Nabble - Firebird

tag:old.nabble.com,2006:forum-4394 Nabble - Firebird 2009-12-03T12:26:28Z Firebird is a relational database offering many ANSI SQL-92 features that runs on Linux, Windows, and a variety of Unix platforms. Firebird offers excellent concurrency, high performance, and powerful language support for stored procedures and triggers. Firebird home is <a href="http://sourceforge.net/projects/firebird/" target="_top" rel="nofollow">here</a>. tag:old.nabble.com,2006:post-26632655 Re: Fbguard - was FB3 and ... 2009-12-03T12:26:28Z 2009-12-03T12:26:28Z Carlos H. Cantu (TeamFB)-2 AWH> Do be careful if the customers are using cheap disks.  The best way <br>AWH> to make a cheap disk look fast is to cache writes rather than doing <br>AWH> them immediately so they can be applied in a disk-friendly order. <br>AWH> That totally breaks careful write and could corrupt databases for <br>AWH> either architecture. <br><br>But, if I understand it well, this would be a problem only if we have <br>a HD or energy fault, since if the stuff to be written is <br>already in the HD cache, those writings were already "authorized" by <br>FB, so, the problem is not direct related to the "killing process" <br>thing. <br><br>I guess nowadays, most of the actual SATA drives (by Seagate, Maxtor, WD, etc) <br>probably have write cache enabled by default, and I'm not sure even if <br>there is a way to disable this... I'll search for info about this. <br><br>[]s <br>Carlos <br><a href="http://www.warmboot.com.br" target="_top" rel="nofollow">http://www.warmboot.com.br</a><br>FireBase - <a href="http://www.FireBase.com.br" target="_top" rel="nofollow">http://www.FireBase.com.br</a><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26632420 Re: Fbguard - was FB3 and ... 2009-12-03T12:09:07Z 2009-12-03T12:09:07Z Ann W. Harrison Carlos, <br>> <br>> Thanks for the answers. My doubt was because some known friends told <br>> me that killing the process may lead to corruption. I'm glad to know <br>> that I can "kill" all classic processes with no worries, if I want to <br>> kick all the connections out. In fact, this "problem" was blocking me <br>> to move SS to CS in some of my customers. Seems that now I can move on <br>> safely :) <br><br>Do be careful if the customers are using cheap disks.  The best way <br>to make a cheap disk look fast is to cache writes rather than doing <br>them immediately so they can be applied in a disk-friendly order. <br>That totally breaks careful write and could corrupt databases for <br>either architecture. <br><br><br>Best regards, <br><br>Ann <br><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26632599 Re: [FB-Tracker] Created: (CORE-2789) Insert/Update Binary data via text SQL 2009-12-03T12:07:29Z 2009-12-03T12:07:29Z Björn Reimer <br>Nice feature, but I think FB should be able to read and write that format e.g. as a text representation of binary blobs e.g. for isql in hex. <br><br><br>And the main problem, which can only be solved by an engine enhancement is that adding 16k (32k chars as Hex) (currently maximum possible) via concatenation operator multiple times is very slow on larger binary blobs because fb has always to copy all the blob data on every update. <br><br><br>> Insert/Update Binary data via text SQL <br>> -------------------------------------- <br><br>>                  Key: CORE-2789 <br>>                  URL: <a href="http://tracker.firebirdsql.org/browse/CORE-2789" target="_top" rel="nofollow">http://tracker.firebirdsql.org/browse/CORE-2789</a><br>>              Project: Firebird Core <br>>           Issue Type: New Feature <br>>          Environment: All <br>>             Reporter: Stefan Heymann <br><br><div class='shrinkable-quote'><br>> With Firebird 2.5 it is now possible to insert binary data into VARCHAR or BLOB fields <br>> (by prefixing hexadecimal strings with x, like in x'0D0A'). <br>> However, these strings are limited to a string literal size of 32K and a whole command is limited to 64K. <br>> Inserting oder updating large binary things into a Blob field via a script is difficult <br>> (it is possible by multiple concatenations). <br>> My suggestion is to include a feature into the SQL language so that large <br>> (multi-Megabyte) binary streams can be stored to a blob field. <br>> (Think of photos or videos that a script will insert into a newly created database) <br>> Maybe it would also be useful to add base64 encoding, which is somewhat smaller than hex. <br>> I already discussed that with Vlad Khorsun on the FbConf 2009 in Munich/Germany. </div><br>Well, Stephan, you've implemented a quick Delphi example which shows the problem. <br><br>Björn <br><br><br>-- <br>Björn Reimer - RRZE <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26632152 Re: Fbguard - was FB3 and ... 