FAQ

PostgreSQL > pgsql-hackers > August 1997

Search Discussions

161 discussions - 404 posts

  • I have been thinking about doing a universal source code update to make it more consistent. The options I have considered are (in order of severity): run a program to convert spaces to tabs convert ...
    Bruce MomjianBruce Momjian
    Aug 17, 1997 at 3:28 pm
    Aug 20, 1997 at 11:02 pm
  • I just added in Adrian's JDBC driver so that its part of the source tree... Adrian, you might want to add some sort of README to this as to how exactly it should be installed/used? I saw the ...
    The Hermit HackerThe Hermit Hacker
    Aug 16, 1997 at 8:51 pm
    Aug 19, 1997 at 5:46 am
  • Hi, Robson Paniago de Miranda wrote: ^^^^^^^^^^^^^^^^ What is the state of this project, Robson ? I've been asked for this feature by my friend and I see that I can implement this in 3-4 hours... We ...
    Vadim B. MikheevVadim B. Mikheev
    Aug 14, 1997 at 5:13 am
    Aug 19, 1997 at 4:35 am
  • Hi! 1. CHECK/DEFAULT implemented and I just added regression tests for them. NOTE that I use some tricks to get unparsed definition of them. I save unparsed definition in catalog to simplify pg_dump. ...
    Vadim B. MikheevVadim B. Mikheev
    Aug 28, 1997 at 7:47 am
    Aug 30, 1997 at 7:44 am
  • Hi Bruce, (and all), main.c splits off to either PostgresMain or PostmasterMain *really* early in the execution path... Since you seem to indicate that the execv() doesn't cost much (any measurable?) ...
    Wayde NieWayde Nie
    Aug 8, 1997 at 2:30 am
    Aug 11, 1997 at 12:45 pm
  • What functionality will this provide, exactly? If I'm understanding correctly, it will mean the ability to see how many records were changed on something like an 'update'? Marc G. Fournier ...
    The Hermit HackerThe Hermit Hacker
    Aug 26, 1997 at 10:08 am
    Aug 27, 1997 at 2:04 am
  • With Vadim busy and Thomas on vacation until the end of the month, Marc agreed to a new release to get the four new ports added, and the new psort code in use. I believe the source is ready for a ...
    Bruce MomjianBruce Momjian
    Aug 12, 1997 at 10:57 pm
    Aug 14, 1997 at 4:19 am
  • I believe the new create synax is: CREATE TABLE test (x NULL, y NOT NULL); Are we sure the first example is SQL standard. I have always see cases where NOT NULL is specified, and if it is not ...
    Bruce MomjianBruce Momjian
    Aug 21, 1997 at 2:45 pm
    Aug 22, 1997 at 4:29 pm
  • The variance/std. deviation requires two passes through the data; one to calculate an average and one to calculate the sum of the differences from the average. The postgres aggregate support ...
    Thomas G. LockhartThomas G. Lockhart
    Aug 29, 1997 at 1:05 am
    Aug 29, 1997 at 12:55 pm
  • I noticed a few of these go by in committers: Log Message: Get rid of attproc, atttyparg, attcanindex, attdefrel, attbound from pg_attribute. atthasdef added. Sice we are changing the system tables ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Aug 21, 1997 at 1:52 am
    Aug 21, 1997 at 1:39 pm
  • I have applied my patch to make hundreds of functions static, and others wrapped in #ifdef NOT_USED. Hope this helps. - -- Bruce Momjian maillist@candle.pha.pa.us ------------------------------
    Bruce MomjianBruce Momjian
    Aug 19, 1997 at 9:47 pm
    Aug 21, 1997 at 1:14 am
  • Somehow, and error has cropt up when doing an UPDATE. Any idea what is causing it? #0 PortalHeapMemoryFree (this=0x28b12c, pointer=0x53ab10 "f") at portalmem.c:264 #1 0xbe19e in MemoryContextFree ...
    Bruce MomjianBruce Momjian
    Aug 17, 1997 at 9:02 pm
    Aug 18, 1997 at 6:48 pm
  • Hi all, In backend/storage/ipc/s_lock.c, the OSF alpha definition tries # if defined(__alpha__) || defined(__alpha) That line breaks linuxalpha (coz there, __alpha__ is defined). But, I think that ...
