FAQ

Search Discussions

228 discussions - 575 posts

  • Well, it is possible, and now is a good time to ask the list if the following is a _bad_ idea: To set the backend timezone, call setenv() from the backend to set the "TZ" environment variable. Is ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 7, 1997 at 1:14 am
    Oct 14, 1997 at 3:13 pm
  • Is there any compelling reason to retain the various "specialty" character types like char2, char4, etc.? I'm not sure if they are used internally, but I don't think that they provide any real ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 24, 1997 at 3:38 pm
    Oct 27, 1997 at 4:35 pm
  • Okay, well, I don't really expect anyone to be able to respond *that* fast, but I'm wrapping up v6.2 in ~10minutes...I believe that everyone got in what they were hoping to get in *cross fingers* ...
    The Hermit HackerThe Hermit Hacker
    Oct 2, 1997 at 7:19 pm
    Oct 3, 1997 at 3:19 am
  • In exactly one hour, I will be making the final v6.2 RELEASE tar file, just in case someone has one last man page that they want to modify...no source changes are to go in, only text related. Marc G. ...
    The Hermit HackerThe Hermit Hacker
    Oct 2, 1997 at 1:53 am
    Oct 2, 1997 at 9:10 am
  • This is a multi-part message in MIME format. - --------------06E108FE176F2FFAC57CD0D0 Content-Type: text/plain; charset=iso-8859-2 Content-Transfer-Encoding: 7bit unsubscribe - ...
    Kamil NagrodzkiKamil Nagrodzki
    Oct 11, 1997 at 5:18 pm
    Oct 30, 1997 at 1:10 am
  • Do we have to do a dump/reload of v6.0 to v6.1 before its possilbe to go to v6.2? My upgrade today failed *miserably* :( Just a little bit of what is scrolling up my screen :( query buffer max length ...
    The Hermit HackerThe Hermit Hacker
    Oct 3, 1997 at 8:16 pm
    Oct 6, 1997 at 2:17 am
  • Hi, vadim has already added the changes for loadable PL support to the backend and said he would add PL/Tcl to CVS these days - thanks vadim. I really like Tcl and it wasn't that much work to build ...
    WieckWieck
    Oct 29, 1997 at 12:22 pm
    Oct 30, 1997 at 3:29 pm
  • I'd like to discuss this too. Would like to keep the parser fairly clean (it's gunked up a lot just with SQL syntax :( but there should be a good place to put this. Another (similar?) thing to look ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 15, 1997 at 1:57 pm
    Oct 16, 1997 at 5:23 am
  • When I do an initdb I get the error messages at the end of this post. This has happened not only in 6.2, but a re-install of 6.1.1. Funny thing is, I had the first version of 6.1.1 up and running. ...
    Ted.rolleTed.rolle
    Oct 8, 1997 at 2:28 am
    Oct 9, 1997 at 5:20 am
  • [warning - I've done a LOT of clipping - just sampled from the message to get some applicable data] To be quite honest this whole problem wouldn't exist if "numeric"-type scanning were implemented... ...
    TeunisTeunis
    Oct 29, 1997 at 2:12 pm
    Oct 30, 1997 at 8:56 pm
  • Hi, I'm sending you this mail because you have expressed interest on the subject in the past. Short refresher: I want functionallity in postgresql so that when I have a field with contents 'EXILE ON ...
    Maarten BoekholdMaarten Boekhold
    Oct 2, 1997 at 10:13 am
    Oct 5, 1997 at 6:08 pm
  • Is the new function returning the number of modified tuples documented in the manual page for libpq? I tried to identify the new function by name and either missed the reference to it or it was not ...
    Marc Howard ZuckmanMarc Howard Zuckman
    Oct 16, 1997 at 2:37 am
    Oct 16, 1997 at 4:43 am
  • And just as I started using oid's to handle relationships between objects.... (better handled using keys or the like I know) [and how would objects be handled?] and dynamic type lookup of objects? ...
    TeunisTeunis
    Oct 8, 1997 at 2:46 am
    Oct 13, 1997 at 3:51 pm
  • Hi, just compiled Tcl8.0 and rebuild pltcl.so against libtcl8.0.so. Worked without a single change and speeds up things as expected. The Tcl version of check_primary_key executes in 60 to 70 % of the ...