2009-12-03T11:55:06Z 2009-12-03T11:55:06Z Carlos H. Cantu (TeamFB)-2 Sean, Ann, <br><br>Thanks for the answers. My doubt was because some known friends told <br>me that killing the process may lead to corruption. I'm glad to know <br>that I can "kill" all classic processes with no worries, if I want to <br>kick all the connections out. In fact, this "problem" was blocking me <br>to move SS to CS in some of my customers. Seems that now I can move on <br>safely :) <br><br>[]s <br>Carlos <br><a href="http://www.warmboot.com.br" target="_top" rel="nofollow">http://www.warmboot.com.br</a><br>FireBase - <a href="http://www.FireBase.com.br" target="_top" rel="nofollow">http://www.FireBase.com.br</a><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26631830 Re: Fbguard - was FB3 and ... 2009-12-03T11:33:48Z 2009-12-03T11:33:48Z Ann W. Harrison Carlos H. Cantu wrote: <br>> AP> No. Since FB2.0 engine can be safely stopped killing it with signal 2 (ctrl-C) <br>> AP> or 15 (default for kill <pid>). <br>> <br>> Is this true for CS too? I mean, can i use kill to stop all the <br>> running CS fbserver process and be sure that will not end with <br>> corrupted DB? <br><br>It has to be true if the database isn't going to be corrupted by <br>operating system crashes, power failures, etc.  As long as forced <br>write is on and actually works, Firebird is crash/kill safe in <br>all modes. <br><br>Warning!  Some disk manufacturers set their defaults to use a <br>disk cache, so it's not enough to tell Firebird to tell the <br>file system to force writes to disk.  You must be certain that <br>all parts of the system are writing bits to oxide (or some <br>stable storage) to get transaction durability and crash safety. <br><br>Firebird is crash safe despite not having a write ahead log <br>because it use "careful write", a technique that basically <br>boils down to "write the page that's pointed to before you <br>write the page that points to it."  The "pointing to" and <br>"pointed at" relationship is called a dependency graph. In <br>some cases pages are written earlier than would otherwise <br>be necessary to avoid having a cycle in the dependency graph. <br><br>Maintaining a log means that every change is written twice, <br>once to the log and once to the database.  The occasional <br>need to write a page to avoid a cycle in the dependency <br>graph is usually less of a cost that writing everything <br>twice.  On the other hand, the log writes are sequential <br>while the database writes are random and cannot easily be <br>batched because of the dependencies. <br><br>Jim's first two databases used careful write to avoid the <br>cost of the log.  His third, Netfrastructure/Falcon, does use <br>a multi-purpose log because it was designed for applications <br>that have an uneven usage pattern.  When transactions are <br>hot and heavy, the sequential write reduces latency.  When <br>calm descends, the changes migrate from the log to the database. <br>Jim's fourth database doesn't use disks for durability. <br><br>Cheers, <br><br>Ann <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26631488 Re: Fbguard - was FB3 and ... 2009-12-03T11:13:31Z 2009-12-03T11:13:31Z Leyne, Sean Carlos, <br><br>> Is this true for CS too? I mean, can i use kill to stop all the <br>> running CS fbserver process and be sure that will not end with <br>> corrupted DB? <br><br>Yes, you can kill all FBServer instances without corrupting the DB. <br><br><br>Sean <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26631287 Re: Fbguard - was FB3 and ... 2009-12-03T09:46:04Z 2009-12-03T09:46:04Z Carlos H. Cantu (TeamFB)-2 AP> No. Since FB2.0 engine can be safely stopped killing it with signal 2 (ctrl-C) <br>AP> or 15 (default for kill <pid>). <br><br>Is this true for CS too? I mean, can i use kill to stop all the <br>running CS fbserver process and be sure that will not end with <br>corrupted DB? <br><br>[]s <br>Carlos <br><a href="http://www.warmboot.com.br" target="_top" rel="nofollow">http://www.warmboot.com.br</a><br>FireBase - <a href="http://www.FireBase.com.br" target="_top" rel="nofollow">http://www.FireBase.com.br</a><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26629510 Re: plugins makefile is missing 2009-12-03T09:06:26Z 2009-12-03T09:06:26Z marius adrian popa On Wed, Dec 2, 2009 at 12:34 PM, marius adrian popa <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26629510&i=0" target="_top" rel="nofollow">mapopa@...</a>> wrote: <div class='shrinkable-quote'><br>> On Wed, Dec 2, 2009 at 11:38 AM, Adriano dos Santos Fernandes <br>> <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26629510&i=1" target="_top" rel="nofollow">adrianosf@...