Search Discussions

204 discussions - 526 posts

  • First, little mistyping for CYCLEd sequences, sorry, it's diff: 142c142 < result = si- max; - --- And it should be mentioned: to create sequence You may run select create_sequence ('sequence_name', ...
    Vadim B. MikheevVadim B. Mikheev
    Mar 5, 1997 at 4:31 am
    Mar 11, 1997 at 1:44 am
  • I wonder if anyone is working on this yet? I have been writing code to do this outside of the engine itself and I would like to couple it more tightly if I can. Basically I create a new table called ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 14, 1997 at 7:42 pm
    Mar 17, 1997 at 3:37 am
  • I can't answer that question. However, I have usually found in the past for my own projects that the simplest way to store time is to count the number of time units since the earliest representable ...
    David FriendDavid Friend
    Mar 21, 1997 at 3:22 pm
    Mar 29, 1997 at 9:53 am
  • There is an interesting twist with folding everything to lower case, which should perhaps be mentioned in some readme: Column names which where mixed upper- and lowercase can't be accessed any more ...
    Martin J. LaubachMartin J. Laubach
    Mar 14, 1997 at 11:44 pm
    Mar 21, 1997 at 7:23 pm
  • Hi, Has anyone else run into problems with the new Oracle compatability overloaded functions? My compile of the new source falls over complaining about the re-definition of several values that appear ...
    Keith ParksKeith Parks
    Mar 15, 1997 at 1:12 pm
    Mar 16, 1997 at 6:32 am
  • About the slowness issue.... I've been playing with Derek GLidden's test program and got some strange results. He suggested I post on this list. So here is the message I sent to glidden: [I had been ...
    John EdstromJohn Edstrom
    Mar 11, 1997 at 10:56 pm
    Mar 13, 1997 at 9:23 pm
  • Does anyone know when this happened? Thomas's date additions aren't installed yet, are they. How do we fix them? Was this true in 6.0 too. ------------------------------
    Bruce MomjianBruce Momjian
    Mar 2, 1997 at 2:13 am
    Mar 27, 1997 at 8:00 pm
  • I couldn't agree more! One more time, as I didn't actually send the last one directly to the list, and no one said anything when Edmund reposted it... I propose we let Postgres overload C language ...
    Mar 17, 1997 at 11:10 am
    Mar 18, 1997 at 2:32 pm
  • Back to this timezone stuff. The struct tm has a field (tm_gmtoff) which is the offset from UTC (GMT is archaic BTW) in seconds. Is this the value you are looking for when you use timezone? Note that ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 21, 1997 at 4:31 pm
    Mar 25, 1997 at 5:03 pm
  • Hi, Does anyone know which PostgresPerl is better? So far I know I could have two choices: psotgresPerl-1.3 written by Vivek Khera. I got the software from ...
    Mar 25, 1997 at 8:48 pm
    Mar 26, 1997 at 1:46 am
  • I decided to go ahead with the required changes since no one else seems to. I don't guarantee that it is perfect but with these changes the package actually compiles. While I was at it I added to the ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 20, 1997 at 4:15 pm
    Mar 21, 1997 at 1:19 pm
  • OK, I've tried to decouple the POSIX/tzset variables from Postgres time handling. I hadn't known that some non-POSIX machines have timezone defined as a subroutine :(. Since we were discussing the ...
    Thomas LockhartThomas Lockhart
    Mar 18, 1997 at 3:25 am
    Mar 19, 1997 at 2:46 am
  • Thanks Hermy for applying the date patches. I'll pick up the snapshot and try it out. I will be making changes soon to support various _output_ formats also. I am planning the following pieces: (1) a ...
    Thomas LockhartThomas Lockhart
    Mar 15, 1997 at 2:51 pm
    Mar 17, 1997 at 2:11 am
  • Here is a bunch of patches to libpq and the backend against the 970317 tarball. The modifications in detail: - Move most of the I/O in both libpq and the backend to a set of common routines in ...
    Martin J. LaubachMartin J. Laubach
    Mar 18, 1997 at 9:15 am
    Mar 21, 1997 at 12:10 pm
  • This patch gives you the compile-time option of inputting and outputting the 'date' type in the ANSI SQL format. Index: backend/utils/adt/datetimes.c ...
