Re: [mail] Re: [HACKERS] Native Win32 sources
Is there a rough date for when they'll be available? I have a development team at work who currently have an M$-Windows box and a Linux box each in order to allow them to read M$-Office documents sent to us and develop against PostgreSQL (which we use in production). I know I could have a shared Linux box with multiple databases and have them bind to that, but one of the important aspects of our application is response time, and you can't accurately measure response times for code changes on a shared system. Having a Win32 native version would save a lot of hassles for me. Al. - Original Message - From: Bruce Momjian [EMAIL PROTECTED] To: Ulrich Neumann [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Monday, November 25, 2002 5:51 PM Subject: [mail] Re: [HACKERS] Native Win32 sources Ulrich Neumann wrote: Hello, i've read that there are 2 different native ports for Windows somewhere. I've searched for them but didn't found them. Is there anyone who can point me to a link or send me a copy of the sources? Oh, you are probably asking about the sources. They are not publically available yet. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED] ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [mail] Re: [HACKERS] Native Win32 sources
Al, to be honest I don't think the Windows native would save hassle, rather it'd probably cause more! No disrespect to those doing the version, read on for reasoning... Yes, you get a beta of a Windows native version just now, yes it probably will not be that long till the source is a available... But how long till it's part of a cosha PostgreSQL release? Version 7.4... Could be up to six months... Do you want to run pre-release versions in the meantime? Don't think so, not in a production environment! So, the real way to save hassle is probably a cheap commodity PC with Linux installed... Or settle for the existing, non-native, Windows version. By the way, just to open Office documents? Have you tried OpenOffice? Regards, Lee Kindness. Al Sutton writes: Is there a rough date for when they'll be available? I have a development team at work who currently have an M$-Windows box and a Linux box each in order to allow them to read M$-Office documents sent to us and develop against PostgreSQL (which we use in production). I know I could have a shared Linux box with multiple databases and have them bind to that, but one of the important aspects of our application is response time, and you can't accurately measure response times for code changes on a shared system. Having a Win32 native version would save a lot of hassles for me. Al. - Original Message - From: Bruce Momjian [EMAIL PROTECTED] Ulrich Neumann wrote: Hello, i've read that there are 2 different native ports for Windows somewhere. I've searched for them but didn't found them. Is there anyone who can point me to a link or send me a copy of the sources? Oh, you are probably asking about the sources. They are not publically available yet. ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Antw: Re: [HACKERS] Native Win32 sources
Bruce, you're right, i'm asking about the sources because I want to help. Is it possible to help in this case or not? Ulrich Bruce Momjian [EMAIL PROTECTED] 25.11.2002 18:51:39 Ulrich Neumann wrote: Hello, i've read that there are 2 different native ports for Windows somewhere. I've searched for them but didn't found them. Is there anyone who can point me to a link or send me a copy of the sources? Oh, you are probably asking about the sources. They are not publically available yet. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED] -- This e-mail is virus scanned Diese e-mail ist virusgeprueft ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] Postgres Security Expert???
Hi Chris, Just received this from them. Look like he was trying to claim stuff that wasn't true. :-/ Thanks for pointing this out Chris. :) Regards and best wishes, Justin Clift *** Original Message Subject: Re: Demande de renseignements Defi SYSDOOR Date: Tue, 26 Nov 2002 11:04:47 +0100 From: Vergoz Michael (SYSDOOR) [EMAIL PROTECTED] To: [EMAIL PROTECTED] References: 200211260205.gAQ25GTK009595@jenna Dear Clift, Justin Clift PostgreSQL Global Development Group demande des informations son adresse : Son e-mail : [EMAIL PROTECTED] Son téléphone : +61.393631313 Son message : Hi, Just noticed your website mentioning that Michael Vergoz is well known to created security patches for PostgreSQL: http://kernel.sysdoor.com/eng/ Can you please point us in their direction, as we don't know him by name. Right, it's true that i never make _security_ patches for PostGreSQL... As a side thought, would you please be able to correct the spelling of PostgreSQL on the same page. Presently it's spelt PostGreSQL, which is incorrect. Better way, i'v remove postgresql name in the site, as i think you want. Regards and best wishes, Justin Clift -- - Source IP : 203.173.161.124 (p378-tnt1.mel.ihug.com.au) Secure ID : [EMAIL PROTECTED] -- - Best Regards, Vergoz Michael SYSDOOR Founder *** -- My grandfather once told me that there are two kinds of people: those who work and those who take the credit. He told me to try to be in the first group; there was less competition there. - Indira Gandhi ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
Re: [HACKERS] Postgres Security Expert???
FWIW, a search on Google gives some hits for the name on the lists this year. First impressions are that it's not Sir Mondred (or whatever the spelling was). On Tue, 26 Nov 2002, Justin Clift wrote: Hi Chris, Just received this from them. Look like he was trying to claim stuff that wasn't true. :-/ Thanks for pointing this out Chris. :) Regards and best wishes, Justin Clift *** Original Message Subject: Re: Demande de renseignements Defi SYSDOOR Date: Tue, 26 Nov 2002 11:04:47 +0100 From: Vergoz Michael (SYSDOOR) [EMAIL PROTECTED] To: [EMAIL PROTECTED] References: 200211260205.gAQ25GTK009595@jenna Dear Clift, Justin Clift PostgreSQL Global Development Group demande des informations son adresse : Son e-mail : [EMAIL PROTECTED] Son téléphone : +61.393631313 Son message : Hi, Just noticed your website mentioning that Michael Vergoz is well known to created security patches for PostgreSQL: http://kernel.sysdoor.com/eng/ Can you please point us in their direction, as we don't know him by name. Right, it's true that i never make _security_ patches for PostGreSQL... As a side thought, would you please be able to correct the spelling of PostgreSQL on the same page. Presently it's spelt PostGreSQL, which is incorrect. Better way, i'v remove postgresql name in the site, as i think you want. Regards and best wishes, Justin Clift -- - Source IP : 203.173.161.124 (p378-tnt1.mel.ihug.com.au) Secure ID : [EMAIL PROTECTED] -- - Best Regards, Vergoz Michael SYSDOOR Founder *** ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [mail] Re: [HACKERS] Native Win32 sources
Lee, I wouldn't go for 7.4 in production until after it's gone gold, but being able to cut the number of boxes per developer by giving them a Win32 native version would save on everything from the overhead of getting the developers familiar enough with Linux to be able to admin their own systems, to cutting the network usage by having the DB and app on the same system, through to cutting the cost of electricity by only having one box per developer. It would also be a good way of testing 7.4 against our app so we can plan for an upgrade when it's released ;). I've tried open office 1.0.1 and had to ditch it. It had problems with font rendering and tables that ment many of the forms that people sent as word documents had chunks that weren't displayed or printed. We did try it on a box with MS-Word on it to ensure that the setup of the machine wasn't the issue, Word had no problems, OO failed horribly. Thanks for the ideas, Al. - Original Message - From: Lee Kindness [EMAIL PROTECTED] To: Al Sutton [EMAIL PROTECTED] Cc: Bruce Momjian [EMAIL PROTECTED]; Ulrich Neumann [EMAIL PROTECTED]; [EMAIL PROTECTED]; Lee Kindness [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 9:08 AM Subject: Re: [mail] Re: [HACKERS] Native Win32 sources Al, to be honest I don't think the Windows native would save hassle, rather it'd probably cause more! No disrespect to those doing the version, read on for reasoning... Yes, you get a beta of a Windows native version just now, yes it probably will not be that long till the source is a available... But how long till it's part of a cosha PostgreSQL release? Version 7.4... Could be up to six months... Do you want to run pre-release versions in the meantime? Don't think so, not in a production environment! So, the real way to save hassle is probably a cheap commodity PC with Linux installed... Or settle for the existing, non-native, Windows version. By the way, just to open Office documents? Have you tried OpenOffice? Regards, Lee Kindness. Al Sutton writes: Is there a rough date for when they'll be available? I have a development team at work who currently have an M$-Windows box and a Linux box each in order to allow them to read M$-Office documents sent to us and develop against PostgreSQL (which we use in production). I know I could have a shared Linux box with multiple databases and have them bind to that, but one of the important aspects of our application is response time, and you can't accurately measure response times for code changes on a shared system. Having a Win32 native version would save a lot of hassles for me. Al. - Original Message - From: Bruce Momjian [EMAIL PROTECTED] Ulrich Neumann wrote: Hello, i've read that there are 2 different native ports for Windows somewhere. I've searched for them but didn't found them. Is there anyone who can point me to a link or send me a copy of the sources? Oh, you are probably asking about the sources. They are not publically available yet. ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [mail] Re: [HACKERS] Native Win32 sources
On November 26, 2002 06:33 am, Al Sutton wrote:
I wouldn't go for 7.4 in production until after it's gone gold, but being
able to cut the number of boxes per developer by giving them a Win32 native
version would save on everything from the overhead of getting the
developers familiar enough with Linux to be able to admin their own
systems, to cutting the network usage by having the DB and app on the same
system, through to cutting the cost of electricity by only having one box
per developer. It would also be a good way of testing 7.4 against our app
so we can plan for an upgrade when it's released ;).
If your database is of any significant size you probably want a separate
database machine anyway. We run NetBSD everywhere and could easily put the
apps on the database machine but choose not to. We have 6 production servers
running various apps and web servers and they all talk to a central database
machine which has lots of RAM. Forget about bandwidth. Just get a 100MBit
switch and plug everything into it. Network bandwidth won't normally be your
bottleneck. Memory and CPU will be.