    WieckWieck
    Oct 20, 1997 at 2:39 pm
    Oct 27, 1997 at 10:33 am
  • Hi, just compiled 6.2 on a BSDI box: uname -a BSD/OS www.clever.net 2.1 BSDI BSD/OS 2.1 and got some strange tests to fail: int2 .. failed int4 .. failed Caused by the error-message spit out by the ...
    Maarten BoekholdMaarten Boekhold
    Oct 12, 1997 at 2:57 pm
    Oct 14, 1997 at 5:40 pm
  • 'now' is evaluated by parser and constant value is returned. This value is stored in catalog... To get what you need a function returning current datetime should be used as default expression. ...
    Vadim B. MikheevVadim B. Mikheev
    Oct 7, 1997 at 7:45 am
    Oct 9, 1997 at 4:13 pm
  • Hi all, here comes something for the wish-list: the current implementation accepts almost everything as input for the datatype bool, but it treats most of the input as false: 0 - false 1 - false '0' ...
    Edmund MerglEdmund Mergl
    Oct 5, 1997 at 7:42 am
    Oct 6, 1997 at 5:51 pm
  • So, I have a stupid question. When we implement subselects and unions, how does one eliminate duplicate rows from queries? For example, if a table t has one integer field i, and 3 rows with values 1, ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 24, 1997 at 3:41 am
    Oct 31, 1997 at 4:15 pm
  • Hi! Bruce, did you begin with them ? I agreed that subqueries should be implemented like SQL-funcs, but I would suggest to don't CREATE FUNCTION - this is quite bad for performance, but use some new ...
    Vadim B. MikheevVadim B. Mikheev
    Oct 30, 1997 at 6:16 am
    Oct 31, 1997 at 3:28 am
  • Hi all, the following session shows an example with a transaction behavior, which is somewhat non-standard (postgresql-6.2): pgperltest= create table gaga (id int, lala bool); CREATE pgperltest= ...
    Edmund MerglEdmund Mergl
    Oct 5, 1997 at 5:57 pm
    Oct 31, 1997 at 2:54 am
  • Dear all, Could it be possible to have 8bit clean (according to locales) identifiers? At least for database names), In the same mood, could ORDER BY be locales-aware ? Allways on the same topic, the ...
    Pierre-Yves FOUCOUPierre-Yves FOUCOU
    Oct 21, 1997 at 8:34 am
    Oct 30, 1997 at 3:25 pm
  • It seems that we can build btree indices near twice faster than now. Sorting in nbtsort.c is not good: create index on table with =~ 800 000 tuples (int4 column) takes =~ 450 sec on i586, 150 Mz, ...
    Vadim B. MikheevVadim B. Mikheev
    Oct 28, 1997 at 6:03 am
    Oct 30, 1997 at 2:39 am
  • Hello, There is a "problem" in the src/backend/utils/adt/datetime.c module function date2tm near line 377. 0ccurs when you do: select datetime('now','now') it returns the correct date and time plus ...
    Lenzi, SergioLenzi, Sergio
    Oct 27, 1997 at 10:07 am
    Oct 28, 1997 at 2:45 am
  • There is still support in gram.y for the binary cursor declaration statement. How does it actually get used? Are fields sent from the backend to the frontend in binary form, and if so how does this ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 24, 1997 at 3:45 pm
    Oct 26, 1997 at 2:33 am
  • Are we finished with non-dump/reload releases for v6.2.x? When we are, I will start making changes to the source tree which require dump/reload... If desirable, I would be happy to maintain a v6.2.x ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 23, 1997 at 1:51 pm
    Oct 24, 1997 at 2:14 am
  • Can I rename PQcmdTuples to PQcmdtuples, so it is consistent with other libpq functions? - -- Bruce Momjian maillist@candle.pha.pa.us ------------------------------
    Bruce MomjianBruce Momjian
    Oct 17, 1997 at 2:20 am
    Oct 17, 1997 at 7:26 pm
  • In order to optimise the table opening in PgAccess, I have choosen the method of using cursors. So, I make a cursor for "select * from thattable" and fetch 100 in it. The ammount of information ...