    Dan McGuirkDan McGuirk
    Mar 12, 1997 at 11:55 am
    Mar 15, 1997 at 1:21 am
  • Hi... I've just brought in some extensions(?) that Edmund sent to me so that we have some Oracle compatibility. So far as I can tell, I've modified catalog/pg_proc.h properly (this stuff really isn't ...
    The Hermit HackerThe Hermit Hacker
    Mar 4, 1997 at 5:37 am
    Mar 4, 1997 at 3:26 pm
  • Is this dd-mm-yyyy (or mm-dd-yyyy) format mandated by ANSI. For an output format it seems less than ideal. The following patch will, if applied, will change the dashes to spaces and use month ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 25, 1997 at 3:43 pm
    Mar 28, 1997 at 10:30 pm
  • Hello! I found this bug, that's why Postgres didn'c compile on my linux. Please include this patch, which fixes the problem: This should be applied in /src/backend/libpq with patch -p0 - --- ...
    Gabriel AkosGabriel Akos
    Mar 28, 1997 at 8:27 am
    Mar 28, 1997 at 9:07 pm
  • Hello all. I have some DB with cgi interface. Today, I test overload Apache and Postgers and fast pressed submit button on my cgi-form. 10 or 20 or 30 time succession. It's part of my cgi-script's, ...
    Maxim KozinMaxim Kozin
    Mar 25, 1997 at 8:27 am
    Mar 25, 1997 at 10:03 am
  • On Solaris 2.5.1 I can't compile pqcomprim.c because it defines the endianness macros twice. The macro BYTE_ORDER acts as if it is defined as both LITTLE_ENDIAN *and* BIG_ENDIAN. In fact it is ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 19, 1997 at 1:12 am
    Mar 19, 1997 at 12:44 pm
  • Hi... I just went through the trouble of compiling Edmund's 'contrib' stuff as a contrib/create function "library"...with a March 3rd server... I hate to eat crow, since it has this really really bad ...
    The Hermit HackerThe Hermit Hacker
    Mar 16, 1997 at 6:47 am
    Mar 17, 1997 at 12:14 am
  • Here is what I get compiling on Solaris 2.5.1 In file included from heapvalid.c:17: ../../fmgr.h:507: warning: `F_LPAD' redefined ../../fmgr.h:501: warning: this is the location of the previous ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 16, 1997 at 1:00 am
    Mar 16, 1997 at 9:25 pm
  • I had a peek at libpq yesterday, and most of the strings it reads from the connection are just null-terminated, ie. it does something along while((c = getchar()) != EOF) if(!c) break; which explains ...
    Martin J. LaubachMartin J. Laubach
    Mar 16, 1997 at 12:04 am
    Mar 17, 1997 at 4:52 pm
  • I asked about this in questions, but didn't get much help. It seems like V6 is basically non-functional OOTB on FreeBSD 2.1.5-RELEASE, unless I've done something really stupid: 1) Doesn't compile ...
    Robert WithrowRobert Withrow
    Mar 4, 1997 at 12:21 am
    Mar 5, 1997 at 11:36 pm
  • I'll get it someday. Index: backend/utils/time/tqual.c =================================================================== RCS file: /usr/local/cvsroot/postgres95/backend/utils/time/tqual.c,v ...
    Dan McGuirkDan McGuirk
    Mar 26, 1997 at 11:24 pm
    Mar 28, 1997 at 7:08 pm
  • Hi folks, Building 6.1 from a couple of days ago fails with several errors in backend/utils/adt/dt.c. I have changed it to use DBL_MIN instead of MINDOUBLE, but can't find NAN defined anywhere in ...
    Colin KabaaraColin Kabaara
    Mar 26, 1997 at 1:04 pm
    Mar 27, 1997 at 12:52 am
  • OK, here are my design criteria: 1) data types should be optimized for internal efficiency, on the assumption that most RDBMS transactions return a few rows from a large number of candidate rows, ...