We actually have 4 database machines, 3 running transaction databases and 1
with an rsynced copy for reporting purposes. We use 3 networks, 1 for the
app servers to talk to the Internet, 1 for the app servers to talk to the
databases and one for the databases to talk amongst themselves.
Even for development we keep a separate database machine that developers all
use. They run whatever they want - we have people using NetBSD, Linux and
Windows - but they work on one database which is tuned for the purpose. They
can even create their own databases on that system if they want for local
testing.
--
D'Arcy J.M. Cain darcy@{druid|vex}.net | Democracy is three wolves
http://www.druid.net/darcy/| and a sheep voting on
+1 416 425 1212 (DoD#0082)(eNTP) | what's for dinner.
---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to [EMAIL PROTECTED] so that your
message can get through to the mailing list cleanly
Re: [HACKERS] RC2? AIX 4.2.1.0 (fwd)
Problem still exists in RC2 [EMAIL PROTECTED] (Samuel A Horwitz) -- Forwarded message -- Date: Mon, 25 Nov 2002 09:12:55 -0500 (EST) From: Samuel A Horwitz [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: [HACKERS] RC1? AIX 4.2.1.0 (fwd) Sorry forgot to include that I had to add -lssl and -lcrypto t0 the libpq line in Makefile.global.in in the src directory to get ecpg to link as follows 286c286 libpq = -L$(libpq_builddir) -lpq --- libpq = -L$(libpq_builddir) -lpq -lssl -lcrypto [EMAIL PROTECTED] (Samuel A Horwitz) -- Forwarded message -- Date: Mon, 25 Nov 2002 08:45:46 -0500 (EST) From: Samuel A Horwitz [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: [HACKERS] RC1? AIX 4.2.1.0 system = powerpc-ibm-aix4.2.1.0 configure command env CC=gcc ./configure --with-maxbackends=1024 --with-openssl=/usr/local/ssl --enable-syslog --enable-odbc --disable-nls gmake check output file regression.out -- parallel group (13 tests): text varchar oid int2 char boolean float4 int4 name int8 float8 bit numeric boolean ... ok char ... ok name ... ok varchar ... ok text ... ok int2 ... ok int4 ... ok int8 ... ok oid ... ok float4 ... ok float8 ... ok bit ... ok numeric ... ok test strings ... ok test numerology ... ok parallel group (20 tests): lseg date path circle polygon box point time timetz tinterval abstime interval reltime comments inet timestamptz timestamp type_sanity opr_sanity oidjoins point... ok lseg ... ok box ... ok path ... ok polygon ... ok circle ... ok date ... ok time ... ok timetz ... ok timestamp... ok timestamptz ... ok interval ... ok abstime ... ok reltime ... ok tinterval... ok inet ... ok comments ... ok oidjoins ... ok type_sanity ... ok opr_sanity ... ok test geometry ... FAILED test horology ... ok test insert ... ok test create_function_1... ok test create_type ... ok test create_table ... ok test create_function_2... ok test copy ... ok parallel group (7 tests): create_aggregate create_operator triggers vacuum inherit constraints create_misc constraints ... ok triggers ... ok create_misc ... ok create_aggregate ... ok create_operator ... ok inherit ... ok vacuum ... ok parallel group (2 tests): create_view create_index create_index ... ok create_view ... ok test sanity_check ... ok test errors ... ok test select ... ok parallel group (16 tests): select_distinct_on select_into select_having transactions select_distinct random subselect portals arrays union select_implicit case aggregates hash_index join btree_index select_into ... ok select_distinct ... ok select_distinct_on ... ok select_implicit ... ok select_having... ok subselect... ok union... ok case ... ok join ... ok aggregates ... ok transactions ... ok random ... ok portals ... ok arrays ... ok btree_index ... ok hash_index ... ok test privileges ... ok test misc ... ok parallel group (5 tests): portals_p2 cluster rules select_views foreign_key select_views ... ok portals_p2 ... ok rules... ok foreign_key ... ok cluster ... ok parallel group (11 tests): limit truncate temp copy2 domain rangefuncs conversion prepare without_oid plpgsql alter_table limit... ok plpgsql ... ok copy2... ok temp ... ok domain ... ok rangefuncs ... ok prepare ... ok without_oid ... ok conversion ... ok truncate ... ok alter_table ... ok regression.diffs - *** ./expected/geometry-powerpc-aix4.outTue Sep 12 17:07:16 2000 --- ./results/geometry.out Thu Nov 21 21:46:01 2002 *** *** 114,120 | (5.1,34.5)
[HACKERS] Hirarchical queries a la Oracle. Patch.
Hi there! Patch is posted to pgsql-patches. docs inside. SQL 99 version will be later. regards, --- .evgen ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
[HACKERS] How can I view the content of Logs?
Im a new comer of PostgreSQL, could anyone tell me can I view the content of logs( updating a tuple, etc.) ? And if I can, how to do it? Thx!
[HACKERS] possible obvious bug?
I was playing with the Japanese win32 7.2.1 port and I noticed that select 0 / 0 caused the server to crash and restart. I understand that it is a totally unsupported version, but it should be easy enough to check vs. the current version. Note that select 0.0/0.0 worked fine! Merlin ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] Need Quote for 7.3
[EMAIL PROTECTED] (Nigel J. Andrews) writes: On Tue, 19 Nov 2002, Josh Berkus wrote: Folks, We need a quote from a major code contributor to PostgreSQL about the upcoming 7.3 release -- something about how great the new release is, or some of the features in the release. We need this for the 7.3 press release, which will be drafted in 2 days. If you have something to say, please e-mail me, Marc ([EMAIL PROTECTED]) and Justin ([EMAIL PROTECTED]) off-list so we can quote you! I think it's great - but don't quote me on that. :) PostgreSQL. Because life's too short to learn Oracle. :) Billy O'Connor ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
[HACKERS] PostGres and WIN32, a plea!
I had read on one of the newsgroups that there is a planned native port to the win32 platform, is this true? I read most of the win32 thread off of the dev site and it was not clear if this was true. In either case, I would like to advocate such a port to be done, and soon, not for any altruistic reasons, but simply on behalf of myself (a windows applications developer) and the many others who are like me. I personally believe that Postgres has a great deal of potention in the applications market, with the database server packaged along with the application. There is a great deal of need for this for medium to high end windows applications, because there as of yet no existing Microsoft package that can handle it. Postgres is ideally suited for this need because of its rich server side programming interfaces, liberal licensing, and high performance. Mysql, despite their sucking up to the windows crowd, fails on all three of those counts. However they have realized the need for a database embedded application by allowing the mysql server to be linked directly with a windows app (at least, on a technical level), and have talked about providing a single user database .dll. I believe that mysql is not well suited for these types of applications though for stability and performance reasons. For all the talk of speed, I think postgres is the fastest database ever written for pc hardware, with the one possible exception of Microsoft Foxpro (note: not written by Microsoft). Sql server costs to much to ship with an app, and, quite frankly, is rather slow. Postgres could easily springboard into a very strong niche market in embedded applicaions. From there, with increased awareness and developer support on the windows side, it could start pecking at more traditional data services currently dominated by sql server, and (yuck!) access, and their eveil fraternal twin, visual basic. Site note: good strategic positioning in this regard would be an XML shell for postgres (pgxml) and a data provider for .net. Thats my .02$. Many great thanks to the dev team. Please don't let postgres continue to be the software wold's best kept secret. Merlin ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
[HACKERS] 7.3rc2 Test Failures
regression.diffs Description: Binary data regression.out Description: Binary data -- David Wheeler AIM: dwTheory [EMAIL PROTECTED] ICQ: 15726394 http://david.wheeler.net/ Yahoo!: dew7e Jabber: [EMAIL PROTECTED] ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
[HACKERS] re [ANNOUNCE] RC1 Packaged for Testing ... AIX 4.2.1 result
system = powerpc-ibm-aix4.2.1.0 configure command env CC=gcc ./configure --with-maxbackends=1024 --with-openssl=/usr/local/ssl --enable-syslog --enable-odbc --disable-nls gmake check output file regression.out -- parallel group (13 tests): text varchar oid int2 char boolean float4 int4 name int8 float8 bit numeric boolean ... ok char ... ok name ... ok varchar ... ok text ... ok int2 ... ok int4 ... ok int8 ... ok oid ... ok float4 ... ok float8 ... ok bit ... ok numeric ... ok test strings ... ok test numerology ... ok parallel group (20 tests): lseg date path circle polygon box point time timetz tinterval abstime interval reltime comments inet timestamptz timestamp type_sanity opr_sanity oidjoins point... ok lseg ... ok box ... ok path ... ok polygon ... ok circle ... ok date ... ok time ... ok timetz ... ok timestamp... ok timestamptz ... ok interval ... ok abstime ... ok reltime ... ok tinterval... ok inet ... ok comments ... ok oidjoins ... ok type_sanity ... ok opr_sanity ... ok test geometry ... FAILED test horology ... ok test insert ... ok test create_function_1... ok test create_type ... ok test create_table ... ok test create_function_2... ok test copy ... ok parallel group (7 tests): create_aggregate create_operator triggers vacuum inherit constraints create_misc constraints ... ok triggers ... ok create_misc ... ok create_aggregate ... ok create_operator ... ok inherit ... ok vacuum ... ok parallel group (2 tests): create_view create_index create_index ... ok create_view ... ok test sanity_check ... ok test errors ... ok test select ... ok parallel group (16 tests): select_distinct_on select_into select_having transactions select_distinct random subselect portals arrays union select_implicit case aggregates hash_index join btree_index select_into ... ok select_distinct ... ok select_distinct_on ... ok select_implicit ... ok select_having... ok subselect... ok union... ok case ... ok join ... ok aggregates ... ok transactions ... ok random ... ok portals ... ok arrays ... ok btree_index ... ok hash_index ... ok test privileges ... ok test misc ... ok parallel group (5 tests): portals_p2 cluster rules select_views foreign_key select_views ... ok portals_p2 ... ok rules... ok foreign_key ... ok cluster ... ok parallel group (11 tests): limit truncate temp copy2 domain rangefuncs conversion prepare without_oid plpgsql alter_table limit... ok plpgsql ... ok copy2... ok temp ... ok domain ... ok rangefuncs ... ok prepare ... ok without_oid ... ok conversion ... ok truncate ... ok alter_table ... ok regression.diffs - *** ./expected/geometry-powerpc-aix4.outTue Sep 12 17:07:16 2000 --- ./results/geometry.out Thu Nov 21 21:46:01 2002 *** *** 114,120 | (5.1,34.5) | [(1,2),(3,4)] | (3,4) | (-5,-12) | [(1,2),(3,4)] | (1,2) | (10,10)| [(1,2),(3,4)] | (3,4) ! | (0,0) | [(0,0),(6,6)] | (0,0) | (-10,0)| [(0,0),(6,6)] | (0,0) | (-3,4) | [(0,0),(6,6)] | (0.5,0.5) | (5.1,34.5) | [(0,0),(6,6)] | (6,6) --- 114,120 | (5.1,34.5) | [(1,2),(3,4)] | (3,4) | (-5,-12) | [(1,2),(3,4)] | (1,2) | (10,10)| [(1,2),(3,4)] | (3,4) ! | (0,0) | [(0,0),(6,6)] | (-0,0) | (-10,0)| [(0,0),(6,6)]
[HACKERS] Why an array in pg_group?