    IntegrationIntegration
    Aug 22, 1997 at 4:27 am
    Aug 25, 1997 at 4:08 am
  • A gather that pg_dump still dumps (SCNR ;-) ) core if one tries to dump sequences? Maybe we should document that somewhere. Regards, Hakan - -- Hakan Tandogan hakan@iconsult.com ICONSULT Tandogan - ...
    Hakan TandoganHakan Tandogan
    Aug 13, 1997 at 8:59 pm
    Aug 14, 1997 at 6:20 pm
  • I have to say, I am feeling very positive about PostgreSQL. In the beginning, because so many of us were unknowledgable about the source at the time, I feared that too many people would throw things ...
    Bruce MomjianBruce Momjian
    Aug 29, 1997 at 2:40 am
    Aug 31, 1997 at 10:15 pm
  • Hi Vadim, I'm not sure whether this is a logical thing towant to do but if I try to specify a default that is *not* NOT NULL then I get the following error. postgres= create table test7 ( id1 int4 ...
    Keith ParksKeith Parks
    Aug 28, 1997 at 9:10 pm
    Aug 30, 1997 at 9:38 am
  • A few days ago, I sent a large patch for the JDBC driver to pgsql-patches@postgresql.org. This patch effectively merged the previous two JDBC drivers. Its just that I haven't seen it appear yet when ...
    PostrgreSQL PatchesPostrgreSQL Patches
    Aug 29, 1997 at 4:36 am
    Aug 30, 1997 at 6:38 am
  • In 5 days, I'm disbanding the supserver, which means that everyone has to be switched over to CVSup. The purpose of sup/CVSup was so that a nightly tar file isn't required for those that want to keep ...
    The Hermit HackerThe Hermit Hacker
    Aug 27, 1997 at 12:10 am
    Aug 27, 1997 at 6:08 pm
  • Hi, when I try: plaut= -- Set to ISO plaut= Set DateStyle to 'ISO'; SET VARIABLE plaut= show DateStyle; NOTICE:DateStyle is Postgres with US (NonEuropean) conventions SHOW VARIABLE BTW: A Perldoc ...
    Michael ReifenbergerMichael Reifenberger
    Aug 25, 1997 at 8:43 pm
    Aug 26, 1997 at 7:50 am
  • What about NetBSD/alpha? Are there any other OS's that run on an alpha? Are you trying to distinguish a binary format or an OS type here? Signature failed Preliminary Design Review. Feasibility of a ...
    Henry B. HotzHenry B. Hotz
    Aug 22, 1997 at 11:33 pm
    Aug 26, 1997 at 5:02 am
  • Here is the actual command from initdb and it's output: [postgres@druid.net:src] $ echo $PGDATA /usr/local/postgres/data [postgres@druid.net:src] $ postgres -F -Q -D$PGDATA template1 ---debug info--- ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Aug 22, 1997 at 1:40 pm
    Aug 26, 1997 at 3:18 am
  • I have botched up pg_attribute when I changed attdisbursion. I think it is because I treated float32 as a byvalue attribute. Try select * from pg_attribute for an example. - -- Bruce Momjian ...
    Bruce MomjianBruce Momjian
    Aug 22, 1997 at 1:56 am
    Aug 22, 1997 at 3:14 am
  • Hi Bruce, what's wrong with integration of the oracle_compat stuff ? I wrote also a man page for these functions. It's included in the pgsql-distribution (pgsql/src/man/oracle_compat.3). Edmund - -- ...
    Edmund MerglEdmund Mergl
    Aug 14, 1997 at 9:41 am
    Aug 29, 1997 at 2:18 am
  • As many of you know, the postgres backend has many functions that are defined as extern, but should be static. It is normally hard for a coder to find them easily. I wrote the following script to ...
    Bruce MomjianBruce Momjian
    Aug 19, 1997 at 4:33 am
    Aug 19, 1997 at 6:58 am
  • Hi, I'm missing the above aggregates for timespan whereas min()/max() do exist. Bye! - ---- Michael Reifenberger Plaut Software GmbH, R/3 Basis ------------------------------
    Michael ReifenbergerMichael Reifenberger
    Aug 17, 1997 at 6:55 pm
    Aug 18, 1997 at 1:21 pm
  • Hi Ruth... Please accept this as permission to distribute PostgreSQL on your CD-ROM, but I am curious...will you be distributing the source code? If so, please also give appropriate pointers to ...