    Constantin TeodorescuConstantin Teodorescu
    Oct 14, 1997 at 3:14 pm
    Oct 16, 1997 at 3:47 am
  • Bruce... When initdb runs, it runs its own postgres process, since none already exists. If it somehow possible to have pg_dumpall do something similar, in such a way that no external connections can ...
    The Hermit HackerThe Hermit Hacker
    Oct 2, 1997 at 8:12 pm
    Oct 3, 1997 at 1:17 am
  • Is this what it's supposed to do? I would expect that 'null' is actually a value, and so there should be 'null' in the column. I thought that defaults would act like: create table bb2 (a float ...
    Maarten BoekholdMaarten Boekhold
    Oct 1, 1997 at 4:47 pm
    Oct 2, 1997 at 2:44 am
  • I have attached a diff again vanilla 6.2.1 implementing Unix sockets as an alternative to the normal TCP-sockets. enjoy, - -- - ----------------------------------------------------------- Göran ...
    Goran ThyniGoran Thyni
    Oct 30, 1997 at 10:26 pm
    Oct 31, 1997 at 7:28 pm
  • I like the idea of using PL/SQL as a base (for that matter I would also like the (+)= syntax for outer joins, when we have them) In that case we mainly need to define what extensions to it we need in ...
    Hannu KrosingHannu Krosing
    Oct 30, 1997 at 9:39 am
    Oct 30, 1997 at 11:49 am
  • - --------------546CB36B63960A968BC3A052 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit In poking through my SQL book, I found a description of SQL92 "delimited ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 27, 1997 at 5:30 pm
    Oct 28, 1997 at 4:13 am
  • I have html-ized the tools/BACKEND_DIRS file, adding more descriptions, and making the directory names into links. I see this as an easy way for people to understand the basic backend operation. This ...
    Bruce MomjianBruce Momjian
    Oct 27, 1997 at 4:57 pm
    Oct 27, 1997 at 9:51 pm
  • Altho its been a very short period since the last release, there were a few changes that the developers wished to make in order to improve the server, but with the tight scheduale we had, we couldn't ...
    The Hermit HackerThe Hermit Hacker
    Oct 18, 1997 at 8:48 pm
    Oct 22, 1997 at 2:59 am
  • Hi, what can cause the occurence of the following FATAL elog()? btree: BTP_CHAIN flag was expected This is done from access/nbtree/nbtsearch.c:216 : if (_bt_skeycmp(rel, keysz, scankey, page, ...
    Maarten BoekholdMaarten Boekhold
    Oct 18, 1997 at 9:34 am
    Oct 19, 1997 at 10:33 am
  • Hi PostgreSQL developers, I've been very busy and lazy the last months. But I would like to come back. First of all congratulations for 6.2. Very great. Now that we have basic support for database ...
    WieckWieck
    Oct 7, 1997 at 1:03 pm
    Oct 17, 1997 at 1:38 pm
  • Vadim, are you going to have time to update the regression tests for constraints and triggers? They both are failing due to improvements/changes in the way NULL is handled (I think). btw, I have one ...
    Thomas G. LockhartThomas G. Lockhart
    Oct 17, 1997 at 1:15 am
    Oct 17, 1997 at 3:22 am
  • Is it true our LIKE does not handle []? Is that a standard feature of LIKE? I don't see anything in utils/adt/like.c which would suggest this. - -- Bruce Momjian maillist@candle.pha.pa.us ...
    Bruce MomjianBruce Momjian
    Oct 14, 1997 at 5:47 pm
    Oct 16, 1997 at 2:38 am
  • Hello, To facilitate the writing of documentation it would be helpful if we had some way of knowing when new features were going to be added to pgsql. To this end I have asked Marc to create a new ...
    Eze OgwumaEze Ogwuma
    Oct 14, 1997 at 8:58 pm
    Oct 15, 1997 at 3:24 am
  • But now we know about problems with relation reference counts in bufmgr (fixed) and backward cursor with order by (fixing now). DEFAULT already fixed (need to update regression tests...) So, it's ...