Hi, Is there any reason why the grolist field in the table pg_group is implemented as an array and not as a separate table? According to the documentation: quote source=Postgresql 7.2 User Manual, chapter 6 near the end Arrays are not sets; using arrays in the manner described in the previous paragraph is often a sign of database misdesign. /quote I have trouble implementing a way to easily check whether a user is part of a group. (I use Apache::AuthDBI to implement authentication and wanted to make a view with columns username, userid , groupname. And installing the contrib/array give's me a postgresql that is different from all the others :-( -- __ Nothing is as subjective as reality Reinoud van Leeuwen[EMAIL PROTECTED] http://www.xs4all.nl/~reinoud __ ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] possible obvious bug?
I was playing with the Japanese win32 7.2.1 port and I noticed that select 0 / 0 caused the server to crash and restart. I understand that it is a totally unsupported version, but it should be easy enough to check vs. the current version. Note that select 0.0/0.0 worked fine! Seems to work fine on my system. postgres=# SELECT version(); version - PostgreSQL 7.2.1 on i686-pc-linux-gnu, compiled by GCC egcs-2.91.66 (1 row) postgres=# SELECT 0/0; ERROR: floating point exception! The last floating point operation either exceeded legal ranges or was a divide by zero postgres=# ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] Need Quote for 7.3
I think it's great - but don't quote me on that. :) PostgreSQL. Because life's too short to learn Oracle. PostgreSQL. For those with more to do than babysit a database. -- Rod Taylor [EMAIL PROTECTED] ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] 7.3rc2 Test Failures
David Wheeler [EMAIL PROTECTED] writes: COPY onek FROM '/usr/local/src/postgresql-7.3rc2/src/test/regress/results/onek.data'; + ERROR: COPY command, running in backend with effective uid 77, could not open file '/usr/local/src/postgresql-7.3rc2/src/test/regress/results/onek.data' for reading. Errno = No such file or directory (2). Looks like a problem on your end... Cheers, Neil -- Neil Conway [EMAIL PROTECTED] || PGP Key ID: DB3C29FC ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] 7.3rc2 Test Failures
On Tuesday, November 26, 2002, at 07:12 AM, Neil Conway wrote: Looks like a problem on your end... Oh, the message finally got through, did it? I chatted with Bruce yesterday and ran the tests again and they all passed. Thanks, David -- David Wheeler AIM: dwTheory [EMAIL PROTECTED] ICQ: 15726394 http://david.wheeler.net/ Yahoo!: dew7e Jabber: [EMAIL PROTECTED] ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
[PERFORM] [HACKERS] Realtime VACUUM, was: performance of insert/delete/update
tom lane wrote: Sure, it's just shuffling the housekeeping work from one place to another. The thing that I like about Postgres' approach is that we put the housekeeping in a background task (VACUUM) rather than in the critical path of foreground transaction commit. Thinking with my marketing hat on, MVCC would be a much bigger win if VACUUM was not required (or was done automagically). The need for periodic VACUUM just gives ammunition to the PostgreSQL opponents who can claim we are deferring work but that it amounts to the same thing. A fully automatic background VACUUM will significantly reduce but will not eliminate this perceived weakness. However, it always seemed to me there should be some way to reuse the space more dynamically and quickly than a background VACUUM thereby reducing the percentage of tuples that are expired in heavy update cases. If only a very tiny number of tuples on the disk are expired this will reduce the aggregate performance/space penalty of MVCC into insignificance for the majority of uses. Couldn't we reuse tuple and index space as soon as there are no transactions that depend on the old tuple or index values. I have imagined that this was always part of the long-term master plan. Couldn't we keep a list of dead tuples in shared memory and look in the list first when deciding where to place new values for inserts or updates so we don't have to rely on VACUUM (even a background one)? If there are expired tuple slots in the list these would be used before allocating a new slot from the tuple heap. The only issue is determining the lowest transaction ID for in-process transactions which seems relatively easy to do (if it's not already done somewhere). In the normal shutdown and startup case, a tuple VACUUM could be performed automatically. This would normally be very fast since there would not be many tuples in the list. Index slots would be handled differently since these cannot be substituted one for another. However, these could be recovered as part of every index page update. Pages would be scanned before being written and any expired slots that had transaction ID's lower than the lowest active slot would be removed. This could be done for non-leaf pages as well and would result in only reorganizing a page that is already going to be written thereby not adding much to the overall work. I don't think that internal pages that contain pointers to values in nodes further down the tree that are no longer in the leaf nodes because of this partial expired entry elimination will cause a problem since searches and scans will still work fine. Does VACUUM do something that could not be handled in this realtime manner? - Curtis ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] Postgres Security Expert???
On Tue, 26 Nov 2002, Justin Clift wrote: Dear Clift, As a side thought, would you please be able to correct the spelling of PostgreSQL on the same page. Presently it's spelt PostGreSQL, which is incorrect. Better way, i'v remove postgresql name in the site, as i think you want. Well, they still have PostGreSQL still on their front page, which is dynamic, as it lists new instrusion attempts every time you refresh it. ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [PORTS] Geometry test on NetBSD (was Re: [HACKERS] RC1?)
At 1:15 AM -0500 11/20/02, Tom Lane wrote: Bruce Momjian [EMAIL PROTECTED] writes: Tom, can you clarify why -0 is valid. The IEEE spec absolutely thinks that -0 and +0 are distinct entities. I don't remember why, at one in the morning ... but if you insist I'm sure that plenty sufficient numerical-analysis reasons can be produced. The guys who wrote that spec knew what they were doing (that's why it's been adopted so universally). It's so that 1/(1/-infinity) == -infinity. There are probably other reasons as well. I'm just guessing here, but it's possible NetBSD acquired the bug by trying to be functional on non-IEEE hardware. I hope that whoever found the problem (I don't see that in this thread) filed a bug report with NetBSD. -- The opinions expressed in this message are mine, not those of Caltech, JPL, NASA, or the US Government. [EMAIL PROTECTED], or [EMAIL PROTECTED] ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [PORTS] Geometry test on NetBSD (was Re: [HACKERS] RC1?)
At 1:51 PM -0500 11/20/02, Tom Lane wrote: Patrick Welche [EMAIL PROTECTED] writes: On Wed, Nov 20, 2002 at 01:21:47PM -0500, Tom Lane wrote: Ah-hah, so it is a version issue --- we could make the resultmap line something like geometry/.*-netbsd1.[0-5]=geometry-positive-zeros NetBSD/i386-1.6H i386-unknown-netbsdelf1.6H (checked 7.3rc1) NetBSD/acorn32-1.6K arm-unknown-netbsdelf1.6K (still building 7.3rc1) Hm, is that elf always there? I'm a little uncomfortable with making the pattern be geometry/.*-netbsd.*1.[0-5]=geometry-positive-zeros as this seems way too lax ... A version like 1.6[A-Z] is a -current, not a release version from in between 1.5.x and 1.6. Different NetBSD ports have converted to elf at different times and not all ports are using elf even with 1.6 released. -- The opinions expressed in this message are mine, not those of Caltech, JPL, NASA, or the US Government. [EMAIL PROTECTED], or [EMAIL PROTECTED] ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [mail] Re: [HACKERS] Native Win32 sources
D'Arcy,
In production the database servers are seperate multi-processor machines
with mirrored disks linked via Gigabit ethernet to the app server.