    The Hermit HackerThe Hermit Hacker
    Aug 12, 1997 at 2:14 am
    Aug 13, 1997 at 7:28 pm
  • The following file has been uploaded to ftp.postgresql.org: postgresql-v6.1-little-jdbc.tar.gz This is a JDBC v6.1.1 compatible driver (little endian). pg_jdbc.tar.gz is the big endian version for ...
    AdrianAdrian
    Aug 6, 1997 at 1:16 am
    Aug 8, 1997 at 12:33 pm
  • ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ or SELECT ... INTO TABLE ... BTW, having select into table I don't understand why you guys so much want to have ALTER DROP COLUMN: it's hard to implemet, it's ...
    Vadim B. MikheevVadim B. Mikheev
    Aug 30, 1997 at 8:24 am
    Aug 31, 1997 at 9:17 am
  • Hi, just CVSupped current and the new pg_passwd.c seems to truncated to 75 lines. Mistake? Bye! - ---- Michael Reifenberger Plaut Software GmbH, R/3 Basis ------------------------------
    Michael ReifenbergerMichael Reifenberger
    Aug 25, 1997 at 10:01 pm
    Aug 25, 1997 at 11:27 pm
  • As you know, I have gotten the time it takes to start a backend, and execute a single select statement that returns no rows down from 0.18 to 0.11 seconds. This is time as reported by gprof on the ...
    Bruce MomjianBruce Momjian
    Aug 25, 1997 at 4:01 pm
    Aug 25, 1997 at 7:17 pm
  • echo 4096 /proc/sys/kernel/file-max - -- Massimo +----------------------------------------------------------------------+ +----------------------------------------------------------------------+ ...
    Massimo Dal ZottoMassimo Dal Zotto
    Aug 22, 1997 at 12:00 pm
    Aug 24, 1997 at 8:15 pm
  • Sounds to me that this is something that the Webmaster(tm) should be watching and documenting, no? :) Make some sort of 'Upcoming Features for...' page? As something is added/fixed, add it on, and ...
    The Hermit HackerThe Hermit Hacker
    Aug 23, 1997 at 12:10 am
    Aug 24, 1997 at 1:32 am
  • Hi All, I am looking into the problem with initdb running on a linux alpha. For some reason, when the postgres backend is running in -boot mode, when it gets to parsing the last 10 or so directives ...
    IntegrationIntegration
    Aug 22, 1997 at 6:06 am
    Aug 22, 1997 at 7:23 am
  • Pulled this mornings CVSup and did a full re-compile and the following pops up: make -C parser all PORTNAME=sparc_solaris make[2]: Entering directory `/opt/postgres/pgsql/src/backend/parser' ...
    Adrian HallAdrian Hall
    Aug 20, 1997 at 1:43 pm
    Aug 21, 1997 at 12:46 am
  • I have found out why a backend startup does so many open() calls. The problem is backend/storage/file/fd.c. Every time it opens a file descriptor, it opens /dev/null first, and if that succeeds, it ...
    Bruce MomjianBruce Momjian
    Aug 15, 1997 at 2:25 am
    Aug 20, 1997 at 11:33 pm
  • This has been discussed, and raised, sooooo many times. We will not adopt a Linux kernel like release scheduale, period. End of story. The fat lady has sung. Elvis has *left* the store. I *would* be ...
    The Hermit HackerThe Hermit Hacker
    Aug 19, 1997 at 10:39 pm
    Aug 20, 1997 at 9:33 am
  • Just added in appropriate links to the interfaces directory so that it compiles now... Edmund(sp?)...what, if anything, should be done with the perl5 stuff? Right now, in interfaces/Makefile, I have ...
    The Hermit HackerThe Hermit Hacker
    Aug 16, 1997 at 8:57 pm
    Aug 19, 1997 at 4:46 am
  • Hi, After compiling with the patch provided 8/15, initdb returns: WARN:BuildFuncTupleDesc: function mkoidname(opaque, opaque) does not exist initdb: could not create template database initdb: ...
    IntegrationIntegration
    Aug 16, 1997 at 8:35 pm
    Aug 17, 1997 at 12:23 am
  • Hi everybody, I can't help but feel a teensy bit slighted. I recently posted a relatively detailed message to all postgres mailing lists that I thought appropriate, asking a question that I thought ...
    IntegrationIntegration
    Aug 14, 1997 at 4:06 pm
    Aug 15, 1997 at 2:48 pm
  • Hey... Well, after finally getting a chance to upgrade the server, I've got CVSup up and running. It still needs to be configured/fine tuned but for those interested in trying it out, here is what ...