    Vadim B. MikheevVadim B. Mikheev
    Oct 14, 1997 at 2:53 am
    Oct 14, 1997 at 5:07 pm
  • There are many advantages to things in a database also being generic objects and they aint it if they don't have an OID. The first thing that comes in mind is documenting your database: I think that ...
    Hannu KrosingHannu Krosing
    Oct 9, 1997 at 10:46 am
    Oct 13, 1997 at 6:13 pm
  • Hi... Is this something that should be worried about? Its a v6.2 system that I'm currently running a pg_perl script on copying from another table into this one: acctng= select count(userid) from ...
    The Hermit HackerThe Hermit Hacker
    Oct 8, 1997 at 1:42 am
    Oct 8, 1997 at 3:20 am
  • Hi... I don't know how much use they will be, but do we want to try and do up binary distributions for v6.2? I'm going to work on a FreeBSD one... figuring it might help get rid of part of the ...
    The Hermit HackerThe Hermit Hacker
    Oct 3, 1997 at 1:38 am
    Oct 4, 1997 at 11:40 pm
  • Anyone know whatever happened to this? Marc G. Fournier Systems Administrator @ hub.org primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org - ---------- Forwarded message ---------- ...
    The Hermit HackerThe Hermit Hacker
    Oct 3, 1997 at 1:43 am
    Oct 3, 1997 at 12:00 pm
  • Three times faster. - -- Bruce Momjian maillist@candle.pha.pa.us ------------------------------
    Bruce MomjianBruce Momjian
    Oct 2, 1997 at 3:47 pm
    Oct 2, 1997 at 5:15 pm
  • pgsql-docs isn't part of this archive. - -- Eze Ogwuma ------------------------------
    Eze OgwumaEze Ogwuma
    Oct 1, 1997 at 4:19 pm
    Oct 2, 1997 at 2:04 am
  • Added to TODO. - -- Bruce Momjian maillist@candle.pha.pa.us ------------------------------
    Bruce MomjianBruce Momjian
    Oct 30, 1997 at 3:06 am
    Oct 30, 1997 at 6:45 pm
  • 1. Permissions on tables was not preserved when moving from 6.1 to 6.2 (using 6.1 postmaster and 6.2 pg_dumpall). 2. \z in psql does not show permissions for sequences! /* m */ ...
    Mattias KregertMattias Kregert
    Oct 8, 1997 at 7:02 pm
    Oct 30, 1997 at 4:30 am
  • Hi, Your name : Gerd Thielemann Your email address : gerd@fru19.hub.de System Configuration - --------------------- Architecture (example: Intel Pentium) : Intel PentiumPro Operating System (example: ...
    Gerd Thielemann AEKGerd Thielemann AEK
    Oct 8, 1997 at 12:02 pm
    Oct 30, 1997 at 4:09 am
  • Hi, FreeBSD (esp. outside the USA) defaults to MD5 resulting crypt(). PostgreSQL seems to be somewhat hardcoded for DES. pg_passwd checks for length 13. The resulting file looks ok though. ...
    Michael ReifenbergerMichael Reifenberger
    Oct 3, 1997 at 6:37 pm
    Oct 30, 1997 at 3:37 am
Group Navigation
period‹ prev | Oct 1997 | next ›
Group Overview
grouppgsql-hackers @
categoriespostgresql
discussions228
posts575
users89
websitepostgresql.org...
irc#postgresql

89 users for October 1997

Bruce Momjian: 125 posts Vadim B. Mikheev: 88 posts Thomas G. Lockhart: 58 posts The Hermit Hacker: 42 posts Wieck: 26 posts Teunis Peters: 25 posts Maarten Boekhold: 16 posts D'Arcy J.M. Cain: 9 posts Eze Ogwuma: 9 posts Mattias Kregert: 8 posts Darren King: 7 posts Edmund Mergl: 7 posts Peter T Mount: 7 posts Massimo Dal Zotto: 6 posts Claudio R Zamana: 5 posts Gord Pusnik: 5 posts Hannu Krosing: 5 posts Jongho Won: 5 posts Kamil Nagrodzki: 5 posts Marc Howard Zuckman: 5 posts
show more
Archives