In development I have people extremely familiar with MS, but not very hot
with Unix in any flavour, who are developing Java and PHP code which is then
passed into the QA phase where it's run on a replica of the production
environment.
My goal is to allow my developers to work on the platform they know (MS),
using as many of the aspects of the production environment as possible (JVM
version, PHP version, and hopefully database version), without needing to
buy each new developer two machines, and incur the overhead of them
familiarising themselves with a flavour of Unix.
Hope this helps you understand where I'm comming from,
Al.
- Original Message -
From: D'Arcy J.M. Cain [EMAIL PROTECTED]
To: Al Sutton [EMAIL PROTECTED]; Lee Kindness [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED]
Sent: Tuesday, November 26, 2002 11:59 AM
Subject: Re: [mail] Re: [HACKERS] Native Win32 sources
On November 26, 2002 06:33 am, Al Sutton wrote:
I wouldn't go for 7.4 in production until after it's gone gold, but
being
able to cut the number of boxes per developer by giving them a Win32
native
version would save on everything from the overhead of getting the
developers familiar enough with Linux to be able to admin their own
systems, to cutting the network usage by having the DB and app on the
same
system, through to cutting the cost of electricity by only having one
box
per developer. It would also be a good way of testing 7.4 against our
app
so we can plan for an upgrade when it's released ;).
If your database is of any significant size you probably want a separate
database machine anyway. We run NetBSD everywhere and could easily put
the
apps on the database machine but choose not to. We have 6 production
servers
running various apps and web servers and they all talk to a central
database
machine which has lots of RAM. Forget about bandwidth. Just get a
100MBit
switch and plug everything into it. Network bandwidth won't normally be
your
bottleneck. Memory and CPU will be.
We actually have 4 database machines, 3 running transaction databases and
1
with an rsynced copy for reporting purposes. We use 3 networks, 1 for the
app servers to talk to the Internet, 1 for the app servers to talk to the
databases and one for the databases to talk amongst themselves.
Even for development we keep a separate database machine that developers
all
use. They run whatever they want - we have people using NetBSD, Linux and
Windows - but they work on one database which is tuned for the purpose.
They
can even create their own databases on that system if they want for local
testing.
--
D'Arcy J.M. Cain darcy@{druid|vex}.net | Democracy is three wolves
http://www.druid.net/darcy/| and a sheep voting on
+1 416 425 1212 (DoD#0082)(eNTP) | what's for dinner.
---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to [EMAIL PROTECTED] so that your
message can get through to the mailing list cleanly
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] PostGres and WIN32, a plea!
-Original Message- From: Merlin Moncure [mailto:[EMAIL PROTECTED]] Sent: 22 November 2002 21:25 To: [EMAIL PROTECTED] Subject: [HACKERS] PostGres and WIN32, a plea! I had read on one of the newsgroups that there is a planned native port to the win32 platform, is this true? I read most of the win32 thread off of the dev site and it was not clear if this was true. Hi Merlin, This is true - the port is being actively worked on (see recent posts from Bruce Momjian). A number of us have also been involved with the closed source beta testing recently. I think postgres is the fastest database ever written for pc hardware, with the one possible exception of Microsoft Foxpro (note: not written by Microsoft). Hmm, ever tried using a large multiuser database such as a finance system using a Foxpro database? Network managers have been known to murder for less... :-) Regards, Dave. ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] [GENERAL] Bug with sequence
[EMAIL PROTECTED] (scott.marlowe) wrote in message news:[EMAIL PROTECTED]... On 21 Nov 2002, Rod Taylor wrote: On Thu, 2002-11-21 at 15:09, scott.marlowe wrote: On 21 Nov 2002, Rod Taylor wrote: On Thu, 2002-11-21 at 14:11, Bruce Momjian wrote: Of course, those would be SQL purists who _don't_ understand concurrency issues. ;-) Or they're the kind that locks the entire table for any given insert. Isn't that what Bruce just said? ;^) I suppose so. I took what Bruce said to be that multiple users could get the same ID. I keep having developers want to make their own table for a sequence, then use id = id + 1 -- so they hold a lock on it for the duration of the transaction. I was just funnin' with ya, but the point behind it was that either way (with or without a lock) that using something other than a sequence is probably a bad idea. Either way, under parallel load, you have data consistency issues, or you have poor performance issues. I'm not familiar with these SQL purists (perhaps the reference is to J. Celko?) but the fact is that it's hard to call SEQUENCE product-specific now that it's in Oracle, DB2, and SQL:2003. The syntaxes do differ a little, usually due to choice of abbreviation, but as far as I can tell the internals are similar across implementations. Peter Gulutzan Author of Sequences And Identity Columns (http://dbazine.com/gulutzan4.html) ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] [GENERAL] RC2 Packaged in Preparation for a Wednesday Release ...
SuSE 7..3 (2.4.10-4GB) Compiles and passes regression fine: All 89 tests passed. Installing to dev server next. Cheers, Steve On Monday 25 November 2002 8:19 am, you wrote: Morning all ... On Sunday this weekend, we packaged up PostgreSQL v7.3rc2 for testing ... this release, if all goes well, will become the Final Release on Wednesday, unless anyone comes up with any outstanding issues. At this point, we need as many ppl as possible to try and break it, so that when we do release, its as solid as we can possibly make it. If all goes well, v7.3 will be released by December 1st. Downloads are available at all mirrors, or the main site: ftp://ftp.postgresql.org/pub/beta Bugs should be reported to [EMAIL PROTECTED] Thanks ... ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
Re: [mail] Re: [HACKERS] Native Win32 sources
On Tue, 26 Nov 2002, Al Sutton wrote: D'Arcy, In production the database servers are seperate multi-processor machines with mirrored disks linked via Gigabit ethernet to the app server. In development I have people extremely familiar with MS, but not very hot with Unix in any flavour, who are developing Java and PHP code which is then passed into the QA phase where it's run on a replica of the production environment. My goal is to allow my developers to work on the platform they know (MS), using as many of the aspects of the production environment as possible (JVM version, PHP version, and hopefully database version), without needing to buy each new developer two machines, and incur the overhead of them familiarising themselves with a flavour of Unix. Hope this helps you understand where I'm comming from, I know it's not windows native but using Cygwin would at least get you out of the two boxes on everybody's desktop business. And for deveopers the difference in performance isn't all that great, as the only real performance issue is the one of creating / dropping backend connections is kinda slow. Since they'd be running on their own boxes for testing, you could probably just use persistant connections and get pretty good performance. What web server are they using? If it's apache, just set the number of max children down to something like 20 or so and they should be fine. ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
[HACKERS] Article on Linkers/Loaders
Here is a good article on linking/loading: http://www.linuxjournal.com/article.php?sid=6463 -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
[HACKERS] Boolean casting in 7.3 - changed?
A quick question: in 7.3 the following no longer works: template1= select 0::bool; ERROR: Cannot cast type integer to boolean The statement must be rewritten as this: template1= select '0'::bool; bool -- f (1 row) Is there a reason for this? I ask because the former query works in 7.1.3 and 7.2.1, but I haven't seen any mention of a change in 7.3 (at least not in the release notes). Apologies if this has been discussed to death previously, but it might be worth mentioning somewhere as a gotcha. Ian Barwick [EMAIL PROTECTED] ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] Request from eWeek for 7.3 comments
Hi Lisa, I'm the Lead Programmer for a company that specialises in online weight loss programs in Australia and the USA. We started off using MySQL but turned to PostgreSQL when it became obvious that MySQL lacked many necessary features required for a maximum power and data integrity. We have followed PostgreSQL through from 7.0 and will upgrade to 7.3 when it comes out. Each new version has added an incredible amount of new features. You only have to compare the release notes of Postgres against other databases to see how fast Postgres is being developed. Postgres has a reputation (deserved or not) of being hard to use. In 7.3 a lot of work has been done to 'fill in the gaps'. eg. In 7.3 you are able to drop columns from tables and change the NOT NULL status of a column easily. Foreign keys are smarter and are easier to drop and manipulate. On the other hand, PostgreSQL 7.3 will be great for academic environments. With the addition of SQL schemas, administrators can now create separate workspaces for all their users. Permissions have been greatly improved to match. Why pay for an Oracle license or get your students to use MS Access when PostgreSQL 7.3 will do the job? Lastly, Postgres 7.3 will be great for power users. With functions that can return record sets and prepared queries, more and more enterprises will see PostgreSQL as a viable backend for their operation. I haven't even scratched the surface of what's new in 7.3 and why everyone should be excited! As always, each new version of PostgreSQL is faster, more secure and more stable than ever before. That's why I love PostgreSQL! Chris - Original Message - From: Bruce Momjian [EMAIL PROTECTED] To: PostgreSQL-general [EMAIL PROTECTED]; PostgreSQL-development [EMAIL PROTECTED] Sent: Tuesday, November 26, 2002 3:20 PM Subject: [HACKERS] Request from eWeek for 7.3 comments I just spoke with Lisa Vaas from eWeek. She is writing an article on the upcoming PostgreSQL 7.3 release. (The release of 7.3 is scheduled for tomorrow.) She would like comments from users about the upcoming 7.3 features, listed at: http://developer.postgresql.org/docs/postgres/release.html#RELEASE-7-3 If you are interested, please reply to this email with any comments you might have. I have directed replies to her email address. She would like comments within the next few hours, until midnight EST. ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
[HACKERS] updating on views
Hi! I'd like to bother you with something about updating views. At present, postgresql7.3¡¡doesn't support update on views.When I look up the source code, I see that view is implemented with rules. That is when a view is created, a corresponding ¡®select'¡¡rule is created too. And I see the following code: #ifdef NOTYET RuleStmt *replace_rule; RuleStmt *append_rule; RuleStmt *delete_rule; #endif .. ... #ifdef NOTYET replace_rule = FormViewReplaceRule(view, viewParse); append_rule = FormViewAppendRule(view, viewParse); delete_rule = FormViewDeleteRule(view, viewParse); #endif .. ... In my eyes, it is not very difficult to realize view update with fill the above three functions¡£But I know the developers of postgresql are very learned, now that they didn't realize view update, maybe it is very difficult. I am a beginner of postgresql, so with my limited knowlege, maybe I can't realize the difficulity of doing such a thing. So I wonder if some of you would like to give me some advice about updating on views or why it is not realized¡¢what's the difficulty of doing it? Long for your reply. Thank you! Best regards! Yours XiaojingLi [EMAIL PROTECTED] 2002-11-23 ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [HACKERS] PostGres and WIN32, a plea!
