Re: [HACKERS] PostgreSQL win32 NT4
NT4 is officially dead, IMHO no need for PostgreSQL to officially support it, let's leave place for companies offering commercial postgresql versions to work on it if they have enough customer requests. BTW Win 2000 is more or less 6 years old now ... Regards Paolo Magnus Hagander [EMAIL PROTECTED] ha scritto Hi! I'm looking at a way to fix the issues with admin privileges on Win32 - specifically by finding a way to give up all admin stuff before the server starts, when possible (and otherwise fail just as today). I think I can do this. However, it is not possible to do this in a way that's compatible with NT4. The APIs werent' created then. So from this, I see two ways to proceed: 1) Do it in a Windows 2000+ way. This will once and for all break compatibility with NT4 - pg simply will not work any more. 2) Code around it by dynamically loading these functions when available. With this solution it will still run on NT4 - without this capability. It will however be quite a bit more code, since all the functions have to be loaded at runtime. Now (2) can be done, and it's not hugely much more code for this change. But the amount of places with issues will continue to go up, and eventually we're going to have to pull the plug, IMHO. The question is - is it time to do it now? Or do we wait until we hit a change that leaves us no choice at all? (Oh, and as everybody knows, NT4 isn't supported by Microsoft any more, see http://www.microsoft.com/ntserver/ProductInfo/Availability/Retiring.asp) Comments? //Magnus ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster
Re: [HACKERS] Why don't we allow DNS names in pg_hba.conf?
Yes, it would be good to implement this. A warning about security and (possible) slow connections due to name resolution issues should be placed in the docs. Regards paolo elein [EMAIL PROTECTED] ha scritto I also support this change. My clients have tended to move machines and networks around a lot as well as move databases from machine to machine. It would be nice to let the network gurus concentrate on getting the dns servers up and correct and leverage that work instead of having to change pg_hba.conf when these changes occur. elein [EMAIL PROTECTED] On Sun, Jan 01, 2006 at 01:30:46PM -0500, Tom Lane wrote: I was reminded of $subject by http://archives.postgresql.org/pgsql-admin/2006-01/msg2.php While I haven't tried it, I suspect that allowing a DNS host name would take little work (basically removing the AI_NUMERICHOST flag passed to getaddrinfo in hba.c). There was once a good reason not to allow it: slow DNS lookups would lock up the postmaster. But now that we do this work in an already-forked backend, with an overall timeout that would catch any indefinite blockage, I don't see a good reason why we shouldn't let people use DNS names. Thoughts? regards, tom lane ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [HACKERS] HOOKS for Synchronous Replication?
I can only add that patched code did not build on windows, contacted author about that and never got an answer back. Regards paolo Tom Lane [EMAIL PROTECTED] ha scritto Christopher Kings-Lynne [EMAIL PROTECTED] writes: Anyone remember this patch? http://gorda.di.uminho.pt/community/pgsqlhooks/ The discussion seems to be pretty minimal: http://archives.postgresql.org/pgsql-hackers/2005-06/msg00859.php Does anyone see a need to investigate it further? I had hoped to see some comments from the Slony people about it. I'd feel better about the validity of a set of hooks if more than one project agreed that it was useful/appropriate ... regards, tom lane ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
[HACKERS] Tablespace on ramdisk
Hi, does anyone have experiences about putting a tablespace on ramdisk? Does it work (and keep working after a restart of the server)? Thanks in advance for any insight. Paolo ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
Re: [HACKERS] Tablespace on ramdisk
I'd like to mimic MySQL's in-memory tables (engine=memory), which structure survives a server restart (data lost of course). I suspected that a server restart would be a problem in this case. Thank you anyway. Paolo Tino Wildenhain [EMAIL PROTECTED] ha scritto [EMAIL PROTECTED] schrieb: Hi, does anyone have experiences about putting a tablespace on ramdisk? Does it work (and keep working after a restart of the server)? Thanks in advance for any insight. Yes it does work as long as you dont restart your server. Postgres does not appreciate disappearing cluster data. What are you trying to solve btw? ++Tino ---(end of broadcast)--- TIP 1: 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: Have you searched our list archives? http://archives.postgresql.org
Re: [HACKERS] MERGE vs REPLACE