    The Hermit HackerThe Hermit Hacker
    Aug 9, 1997 at 10:32 pm
    Aug 11, 1997 at 11:48 am
  • Hi... I'm trying to clean things up a bit, and have renamed the following two source trees. Those connecting directly to postgresql.org, having access to the source tree, will want to note the ...
    The Hermit HackerThe Hermit Hacker
    Aug 9, 1997 at 10:47 pm
    Aug 9, 1997 at 11:51 pm
  • I am looking at postmaster.c to try and change it to keep backends up and ready for connections, rather than starting on each time a connection is requested. Looking at the code, I see us doing an ...
    Bruce MomjianBruce Momjian
    Aug 7, 1997 at 4:52 pm
    Aug 7, 1997 at 6:21 pm
  • I have finished testing the new psort code and have updated the source tree. For those with sup, the updates will appear in sup at 6PM, and will make sorts over 512k work properly that I broke last ...
    Bruce MomjianBruce Momjian
    Aug 6, 1997 at 5:13 pm
    Aug 6, 1997 at 6:26 pm
  • I don't have a small test-case for the following ... it was once of the regression-tests; if you really want, I can try narrow it down. - -Stuart Kemp ABW: Array bounds write (31 times) This is ...
    Stuart KempStuart Kemp
    Aug 2, 1997 at 8:10 pm
    Aug 3, 1997 at 6:40 pm
  • I added new pg_trigger relation, so - dump/reload... I'm awaiting for Thomas' changes in parser before doing CREATE/DROP TRIGGER syntax. Proposed syntax: CREATE TRIGGER _name_ [(arg1, ...)] ON ...
    Vadim B. MikheevVadim B. Mikheev
    Aug 31, 1997 at 12:44 pm
    Aug 31, 1997 at 12:44 pm
  • Hi all, I've noticed thet "pg_dump" is not creating restorable dumps when new TYPE's have been defined in the database. eg. bash$ grep -i default ~/alldbdump.230897a CREATE TYPE sotime ( ...
    Keith ParksKeith Parks
    Aug 27, 1997 at 5:31 pm
    Aug 28, 1997 at 3:57 am
  • [ This is the same message sent to patches, but without the patch.] I have found from profiling that heap_getattr() is called over 20,000 times for a single startup/query combination. That is a lot ...
    Bruce MomjianBruce Momjian
    Aug 26, 1997 at 11:41 pm
    Aug 26, 1997 at 11:58 pm
  • For those people who are still experiencing initdb core dumps, please find the line in initdb that is the problem, and put an exit statement just before it and run the initdb. Then, using a postgres ...
    Bruce MomjianBruce Momjian
    Aug 26, 1997 at 2:36 pm
    Aug 26, 1997 at 3:15 pm
  • RESUME Name, surname: Andrew Protasov Looking for opportunity: I am interested in relocation to receive full time or long term contract job in US or Europe. Development tools are Oracle, Delphi, C. I ...
    Andrew ProtasovAndrew Protasov
    Aug 7, 1997 at 5:35 pm
    Aug 25, 1997 at 1:44 pm
  • - -- Posted to hackers for discussion rather than patches for inclusion -- Well, here's a first stab at it. I'm not completely happy with this as I'll explain but first, here's what I have. ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Aug 22, 1997 at 4:21 am
    Aug 22, 1997 at 5:53 am
Aug 1997
1234
Group Navigation
period‹ prev | Aug 1997 | next ›
Group Overview
grouppgsql-hackers @
categoriespostgresql
discussions161
posts404
users60
websitepostgresql.org...
irc#postgresql

60 users for August 1997

Bruce Momjian: 106 posts The Hermit Hacker: 63 posts Vadim B. Mikheev: 57 posts Integration: 15 posts D'Arcy J.M. Cain: 14 posts RHS Linux User: 12 posts Maarten Boekhold: 10 posts Michael Reifenberger: 9 posts Stan Brown: 7 posts Igor: 6 posts Robson Paniago de Miranda: 6 posts Andrew Martin: 5 posts Edmund Mergl: 5 posts John Polstra: 5 posts Massimo Dal Zotto: 5 posts Neil D. Quiogue: 5 posts Stuart Kemp: 5 posts Thomas G. Lockhart: 5 posts Adrian Hall: 4 posts Keith Parks: 4 posts
show more
Archives