Dave Page wrote: -Original Message- From: Merlin Moncure [mailto:[EMAIL PROTECTED]] Sent: 22 November 2002 21:25 To: [EMAIL PROTECTED] Subject: [HACKERS] PostGres and WIN32, a plea! I had read on one of the newsgroups that there is a planned native port to the win32 platform, is this true? I read most of the win32 thread off of the dev site and it was not clear if this was true. Hi Merlin, This is true - the port is being actively worked on (see recent posts from Bruce Momjian). A number of us have also been involved with the closed source beta testing recently. Yes, I expect the 7.4 release, due in mid-2003, to have a native Win32 port. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] Location of language .mo files or
Zdravo, Darko Prenosil writes: When I install, .mo file is copied to: /usr/local/pgsql/share/locale/hr_HR/LC_MESSAGES/postgres.mo (RedHat). In postgresql.conf is already line that looks like this: LC_MESSAGES = 'hr_HR'. So why I do not see the translated messages ? Hard to tell. Try making a PO file for a frontend application (such as psql or pg_dump), set the environment variables, and then see what happens. Possibly, there are less disturbing factors involved that way than on the server side. -- Peter Eisentraut [EMAIL PROTECTED] ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] elog(PANIC) should abort()?
Tom Lane writes: I am thinking it would be useful for debugging if elog(PANIC) were to exit by calling abort() so that a core dump would be produced. Going out via proc_exit(), as it now does, seems like a bad idea in any case, since that will try to do a bunch of cleanup activity that's probably inappropriate after a panic. But is this appropriate? PANIC: The database cluster was initialized with CATALOG_VERSION_NO 200210181, but the backend was compiled with CATALOG_VERSION_NO 200211021. It looks like you need to initdb. Aborted (core dumped) -- Peter Eisentraut [EMAIL PROTECTED] ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [PERFORM] [HACKERS] Realtime VACUUM, was: performance of insert/delete/update
Curtis Faith [EMAIL PROTECTED] writes: tom lane wrote: Sure, it's just shuffling the housekeeping work from one place to another. The thing that I like about Postgres' approach is that we put the housekeeping in a background task (VACUUM) rather than in the critical path of foreground transaction commit. Couldn't we reuse tuple and index space as soon as there are no transactions that depend on the old tuple or index values. I have imagined that this was always part of the long-term master plan. Couldn't we keep a list of dead tuples in shared memory and look in the list first when deciding where to place new values for inserts or updates so we don't have to rely on VACUUM (even a background one)? ISTM that either of these ideas would lead to pushing VACUUM overhead into the foreground transactions, which is exactly what we don't want to do. Keep in mind also that shared memory is finite ... *very* finite. It's bad enough trying to keep per-page status in there (cf FSM) --- per-tuple status is right out. I agree that automatic background VACUUMing would go a long way towards reducing operational problems. regards, tom lane ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [mail] Re: [HACKERS] Native Win32 sources
Al Sutton wrote: Lee, I wouldn't go for 7.4 in production until after it's gone gold, but being able to cut the number of boxes per developer by giving them a Win32 native version would save on everything from the overhead of getting the developers familiar enough with Linux to be able to admin their own systems, to cutting the network usage by having the DB and app on the same system, through to cutting the cost of electricity by only having one box per developer. It would also be a good way of testing 7.4 against our app so we can plan for an upgrade when it's released ;). I've tried open office 1.0.1 and had to ditch it. It had problems with font rendering and tables that ment many of the forms that people sent as word documents had chunks that weren't displayed or printed. We did try it on a box with MS-Word on it to ensure that the setup of the machine wasn't the issue, Word had no problems, OO failed horribly. www.peerdirect.com has a native PostgreSQL 7.2 release that should work fine for you until 7.4. It is in beta, I think. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] Boolean casting in 7.3 - changed?
Ian Barwick [EMAIL PROTECTED] writes: in 7.3 the following no longer works: template1= select 0::bool; ERROR: Cannot cast type integer to boolean Note that both old and new versions reject select 0::int4::bool; I believe the behavioral change is a consequence of Rod Taylor's DOMAIN patch: it essentially eliminated the old parser_typecast_constant() routine in order to ensure that constraints associated with a domain would get applied in examples like select 0::domaintypename. I wasn't totally happy with Rod's patch, for reasons that I couldn't put my finger on at the time, but perhaps my hindbrain understood that there would be noticeable behavioral changes. But be that as it may, the code is in there now and is unlikely to get reverted. There isn't any place in our docs that promises that you can coerce an integer-looking literal to bool --- and one could argue that allowing such is just opening the door for typos. regards, tom lane ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
Re: Antw: Re: [HACKERS] Native Win32 sources
I am told by PeerDirect that they will release the Win32 source as a patch against current CVS by the end of December. At that point, we will make adjustments then apply the patch and start making any other changes required. I don't think there is much we can do until they supply that patch. I thought about starting work on it but most were willing to wait for a 100% functional patch. I am CC'ing Katie Ward [EMAIL PROTECTED], our PeerDirect contact on this, and Jan, who also works for them. --- Ulrich Neumann wrote: Bruce, you're right, i'm asking about the sources because I want to help. Is it possible to help in this case or not? Ulrich Bruce Momjian [EMAIL PROTECTED] 25.11.2002 18:51:39 Ulrich Neumann wrote: Hello, i've read that there are 2 different native ports for Windows somewhere. I've searched for them but didn't found them. Is there anyone who can point me to a link or send me a copy of the sources? Oh, you are probably asking about the sources. They are not publically available yet. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED] -- This e-mail is virus scanned Diese e-mail ist virusgeprueft ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] updating on views
On Tue, 26 Nov 2002, Stephan Szabo wrote: Well, it depends. SQL updatable views are fairly limited from what I remember and the general write your own update rules is much broader, so I doubt anyone got terribly excited by doing the limited version. It would take some work to define what sort of view queries are acceptable for what sort of updates and then make the appropriate rules (imagine queries with SQL99 says that all columns in the view definition (ie, the SQL query which defines the view) must also be updateable. This, of course, requires some checking. We could do this here. I think, however, that it would be much cleaner to implement this correctly through the planner and executor instead of hacking it through the rewriter. Of course, that means lots of code. Insertable-into and updateable views are certainly a very important feature which Postgres is lacking. Maybe we should implement this the easy way first and then, to increase performance, correctly -- at some later point. Gavin ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] Auto Vacuum Daemon (again...)
