sun, 4 Jun 2000 23:51:10 -0400 (EDT) > > okay, since I've obviously screwed up naming conventions on ppl (.7.0.1 vs > -7.0.1) and missed the docs, I'm just going to do a 're-release' with > these problems fixed and call it v7.0.2 ... I'm not going to bother > announcing this, cause there are no changes other then cleaning up the > packaging, but don't want to confuse ppl by just changing the contents of > hte existing tar files ... > > If anyone feels like looking it over and telling me if I'm missing > anythign else, the 'generation script' is in > ftp://ftp.postgresql.org/www/supported-bin/mk-release (as is > mk-snapshot) ... please feel free to suggest any changes ... We will need to add release note changes, and change the install files and other branding to mark it as 7.0.2. Let me know if you want me to do it. === Subject: [HACKERS] [v7.0.2] Very small cleanup release ... From: The Hermit Hacker <scrappy@hub.org> Date: Mon, 5 Jun 2000 15:23:47 -0300 (ADT) This is purely a 'we missed some stuff' release, with the only change being: Changes ------- Added documentation to tarball. So, if you are already running, you don't need this ... this is just to help Lamar out with the RPMs more then anything ... === Subject: [HACKERS] New warning code for missing FROM relations From: Bruce Momjian <pgman@candle.pha.pa.us> Date: Sat, 3 Jun 2000 00:40:48 -0400 (EDT) I have committed new warning code to alert users who auto-create relations without knowing it. The new code does not throw a warning for: SELECT pg_language.*; but does throw one for: SELECT pg_language.* FROM pg_class The code issues the warning if it auto-creates a range table entry, and there is already a range table entry identified as coming from a FROM clause. Correlated subqueries should not be a problem because they are not auto-created. The regression tests run fine, except for: SELECT * INTO TABLE tmp1 FROM tmp WHERE onek.unique1 < 2; NOTICE: Adding missing FROM-clause entry for table onek DROP TABLE tmp1; SELECT * INTO TABLE tmp1 FROM tmp WHERE onek2.unique1 < 2; NOTICE: Adding missing FROM-clause entry for table onek2 Seems those warnings are justified. In fact, I am not even sure what these queries are trying to do. I have modified the expected files so they now expect to see the warnings. A bigger question is whether we should issue ERROR for these queries. If they have already used a FROM clause, why would they have other relations not specified there? If people have other suggestions about this, I would be glad to modify the code. A new function warnAutoRange() does the checking. === Subject: Re: AW: [HACKERS] New warning code for missing FROM relations From: Philip Warner <pjw@rhyme.com.au> Date: Tue, 06 Jun 2000 14:41:35 +1000 At 11:44 5/06/00 -0400, Tom Lane wrote: >Philip Warner <pjw@rhyme.com.au> writes: >> I take it that there is no chance of a compile-time or runtime option to >> disallow this kind of thing in all cases? > >Define "all cases" ... also, just how strict do you want to be? >Disallow use of *any* Postgres feature that is not in SQL92 (that >would include all user datatypes and functions, for example)? Sorry, I should have been a bit more specific! I would like some kind of option to disable all cases of adding tables to FROM clauses by implication. The main issue I have with this feature is it is more likely to be used (by me) by accident (as a result of a typo), and consequently introduce very strange results. I am unaware (yet? ;-}) of any other non-SQL features in PostgreSQL that are likely to cause me the same level of concern. Adding tables to a query seems very dangerous: some people might, for example, expect an automatic natural join on primary/foreign keys if you add a table. === Subject: Re: AW: [HACKERS] New warning code for missing FROM relations From: Peter Eisentraut <peter_e@gmx.net> Date: Tue, 6 Jun 2000 18:08:04 +0200 (CEST) Bruce Momjian writes: > I see the notice as "Hey, you probably did something you didn't want > to do". Again, "probably" means you're not sure, so you leave it up to the user to turn it on or off. === Subject: Re: AW: [HACKERS] New warning code for missing FROM relations From: Bruce Momjian <pgman@candle.pha.pa.us> Date: Tue, 6 Jun 2000 12:13:31 -0400 (EDT) [ Charset ISO-8859-1 unsupported, converting... ] > Zeugswetter Andreas SB writes: > > > Yes, only Bruce and I are of the opinion that it *is* an Error, and I > > guess we want some consensus. > > I agree that it is an error. > > > The notice is imho of the sort: notice this syntax is going to be > > disallowed soon. > > If you can guarantee that each user will only see the notice once, then > okay. :) There is no sense that this is a warning about the syntax changing at some point. It is to warn queries that are probably broken. Seems if they already have a FROM clause, there is no purpose for some tables being in the FROM clause, and others begin auto-created. In other cases, it issues no warning. === Subject: [ANNOUNCE] PostgreSQL 7.0.2-1 RPMset available. From: Lamar Owen <lamar.owen@wgcr.org> Date: Mon, 5 Jun 2000 22:19:51 -0400 RPMs for PostgreSQL 7.0.2 are now available at ftp://ftp.postgresql.org/pub/binary/v7.0.2/redhat-RPM/RPMS/redhat-6.x There is only one major change from the 7.0-3 RPMset (other than the uprev to 7.0.2, which involves the removal of the pre-built PostScript documentation), and that is the removal of the pl/perl subpackage. If you want pl/perl, the lines to build it are still in the spec file, just commented out. There have been reports of difficulties building and/or running pl/perl on platforms other than x86. I still want to thank Karl DeBisschop for getting it to build on x86. There are a number of minor changes; see the spec file's changelog for details. Please read the README.rpm file, distributed in the main RPM in /usr/doc/postgresql-7.0.2/README.rpm, as well as on the ftp site as ftp://ftp.postgresql.org/pub/binary/v7.0.2/redhat-RPM/README. Soon we will also have LinuxPPC binary RPM's available, courtesy of Murray Todd Williams. If you have a platform that we don't have binary RPM's for, and you want to contribute binary RPMs, please let me know. The only condition is that the absolute minimum spec file changes are to be made -- if the changes are more than a little, I'll need your spec file and any patches you may have added. If you are able to get the RPMset to build on an RPM-based OS other than RedHat Linux, send me the patches necessary to do so, along with complete platform information (OS, OS version, CPU Type, CPU model, GCC or other CC version, and any patches made to the main PostgreSQL tarball necessary to build). My goal is to get a single _source_ RPM for all RPM-based OS's (more than just Linux can run RPM....). Thanks in advance for your help. And, of course, if you find packaging problems, let me know, either through the pgsql-ports@postgresql.org mailing list or by e-mailing me directly at lamar.owen@wgcr.org. ===