    Thomas LockhartThomas Lockhart
    Mar 24, 1997 at 6:17 pm
    Mar 25, 1997 at 7:12 pm
  • There was a lot of pre-6.0 discussion, and almost everyone agreed that if standard SQL is caseless, we should be also. ------------------------------
    Bruce MomjianBruce Momjian
    Mar 18, 1997 at 10:14 pm
    Mar 21, 1997 at 12:18 pm
  • Hi, I'm getting the following error compiling the password.c sup'd on 19th Mar:- gcc -Wall -Wmissing-prototypes -O2 -Werror -Dsparc_solaris - -I../../backend/port/sparc_solaris -I.. ...
    Keith ParksKeith Parks
    Mar 19, 1997 at 10:50 pm
    Mar 20, 1997 at 6:32 pm
  • This is a multi-part message in MIME format. - --------------F3615C1493F66A8751F5ECB Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit We have SUBJ now. I failed to connect ...
    Vadim B. MikheevVadim B. Mikheev
    Mar 18, 1997 at 4:02 pm
    Mar 19, 1997 at 2:18 am
  • That's the error I get. The problem is that /usr/include/time.h and /usr/local/postgres95/src/backend/utils/adt/dt.c both declare the variable timezone in different ways. Given that; HISTORY A ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 16, 1997 at 5:03 pm
    Mar 17, 1997 at 5:42 pm
  • I'm looking to make heavy usage of postgres on a couple of different platforms (AIX-3.2.5, BSDi-2.1/3.0, Solaris-2.5.1/sparc/x86). First I'm looking at it to see if it works (reliably) for what I ...
    Roger FlorkowskiRoger Florkowski
    Mar 15, 1997 at 1:43 pm
    Mar 17, 1997 at 1:58 pm
  • It references a function called infnan() which isn't defined. Can the person who put that in please complete the job? Thanks. - -- D'Arcy J.M. Cain | Democracy is three wolves ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 16, 1997 at 5:18 pm
    Mar 16, 1997 at 7:02 pm
  • Er, about that change in error message handling... do you think it would be a good idea? If so, I'll look into it... mjl ------------------------------
    Martin J. LaubachMartin J. Laubach
    Mar 12, 1997 at 12:14 pm
    Mar 14, 1997 at 10:34 pm
  • I've noticed that PostgreSQL does not appear to come with a reference manual. On the web I found a postscript file that turned out to be the reference manual for Postgres95 v1.01. I am just learning ...
    David FriendDavid Friend
    Mar 13, 1997 at 6:49 pm
    Mar 14, 1997 at 6:06 am
  • Dear PostgreSQL gurus, in the current sup source tree, files named password.h and password.c are mentioned (e.g., in postgres95/src/backend/libpq/Makefile and postgres95/src/backend/libpq/auth.c), ...
    Ernst MolitorErnst Molitor
    Mar 13, 1997 at 3:32 pm
    Mar 14, 1997 at 5:58 am
  • OK, last one. This patch adds an ANSI SQL 'timestamp' type. Index: backend/utils/adt/Makefile diff -u backend/utils/adt/Makefile: postgres95/backend/utils/adt/Makefile:1.5 - --- Makefile Tue ...
    Dan McGuirkDan McGuirk
    Mar 12, 1997 at 12:08 pm
    Mar 12, 1997 at 9:27 pm
  • Hi, with the recent talk about the release of 6.1, I just wanted to clarify that = the current snapshot doesn't really work with NEXTSTEP. The recent efforts = made it compile for NEXTSTEP, but it ...
    Gregor HoffleitGregor Hoffleit
    Mar 6, 1997 at 9:36 pm
    Mar 8, 1997 at 9:03 pm
  • Is there some simple way to increment dates? I have to be able to move to the next month but it seems like a generic capability would be very useful. I have tried variations of this; select from_date ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 29, 1997 at 3:53 pm
    Mar 30, 1997 at 4:01 pm
  • Porting all of PostgreSQL to Windows 95 would be a major project. Why not start by simply developing the graphical interface you talked about here? I suspect that there are a lot of people/companies ...