On 26 Nov 2002 at 21:54, Matthew T. O'Connor wrote: First: Do we want AVD integrated into the main source tree, or should it remain a tool that can be downloaded from gborg. I would think it should be controlled by the postmaster, and configured from GUC (at least basic on off settings) Since you have rewritten in C, I think it can be safely added to contrib, after core team agrees. It is a good place for such things. Second: Assuming we want it integrated into the source tree, can it remain a client app? Can a non backend program that connects to the postmaster using libpq be a child of the postmaster that the postmaster can control (start and stop). I would not like postmaster forking into pgavd app. As far as possible, we should not touch the core. This is a client app. and be it that way. Once we integrate it into backend, we need to test the integration as well. Why bother? Anyway for you reading pleasure, I have attached a plot of results from a simple test program I wrote. As you can see from the plot, AVD keeps the file size under control. Also, the first few Xacts are faster in the non AVD case, but after that AVD keeps the average Xact time down. The periodic spikes in the AVD run correspond to when the AVD has fired off a vacuum. Also when the table file gets to approx 450MB performance drops off horribly I assume this is because my system can no longer cache the whole file (I have 512M in my machine). Also, I had been developing against 7.2.3 until recently, and I wound up doing some of these benchmarks against both 7.2.3 and 7.3devel and 7.3 perfoms much better, that is it 7.2 slowed down much sooner under this test. Good to know that it works. I would like to comment w.r.t to my original effort. 1) I intentionally left vacuum full to admin. Disk space is cheap and we all know that but IMO no application should lock a table without admin knowing it. This is kinda microsoftish assumption of user friendliness to make decision on behalf of users. Of course, sending admin a notigication is a good idea.. 2)In a cluster if there are many databases and time taken for serial vacuum is more than time gap between two wake-up intervals of AVD, it would get into a continous vacuum. At some point of time, we are going to need one connection per database in separate process/thread. Thanks for your work.. Bye Shridhar -- Distinctive, adj.: A different color or shape than our competitors. ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [PERFORM] [HACKERS] Realtime VACUUM, was: performance of insert/delete/update
Good ideas. I think the master solution is to hook the statistics daemon information into an automatic vacuum that could _know_ which tables need attention. --- Curtis Faith wrote: tom lane wrote: Sure, it's just shuffling the housekeeping work from one place to another. The thing that I like about Postgres' approach is that we put the housekeeping in a background task (VACUUM) rather than in the critical path of foreground transaction commit. Thinking with my marketing hat on, MVCC would be a much bigger win if VACUUM was not required (or was done automagically). The need for periodic VACUUM just gives ammunition to the PostgreSQL opponents who can claim we are deferring work but that it amounts to the same thing. A fully automatic background VACUUM will significantly reduce but will not eliminate this perceived weakness. However, it always seemed to me there should be some way to reuse the space more dynamically and quickly than a background VACUUM thereby reducing the percentage of tuples that are expired in heavy update cases. If only a very tiny number of tuples on the disk are expired this will reduce the aggregate performance/space penalty of MVCC into insignificance for the majority of uses. Couldn't we reuse tuple and index space as soon as there are no transactions that depend on the old tuple or index values. I have imagined that this was always part of the long-term master plan. Couldn't we keep a list of dead tuples in shared memory and look in the list first when deciding where to place new values for inserts or updates so we don't have to rely on VACUUM (even a background one)? If there are expired tuple slots in the list these would be used before allocating a new slot from the tuple heap. The only issue is determining the lowest transaction ID for in-process transactions which seems relatively easy to do (if it's not already done somewhere). In the normal shutdown and startup case, a tuple VACUUM could be performed automatically. This would normally be very fast since there would not be many tuples in the list. Index slots would be handled differently since these cannot be substituted one for another. However, these could be recovered as part of every index page update. Pages would be scanned before being written and any expired slots that had transaction ID's lower than the lowest active slot would be removed. This could be done for non-leaf pages as well and would result in only reorganizing a page that is already going to be written thereby not adding much to the overall work. I don't think that internal pages that contain pointers to values in nodes further down the tree that are no longer in the leaf nodes because of this partial expired entry elimination will cause a problem since searches and scans will still work fine. Does VACUUM do something that could not be handled in this realtime manner? - Curtis ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
Re: [HACKERS] Boolean casting in 7.3 - changed?
On Wednesday 27 November 2002 06:23, Tom Lane wrote: (B Ian Barwick [EMAIL PROTECTED] writes: (B in 7.3 the following no longer works: (Btemplate1= select 0::bool; (BERROR: Cannot cast type integer to boolean (B (B Note that both old and new versions reject (B select 0::int4::bool; (B (B I believe the behavioral change is a consequence of Rod Taylor's (B DOMAIN patch: it essentially eliminated the old parser_typecast_constant() (B routine in order to ensure that constraints associated with a domain (B would get applied in examples like "select 0::domaintypename". (B (B I wasn't totally happy with Rod's patch, for reasons that I couldn't put (B my finger on at the time, but perhaps my hindbrain understood that there (B would be noticeable behavioral changes. But be that as it may, the code (B is in there now and is unlikely to get reverted. There isn't any place (B in our docs that promises that you can coerce an integer-looking literal (B to bool --- and one could argue that allowing such is just opening the (B door for typos. (B (BThanks for the explanation. I'm not screaming for a reversion ;-), but (Bchanging behaviour which was implicitly valid in previous (Bversions is bound to cause a few people a little head scratching (Bwhen converting applications to 7.3 (I'm sure I can't be the only one). (B (BHow about a line in HISTORY under "Migration to version 7.3" along (Bthe lines of: (B (B"Casting integers to boolean (for example, 0::bool) is no longer allowed, (Buse '0'::bool instead". (B (B (BIan Barwick ([EMAIL PROTECTED] (B (B (B---(end of broadcast)--- (BTIP 5: Have you checked our extensive FAQ? (B (Bhttp://www.postgresql.org/users-lounge/docs/faq.html
Re: [HACKERS] Location of language .mo files or 'Zato postgres ne govori Hrvatski' ???
On Tuesday 26 November 2002 18:42, Peter Eisentraut wrote: Zdravo, Darko Prenosil writes: When I install, .mo file is copied to: /usr/local/pgsql/share/locale/hr_HR/LC_MESSAGES/postgres.mo (RedHat). In postgresql.conf is already line that looks like this: LC_MESSAGES = 'hr_HR'. So why I do not see the translated messages ? Hard to tell. Try making a PO file for a frontend application (such as psql or pg_dump), set the environment variables, and then see what happens. Possibly, there are less disturbing factors involved that way than on the server side. I find out what was wrong. I did not make clean, and postgres backend was already compiled and linked without --enable-nls. So make only maked my hr_HR.mo, but did not recompiled the backend. After make clean and make install evetithing is working just fine. Thank You anyway for Your effort !!! Zdravo i tebi ! ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [mail] Re: [HACKERS] Native Win32 sources
Al Sutton kirjutas T, 26.11.2002 kell 20:37: D'Arcy, In production the database servers are seperate multi-processor machines with mirrored disks linked via Gigabit ethernet to the app server. In development I have people extremely familiar with MS, but not very hot with Unix in any flavour, who are developing Java and PHP code which is then passed into the QA phase where it's run on a replica of the production environment. My goal is to allow my developers to work on the platform they know (MS), using as many of the aspects of the production environment as possible (JVM version, PHP version, and hopefully database version), without needing to buy each new developer two machines, and incur the overhead of them familiarising themselves with a flavour of Unix. You could try out VMWare and run a linux virtual machine under Windows, You could set it up once with all necessary servers and then copy the files to each new developers machine. VMWare is not free, but should be significantly cheaper than buying a whole computer. - Hannu ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [mail] Re: [HACKERS] Native Win32 sources
D'Arcy, In production the database servers are seperate multi-processor machines with mirrored disks linked via Gigabit ethernet to the app server. In development I have people extremely familiar with MS, but not very hot with Unix in any flavour, who are developing Java and PHP code which is then passed into the QA phase where it's run on a replica of the production environment. My goal is to allow my developers to work on the platform they know (MS), using as many of the aspects of the production environment as possible (JVM version, PHP version, and hopefully database version), without needing to buy each new developer two machines, and incur the overhead of them familiarising themselves with a flavour of Unix. (from experience in a large .com web site) Can you have a central DB server? Do all the dev DB servers need to be independent? You could even have a machine w/ ip*(# developers) and bind a postgresql to each ip for each developer (assuming you had enough memory, etc). We used oracle once upon a time at my .com and used seperate schemas for the seperate developers. This may be tricky for your environment because the developers would need to know what schema they would connect to if all schemas were under the same pgsql instance. - Brandon c: 917-697-8665h: 201-798-4983 b. palmer, [EMAIL PROTECTED] pgp:crimelabs.net/bpalmer.pgp5 ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [mail] Re: [HACKERS] Native Win32 sources
On 27 Nov 2002, Hannu Krosing wrote: Al Sutton kirjutas T, 26.11.2002 kell 20:37: D'Arcy, In production the database servers are seperate multi-processor machines with mirrored disks linked via Gigabit ethernet to the app server. In development I have people extremely familiar with MS, but not very hot with Unix in any flavour, who are developing Java and PHP code which is then passed into the QA phase where it's run on a replica of the production environment. My goal is to allow my developers to work on the platform they know (MS), using as many of the aspects of the production environment as possible (JVM version, PHP version, and hopefully database version), without needing to buy each new developer two machines, and incur the overhead of them familiarising themselves with a flavour of Unix. You could try out VMWare and run a linux virtual machine under Windows, You could set it up once with all necessary servers and then copy the files to each new developers machine. VMWare is not free, but should be significantly cheaper than buying a whole computer. If you're gonna go that far, look at reversing that situation, i.e. run a linux box for each person with windows in vmware. It's a much more stable situation than the other way around. Either way, you can then run multiple Windows instances, of different versions of windows if need be, which means you can test and develop for multiple windows environments on one box, no rebooting, not even having to turn your chair around. VMWare likes memory, so get plenty if you go that way. And don't worry about the problems getting familiar with most newer flavors of linux, they're pretty easy to grok for most developers. P.S. a note on windows and vmware: It's not uncommon for companies now to build a large linux box, put vmware gsx on it, and run dozens of windows instances. That way the spare cycles for one server can be used by another, you can consolidate your windows servers onto a couple of boxen, and you get much more reliable operation from windows when the hardware is abstracted away from underneath it. ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] Hirarchical queries a la Oracle. Patch.
Evgen Potemkin kirjutas R, 22.11.2002 kell 15:57:
Hi there!
Patch is posted to pgsql-patches. docs inside.
It would of course be nice to support both Oracle and ISO/ANSI syntaxes,
but I'm afraid that the (+) may clash with our overloadable operators
feature.
SQL 99 version will be later.
I attach a railroad diagram of SQL99 WITH RECURSIVE and a diff against
mid-summer gram.y which implements half of SQL99 _syntax_ (just the WITH
{RECURSIVE} part, SEARCH (tree search order order) and CYCLE (recursion
control) clauses are missing).