I think you translated it correctly, MySQL has another way of specifying this which is INSERT ... ON DUPLICATE KEY UPDATE ... (http://dev.mysql.com/doc/refman/5.0/en/insert.html) Regards Paolo Jochem van Dieten [EMAIL PROTECTED] ha scritto On 11/13/05, Petr Jelinek wrote: I am really not db expert and I don't have copy of sql standard but you don't need to use 2 tables I think - USING part can also be subquery (some SELECT) and if I am right then you could simulate what REPLACE does because in PostgreSQL you are not forced to specify FROM clause in SELECT. So you could in theory do MERGE INTO tablename USING (SELECT 1 AS myid) ON (tablename.id = myid) ... I think the MySQL statement: REPLACE INTO table (pk, col1, col2, col3) VALUES (2, '-00-00', NULL, 3) would translate into the following MERGE statement: MERGE INTO table target USING (2 as pknew , NULL as col1new, NULL as col2new, 3 as col3new) source ON target.pknew = source.pk WHEN MATCHED THEN UPDATE SET col1 = col1new, col2 = col2new, col3 = col3new WHEN NOT MATCHED THEN INSERT (pk, col1, col2, col3) VALUES (pknew, col1new, col2new, col3new) It might not be the most elegant solution, but I don't see why it won't work. Jochem ---(end of broadcast)--- TIP 4: Have you searched our list archives? http://archives.postgresql.org ---(end of broadcast)--- TIP 4: Have you searched our list archives? http://archives.postgresql.org
[HACKERS] Postgresql 8.1 XML2
Hi all, I just installed PostgreSQL 8.1 win32 and didn't find option to install contrib/xml2, is it available on win32? Shall I build it on my own? Thanks in advance Paolo ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster
[HACKERS] Constraint exclusion, some questions
Hi, I'm testing constraint exclusion on PostgreSQL 8.1 beta 1 on windows with the GEOIP countries database and I have a few questions: 1. Can I say that an index created on the parent table is a global index, an index that spans over data in all derived tables? 2. Can I say that an index created on any of the child or inheriting table is a partition index, that spans data belonging that single table only? (explain plan seems to confirm that, just to be shure) 3. Constraint exclusion works (prunes child tables) only if your query contains a condition based on the check constraint (say constraint A) and it doesn't get the fact that the data requested lies just in one child table if the condition is on another column (say constraint B), even if global and partition indexes are present for constraint B Based on that quick look it would be great to add the ability for the optimizer to find the good child table from a global index, which would make table partitioning or constraint exclusion even more effective. Hope what I wrote makes sense. Best regards paolo ---(end of broadcast)--- TIP 5: don't forget to increase your free space map settings
[HACKERS] Rendezvous/Bonjour broken in 8.1 beta
Hi, I'm trying to compile Postgresql 8.1 beta on my own in mingw/windows. I tried to compile it with --with-bonjour but it is looking for the wrong file, it looks for DNSServiceDiscovery.h but Apple provides a file named dns_sd.h (Bonjour SDK for windows as of May 5 2005), after renaming it ./configure went ok but make ended with errors: ... In file included from postmaster.c:92: d:/mingw/bin/../lib/gcc/mingw32/3.4.2/../../../../include/DNSServiceDiscovery/DNSServiceDiscovery.h:52: error: syntax error before uint8_t d:/mingw/bin/../lib/gcc/mingw32/3.4.2/../../../../include/DNSServiceDiscovery/DNSServiceDiscovery.h:53: error: syntax error before int8_t d:/mingw/bin/../lib/gcc/mingw32/3.4.2/../../../../include/DNSServiceDiscovery/DNSServiceDiscovery.h:54: error: syntax error before uint16_t d:/mingw/bin/../lib/gcc/mingw32/3.4.2/../../../../include/DNSServiceDiscovery/DNSServiceDiscovery.h:55: error: syntax error before int16_t postmaster.c:249: error: syntax error before errorCode postmaster.c: In function `PostmasterMain': postmaster.c:769: warning: implicit declaration of function `DNSServiceRegistrationCreate' postmaster.c:774: error: `DNSServiceRegistrationReply' undeclared (first use in this function) postmaster.c:774: error: (Each undeclared identifier is reported only once postmaster.c:774: error: for each function it appears in.) postmaster.c:774: error: syntax error before reg_reply postmaster.c: At top level: postmaster.c:1042: error: syntax error before errorCode postmaster.c: In function `reg_reply': postmaster.c:1043: warning: old-style parameter declaration postmaster.c: At top level: postmaster.c:1043: warning: 'reg_reply' defined but not used is this supported on windows? Do you have any plans to restore this function? Thanks in advance. ---(end of broadcast)--- TIP 5: don't forget to increase your free space map settings
Re: [HACKERS] Rendezvous/Bonjour broken in 8.1 beta
Hi Andrew, you wrote: I don't recall ever hearing that this has been done. bonjour is the new name of Apple's rendezvous technology, an old email (http://archives.postgresql.org/pgsql-hackers/2005-05/msg00739.php) stated that rendezvous support was working at least on macosx and windows, so I tried. We should not be renaming other people's files - if necessary we should adjust our configure tests and #include directives. But I guess you did that just to make some progress. Yes, also keep in mind that I'm a programming illiterate and I was doing some do it yourself tests What's on lines 52 .. 55 of that file? Lines 51 and 52 are: #ifndef _MSL_STDINT_H typedef UINT8 uint8_t; Thanks and best regards paolo ---(end of broadcast)--- TIP 5: don't forget to increase your free space map settings