    David FriendDavid Friend
    Mar 27, 1997 at 5:29 pm
    Mar 30, 1997 at 2:16 am
  • Hi, the cursor concept in postgreSQL is of limited usability, because the complete block of statements - - including fetch - has to be a single transaction: begin declare mycursor cursor for select * ...
    Edmund MerglEdmund Mergl
    Mar 29, 1997 at 8:00 am
    Mar 29, 1997 at 10:33 pm
  • It looks like the ['epoch', 'now'] thing not working is because the current version is not adjusting for the time zone properly when you enter the value 'now' for an abstime. E.g.: test= create table ...
    Dan McGuirkDan McGuirk
    Mar 26, 1997 at 10:43 pm
    Mar 28, 1997 at 7:09 am
  • I've seen this performance thread pop up again, and I've had time to get a FreeBSD machine up and running (temporarily) using the same hardware as my Linux box, and I'm finding that on a Pentium ...
    Derek GliddenDerek Glidden
    Mar 27, 1997 at 1:06 am
    Mar 27, 1997 at 1:39 am
  • There is a problem with some of the calls to strftime. The second arg is missing. In all cases the buffer is CTZName which, according to the file init/globals.c, is char CTZName[8] so I have added ...
    D'Arcy J.M. CainD'Arcy J.M. Cain
    Mar 21, 1997 at 3:28 pm
    Mar 26, 1997 at 3:57 pm
  • It seems as though the following patch gives a more consistant error message; the classification for each line in this message seemed a bit off: - Tom - --- ../src/backend/parser/catalog_utils.c.orig ...
    Thomas LockhartThomas Lockhart
    Mar 25, 1997 at 3:21 am
    Mar 25, 1997 at 9:22 pm
  • To get latest snapshot (970322.tar.gz) compiled on my Linux box I had to add #include <endian.h to backend/libpq/pqcomprim.c Best regards, Oleg Oleg Bartunov, sci.researcher, hostmaster of AstroNet, ...
    Oleg BartunovOleg Bartunov
    Mar 22, 1997 at 7:15 am
    Mar 25, 1997 at 8:19 am
  • Hi, after succesful compiling of the latest snapshot (the only problem was found with missing endian.h) I tried to make regression test but it just failed. I run postgres as nohup postmaster ...
    Oleg BartunovOleg Bartunov
    Mar 22, 1997 at 7:39 am
    Mar 22, 1997 at 12:13 pm
  • Hi, just a simple question, what are the differences between PostgresSQL and Postgres95 Thanks, Yann. ------------------------------
    Yann CourtiesYann Courties
    Mar 21, 1997 at 5:09 pm
    Mar 21, 1997 at 7:00 pm
  • My test platform on AIX-3.2.5 works fine (regression test, etc) with one exception: loading of modules (ie, regression test trying to load regress.so) fails. WARN:Load of file ...
    Roger FlorkowskiRoger Florkowski
    Mar 17, 1997 at 6:03 pm
    Mar 20, 1997 at 2:01 am
  • Here's what I have come up with... mjl #include <stdio.h main() { char c[4] = { 0x11, 0x22, 0x33, 0x44 }; u_long i = *(int *) c; switch(i) { case 0x44332211: printf("little endian\n"); break; case ...
    Martin J. LaubachMartin J. Laubach
    Mar 19, 1997 at 12:06 pm
    Mar 19, 1997 at 3:36 pm
Group Navigation
period‹ prev | Mar 1997 | next ›
Group Overview
grouppgsql-hackers @

70 users for March 1997

The Hermit Hacker: 118 posts Bruce Momjian: 59 posts Dan McGuirk: 38 posts D'Arcy J.M. Cain: 38 posts Thomas Lockhart: 36 posts Vadim B. Mikheev: 32 posts Christopher Hicks: 13 posts Martin J. Laubach: 12 posts Henry B. Hotz: 9 posts John Edstrom: 9 posts Derek Glidden: 8 posts David Friend: 7 posts Edmund Mergl: 7 posts James A. Hillyerd: 7 posts Martin J. Laubach: 7 posts Darren King: 6 posts Keith Parks: 6 posts Bibach: 5 posts Lsh: 5 posts Andrew Martin: 5 posts
show more