WITH clause seems to be quite useful in its own right as well, not just
for recursive queries, so I guess that someone with good knwledge of
postgresql internals could get plain WITH working quite fast -
The main difference between subqueries defined in WITH clause and in
FROM clause is that while subqueries in FROM don't see each other in
their namespaces, the ones in WITH either see all preceeding ones (plain
with) or just all in WITH clause (WITH RECURSIVE)
--
Hannu
attachment: with_clause.gif276a277,285
/* WITH CLAUSE */
%type boolean opt_recursive
%type node with_list_element
%type list with_list
/* /WITH CLAUSE */
374c383,387
READ, REAL, REFERENCES, REINDEX, RELATIVE, RENAME, REPLACE,
---
READ, REAL,
RECURSIVE,
REFERENCES, REINDEX, RELATIVE, RENAME, REPLACE,
4023,4024c4036,4037
SelectStmt: select_no_parens %prec UMINUS
| select_with_parens%prec UMINUS
---
SelectStmt: select_no_parens%prec UMINUS
| select_with_parens%prec UMINUS
4028,4029c4041,4042
'(' select_no_parens ')'{ $$ =
$2; }
| '(' select_with_parens ')'{ $$ = $2; }
---
'(' select_no_parens ')'
{ $$ = $2; }
| '(' select_with_parens ')'
{ $$ = $2; }
4031a4045
4033c4047,4058
simple_select { $$ =
$1; }
---
simple_select
{ $$ = $1; }
| WITH opt_recursive with_list simple_select
{
/* this should actually blend in subselects
from WITH
* just replacing will do the WRONG THING
*/
((SelectStmt *) $4 )-with_recursive = $2;
((SelectStmt *) $4 )-withClause = $3;
$$ = $4;
}
4039a4065,4072
/*
| WITH RECURSIVE with_list select_clause sort_clause
opt_for_update_clause opt_select_limit
{
insertSelectOptions((SelectStmt *) $2, $3, $4,
nth(0,
$5), nth(1, $5));
$$ = $1;
}
*/
4045a4079,4087
/*
| WITH RECURSIVE with_list select_clause for_update_clause
opt_select_limit
{
insertSelectOptions((SelectStmt *) $2, NIL, $3,
nth(0,
$4), nth(1, $4));
$$ = $1;
}
*/
4051a4094,4100
| WITH opt_recursive with_list select_clause select_limit
{
insertSelectOptions((SelectStmt *) $4, NIL,
NIL,
nth(0,
$5), nth(1, $5));
$$ = $4;
}
4053a4103,4141
/* WITH CLAUSE
*
* ANSI/ISO SQL99 p 7.13-7.14
*
*/
opt_recursive:
RECURSIVE { $$ = TRUE; }
| /*EMPTY */{ $$ = FALSE; }
;
with_list:
with_list_element { $$ =
makeList1($1); }
| with_list ',' with_list_element { $$ = lappend($1, $3); }
;
with_list_element:
ColId '(' name_list ')' AS select_with_parens
{
RangeSubselect *n = makeNode(RangeSubselect);
n-subquery = $6;
n-alias =
Re: [HACKERS] Why an array in pg_group?
Reinoud van Leeuwen kirjutas K, 20.11.2002 kell 17:03: Hi, Is there any reason why the grolist field in the table pg_group is implemented as an array and not as a separate table? most likely for performance reasons. According to the documentation: quote source=Postgresql 7.2 User Manual, chapter 6 near the end Arrays are not sets; using arrays in the manner described in the previous paragraph is often a sign of database misdesign. /quote I have trouble implementing a way to easily check whether a user is part of a group. (I use Apache::AuthDBI to implement authentication and wanted to make a view with columns username, userid , groupname. And installing the contrib/array give's me a postgresql that is different from all the others :-( not from those who also have installed contrib/array ;) but you should actually be using contrib/intagg (and perhaps contrib intarray) for performance reasons ;) -- Hannu Krosing [EMAIL PROTECTED] ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [mail] Re: [HACKERS] Native Win32 sources
scott.marlowe kirjutas K, 27.11.2002 kell 01:40: On 27 Nov 2002, Hannu Krosing wrote: You could try out VMWare and run a linux virtual machine under Windows, You could set it up once with all necessary servers and then copy the files to each new developers machine. VMWare is not free, but should be significantly cheaper than buying a whole computer. If you're gonna go that far, look at reversing that situation, i.e. run a linux box for each person with windows in vmware. It's a much more stable situation than the other way around. That's how I use it. It's also nice way to try out new win software - install it, check it out and if you don't like it just say no to save changes? when closing the vmware session ;) Either way, you can then run multiple Windows instances, of different versions of windows if need be, which means you can test and develop for multiple windows environments on one box, no rebooting, not even having to turn your chair around. VMWare likes memory, so get plenty if you go that way. And don't worry about the problems getting familiar with most newer flavors of linux, they're pretty easy to grok for most developers. P.S. a note on windows and vmware: It's not uncommon for companies now to build a large linux box, put vmware gsx on it, and run dozens of windows instances. That way the spare cycles for one server can be used by another, you can consolidate your windows servers onto a couple of boxen, and you get much more reliable operation from windows when the hardware is abstracted away from underneath it. I guess this would be good for win _servers_, but how would you use this setup for developers - will they all sit around a single box ? --- Hannu ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] possible obvious bug?
On 27 Nov 2002, Hannu Krosing wrote: Merlin Moncure kirjutas T, 26.11.2002 kell 08:00: I was playing with the Japanese win32 7.2.1 port and I noticed that select 0 / 0 caused the server to crash and restart. I understand that it is a totally unsupported version, but it should be easy enough to check vs. the current version. Note that select 0.0/0.0 worked fine! So what is the right answer ? Maybe it's a locale oriented thing? ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [mail] Re: [HACKERS] Native Win32 sources
On 27 Nov 2002, Hannu Krosing wrote: scott.marlowe kirjutas K, 27.11.2002 kell 01:40: On 27 Nov 2002, Hannu Krosing wrote: You could try out VMWare and run a linux virtual machine under Windows, You could set it up once with all necessary servers and then copy the files to each new developers machine. VMWare is not free, but should be significantly cheaper than buying a whole computer. If you're gonna go that far, look at reversing that situation, i.e. run a linux box for each person with windows in vmware. It's a much more stable situation than the other way around. That's how I use it. It's also nice way to try out new win software - install it, check it out and if you don't like it just say no to save changes? when closing the vmware session ;) Plus, it's real easy to back up your windows servers. just shut them down, backup their image, and start them back up. P.S. a note on windows and vmware: It's not uncommon for companies now to build a large linux box, put vmware gsx on it, and run dozens of windows instances. That way the spare cycles for one server can be used by another, you can consolidate your windows servers onto a couple of boxen, and you get much more reliable operation from windows when the hardware is abstracted away from underneath it. I guess this would be good for win _servers_, but how would you use this setup for developers - will they all sit around a single box ? You could probably use xwindows remote sessions for something like that, but yeah, I was strictly thinking servers at that point. :-) There is some work being done to put mutiple video cards and keyboard/mice onto a single large box and share it though. I don't think I like taking sharing quite that far though. :-0 ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] possible obvious bug?
Merlin Moncure kirjutas T, 26.11.2002 kell 08:00: I was playing with the Japanese win32 7.2.1 port and I noticed that select 0 / 0 caused the server to crash and restart. I understand that it is a totally unsupported version, but it should be easy enough to check vs. the current version. Note that select 0.0/0.0 worked fine! So what is the right answer ? NaN :) Chris ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] possible obvious bug?
Merlin Moncure kirjutas T, 26.11.2002 kell 08:00: I was playing with the Japanese win32 7.2.1 port and I noticed that select 0 / 0 caused the server to crash and restart. I understand that it is a totally unsupported version, but it should be easy enough to check vs. the current version. Note that select 0.0/0.0 worked fine! So what is the right answer ? Maybe it's a locale oriented thing? In 7.2.3 there seem to be two different messages: usa=# select 0/0; ERROR: floating point exception! The last floating point operation either exceeded legal ranges or was a divide by zero usa=# select 0/0.0; ERROR: float8div: divide by zero error usa=# select 0.0/0.0; ERROR: float8div: divide by zero error usa=# select 0.0/0; ERROR: float8div: divide by zero error usa=# select 1/0; ERROR: floating point exception! The last floating point operation either exceeded legal ranges or was a divide by zero usa=# select 1/0.0; ERROR: float8div: divide by zero error Chris ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/users-lounge/docs/faq.html
[HACKERS] Request from eWeek for 7.3 comments
I just spoke with Lisa Vaas from eWeek. She is writing an article on the upcoming PostgreSQL 7.3 release. (The release of 7.3 is scheduled for tomorrow.) She would like comments from users about the upcoming 7.3 features, listed at: http://developer.postgresql.org/docs/postgres/release.html#RELEASE-7-3 If you are interested, please reply to this email with any comments you might have. I have directed replies to her email address. She would like comments within the next few hours, until midnight EST. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] error codes
Insisting on Andreas suggestion, why can't we just prefix all error message strings with the SQLState code? So all error messages would have the format CCSSS - Where CCSSS is the standard SQLState code and the message text is a more specific description. Note that the standard allows for implementation-defined codes, so we can have our own CC classes and all the SSS subclasses that we need. -- Fernando Nasser Red Hat - Toronto E-Mail: [EMAIL PROTECTED] 2323 Yonge Street, Suite #300 Toronto, Ontario M4P 2C9 ---(end of broadcast)--- TIP 6: Have you searched our list archives? http://archives.postgresql.org
[HACKERS] contrib/ltree patches
I have been looking at contrib/ltree in the PostgreSQL repository. I've
modified the code to allow / as a node delimiter instead of . which is the
default.
Below are the patches to make this change. I have also moved the
delimiter to a DEFINE so that other customizations are easily done. This
is a work in progress.
My thanks to DarbyD for assistance.
cheers
--- ltree.h.origTue Nov 26 18:57:58 2002
+++ ltree.h Tue Nov 26 20:16:40 2002
@@ -6,6 +6,8 @@
#include utils/palloc.h
#include utils/builtins.h
+#defineNODE_DELIMITER '/'
+
typedef struct
{
uint8 len;
@@ -88,7 +90,7 @@
#ifndef abs
#define abs(a) ((a) (0) ? -(a) : (a))
#endif
-#define ISALNUM(x) ( isalnum((unsigned int)(x)) || (x) == '_' )
+#define ISALNUM(x) ( isalnum((unsigned int)(x)) || (x) == '_' || (x) ==
+NODE_DELIMITER )
/* full text query */
--- ltree_io.c Tue Nov 26 20:23:45 2002
+++ ltree_io.c.orig Tue Nov 26 18:57:26 2002
@@ -48,7 +48,7 @@
ptr = buf;
while (*ptr)
{
- if (*ptr == NODE_DELIMITER)
+ if (*ptr == '.')
num++;
ptr++;
}
@@ -69,7 +69,7 @@
}
else if (state == LTPRS_WAITDELIM)
{
- if (*ptr == NODE_DELIMITER)
+ if (*ptr == '.')
{
lptr-len = ptr - lptr-start;
if (lptr-len 255)
@@ -131,7 +131,7 @@
{
if (i != 0)
{
- *ptr = NODE_DELIMITER;
+ *ptr = '.';
ptr++;
}
memcpy(ptr, curlevel-name, curlevel-len);
@@ -181,7 +181,7 @@
ptr = buf;
while (*ptr)
{
- if (*ptr == NODE_DELIMITER)
+ if (*ptr == '.')
num++;
else if (*ptr == '|')
numOR++;
@@ -265,7 +265,7 @@
lptr-len, (int) (lptr-start - buf));
state = LQPRS_WAITVAR;
}
- else if (*ptr == NODE_DELIMITER)
+ else if (*ptr == '.')
{
lptr-len = ptr - lptr-start -
((lptr-flag LVAR_SUBLEXEM) ? 1 : 0) -
@@ -289,7 +289,7 @@
{
if (*ptr == '{')
state = LQPRS_WAITFNUM;
- else if (*ptr == NODE_DELIMITER)
+ else if (*ptr == '.')
{
curqlevel-low = 0;
curqlevel-high = 0x;
@@ -347,7 +347,7 @@
}
else if (state == LQPRS_WAITEND)
{
- if (*ptr == NODE_DELIMITER)
+ if (*ptr == '.')
{
state = LQPRS_WAITLEVEL;
curqlevel = NEXTLEV(curqlevel);
@@ -471,7 +471,7 @@
{
if (i != 0)
{
- *ptr = NODE_DELIMITER;
+ *ptr = '.';
ptr++;
}
if (curqlevel-numvar)
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster
Re: [HACKERS] Why an array in pg_group?
On Tuesday 26 November 2002 09:05 pm, Hannu Krosing wrote: Reinoud van Leeuwen kirjutas K, 20.11.2002 kell 17:03: Hi, Is there any reason why the grolist field in the table pg_group is implemented as an array and not as a separate table? most likely for performance reasons. According to the documentation: quote source=Postgresql 7.2 User Manual, chapter 6 near the end Arrays are not sets; using arrays in the manner described in the previous paragraph is often a sign of database misdesign. /quote I have trouble implementing a way to easily check whether a user is part of a group. (I use Apache::AuthDBI to implement authentication and wanted to make a view with columns username, userid , groupname. And installing the contrib/array give's me a postgresql that is different from all the others :-( not from those who also have installed contrib/array ;) but you should actually be using contrib/intagg (and perhaps contrib intarray) for performance reasons ;) Can You make syntax (operator or function) like : scalar integer IN array integers for join scalar and array fileds ? in base PostgreSQL ? regards Haris Peco ---(end of broadcast)--- TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
Re: [HACKERS] elog(PANIC) should abort()?
Peter Eisentraut [EMAIL PROTECTED] writes: Tom Lane writes: I am thinking it would be useful for debugging if elog(PANIC) were to exit by calling abort() so that a core dump would be produced. But is this appropriate? PANIC: The database cluster was initialized with CATALOG_VERSION_NO 200210181, but the backend was compiled with CATALOG_VERSION_NO 200211021. It looks like you need to initdb. Aborted (core dumped) Hm. We could possibly reduce those particular messages to FATAL. OTOH, it's not unreasonable that seeing those messages *in the field* might be an appropriate situation for a core dump. I think as developers we sometimes have a skewed sense of what's common ;-) Ever since Bruce introduced the additional elog levels, I have felt it would be a good idea to go through all the elog calls and re-evaluate what levels they should have. It's a lot o' work though... regards, tom lane ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
Re: [HACKERS] Request from eWeek for 7.3 comments
I am the maintainer and lead developer for Bricolage, an enterprise-class open-source content management system built on a PostgreSQL database. The Bricolage developers really look forward to 7.3's new features, which, like every release of PostgreSQL, set a new standard against which other databases measure themselves. Particularly important for Bricolage are the ability to drop columns and the new support for prepared SQL statements. Every major release of Bricolage requires changes to the database, often including the removal or change of a table column. Now that PostgreSQL can drop columns, future Bricolage upgrades can change database columns without leaving deprecated columns in the database. Furthermore, Bricolage runs in an Apache/mod_perl environment where many of the same database queries are executed many times over. The new support for prepared statements in PostgreSQL 7.3 will greatly enhance performance by reducing the number of times each of those SQL statements is prepared by PostgreSQL to once per Apache process -- for the lifetime of the process. These enhancements in PostgreSQL are great because they'll improve not just the database, but all applications that are built upon it. The speed with which PostgreSQL continues to develop and provide trickle-down benefits to the applications that depend upon it is simply second-to-none. Regards, David PS: eWeek has covered Bricolage here: http://www.eweek.com/article2/0,3959,652977,00.asp -- David Wheeler AIM: dwTheory [EMAIL PROTECTED] ICQ: 15726394 http://david.wheeler.net/ Yahoo!: dew7e Jabber: [EMAIL PROTECTED] On Tuesday, November 26, 2002, at 03:20 PM, Bruce Momjian wrote: I just spoke with Lisa Vaas from eWeek. She is writing an article on the upcoming PostgreSQL 7.3 release. (The release of 7.3 is scheduled for tomorrow.) She would like comments from users about the upcoming 7.3 features, listed at: http://developer.postgresql.org/docs/postgres/release.html#RELEASE-7-3 If you are interested, please reply to this email with any comments you might have. I have directed replies to her email address. She would like comments within the next few hours, until midnight EST. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED]) ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
[HACKERS] Auto Vacuum Daemon (again...)
Several months ago tried to implement a special postgres backend as an Auto Vacuum Daemon (AVD), somewhat like the stats collector. I failed due to my lack of experience with the postgres source. On Sep 23, Shridhar Daithankar released an AVD written in C++ that acted as a client program rather than part of the backend. I rewrote it in C, and have been playing with it ever since. At this point I need feedback and direction from the hacker group. First: Do we want AVD integrated into the main source tree, or should it remain a tool that can be downloaded from gborg. I would think it should be controlled by the postmaster, and configured from GUC (at least basic on off settings) Second: Assuming we want it integrated into the source tree, can it remain a client app? Can a non backend program that connects to the postmaster using libpq be a child of the postmaster that the postmaster can control (start and stop). Third: If a special backend version is preferred, I don't personally know how to have a backend monitor and vacuum multiple databases. I guess it could be similar to the client app and fire up new back everytime a database needs to be vacuumed. Fourth: I think AVD is a feature that is needed in some form or fashion. I am willing to work on it, but if it needs to be a backend version I will probably need some help. Anyway for you reading pleasure, I have attached a plot of results from a simple test program I wrote. As you can see from the plot, AVD keeps the file size under control. Also, the first few Xacts are faster in the non AVD case, but after that AVD keeps the average Xact time down. The periodic spikes in the AVD run correspond to when the AVD has fired off a vacuum. Also when the table file gets to approx 450MB performance drops off horribly I assume this is because my system can no longer cache the whole file (I have 512M in my machine). Also, I had been developing against 7.2.3 until recently, and I wound up doing some of these benchmarks against both 7.2.3 and 7.3devel and 7.3 perfoms much better, that is it 7.2 slowed down much sooner under this test. Thanks, Matthew ps, The test program performs the following: create table pgavdtest_table (id int,num numeric(10,2),txt char(512)) while i1000 insert into pgavdtest_table (id,num,txt) values (i,i.i,'string i') while i1000 update pgavdtest_table set num=num+i, txt='update string %i' pps, I can post the source (both the AVD and the test progam) to the list, or email it to individuals if they would like. attachment: avdtest.png ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
[HACKERS] Interface update for 7.3
I have worked on several gborg projects migrated from our CVS tree, and all but one have new packaged releases, ready for 7.3: libpq++ pgeasy pgperl I am working with David Wheeler on DBD:pg and hope to have a release packaged up tomorrow. -- Bruce Momjian| http://candle.pha.pa.us [EMAIL PROTECTED] | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup.| Newtown Square, Pennsylvania 19073 ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send unregister YourEmailAddressHere to [EMAIL PROTECTED])