</a>> wrote: <br>>> marius adrian popa escreveu: <br>>>> with head and <br>>>> /autogen.sh <br>>>> <br>>>> config.status: error: cannot find input file: `builds/posix/Makefile.in.plugins' <br>>> It's certainly there. Maybe a problem in your git tree integration? <br>> <br>> sorry that seems to be a bug in git-cvsimport </div>I'm not the only one that found the bug <br><a href="http://feeding.cloud.geek.nz/2009/08/migrating-sahana-from-cvs-to-git.htm" target="_top" rel="nofollow">http://feeding.cloud.geek.nz/2009/08/migrating-sahana-from-cvs-to-git.htm</a><br>so i will use cvs2git to import the branches too <br><br>mkdir firebird2 <br>cd firebird2 <br>rsync -av rsync://firebird.cvs.sourceforge.net/cvsroot/firebird/firebird2 . <br>rsync -av rsync://firebird.cvs.sourceforge.net/cvsroot/firebird/CVSROOT . <br><br>cd .. <br>svn co --username=guest <a href="http://cvs2svn.tigris.org/svn/cvs2svn/trunk" target="_top" rel="nofollow">http://cvs2svn.tigris.org/svn/cvs2svn/trunk</a><br>cvs2svn-trunk <br><br># Dump the CVS history in a format ready for git-fast-import <br>cd cvs2svn-trunk <br>./cvs2git --blobfile ../cvs2git.blob --dumpfile ../cvs2git.dump \ <br>--username '(no author)' --fallback-encoding utf-8 ../firebird2 <br><br># Create and fill the Git repo <br>mkdir ../firebirdgit <br>cd ../firebirdgit <br>git init <br>cat ../cvs2git.{blob,dump} | git fast-import <br><br><div class='shrinkable-quote'><br>>> <br>>> <br>>> Adriano <br>>> <br>>> <br>>> ------------------------------------------------------------------------------ <br>>> Join us December 9, 2009 for the Red Hat Virtual Experience, <br>>> a free event focused on virtualization and cloud computing. <br>>> Attend in-depth sessions from your desk. Your couch. Anywhere. <br>>> <a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>>> Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br>>> <br>> </div><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26629311 Re: Fbguard - was FB3 and ... 2009-12-03T08:56:53Z 2009-12-03T08:56:53Z Alexander Peshkoff On Thursday 03 December 2009 18:22:00 Dunbar, Norman wrote: <br>> Afternoon all, <br>> <br>> following on from discussions about the deprecation of fbmgr from 2.1 <br>> and it's eventual replacement by fbguard, I have a couple of questions. <br>> <br>> When will fbguard take over all the duties of fbmgr - version 3.x or <br>> before that in 2.5? <br><br>Not all - the only duty it takes from fbmgr is starting a daemon. <br><br>> Will (does) fbguard have an option to stop the database engine as in <br>> fbmgr -shut? I'm running 2.0 on Windows and 2.1.3 on Linux and in the <br>> Linux version at least, it doesn't do anything to stop the engine. <br><br>No. Since FB2.0 engine can be safely stopped killing it with signal 2 (ctrl-C) <br>or 15 (default for kill <pid>). Use of fbmgr has always been a kind of <br>security risk - we had to keep sysdba's password in startup script. And it <br>was always possible to modify sysdba's password by some other mean instead <br>special script, which lead to inability to safely shutdown server when linux <br>was going down. Therefore this fbmgr duty was replaced by direct signal <br>processing - like 99.9% of unix daemons do. <br><br>> What will the -daemon option (from 2.5 on I believe) do? I assume it <br>> will start the engine as a daemon. (Pretty obvious really, but best to <br>> be sure). <br><br>Yes, exactly. <br><br>> I also assume that running "fbguard -daemon -pidfile ..." will <br>> not require me to use "startproc" or "nohup fbguard ... &" as I need to <br>> do now if I wish to use fbguard directly. <br><br>-pidfile is required to support multi-instancing. Imagine you have 3 <br>instances, all are running process fbserver. How does startup script know, <br>what to shutdown? It keeps PID of the process in file with known name <br>(different for each instance) and when needed does something like <br>kill `cat $PidFileName` <br>to stop particular instance. It's not directly related with use of <br>distro-specific commands like "startproc". And if one wants to run fbguard <br>directly (not recommended, better use "/etc/init.d/firebird start"), that <br>switch is sooner of all not needed. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26628397 Re: Remove not used system fields 2009-12-03T07:58:42Z 2009-12-03T07:58:42Z Ann W. Harrison Adriano dos Santos Fernandes wrote: <br><div class='shrinkable-quote'><br>> GDML BLR references only fields used in the query, so this is not GDML <br>> specific problem. <br>> <br>> There are two situations: <br>> 1) SELECT * in SQL <br>> 2) Referencing a field in GDML or SQL <br>> <br>> Removing them does not direct break (1), but depends on what the <br>> application do. <br>> <br>> IMO, both cases could be considered as not good written application <br>> reading non-useful fields. </div><br>Document them as unused, but don't remove them!  An unused field <br>occupies next to no space, so carrying them around is a very low <br>cost way to maintain compatibility with old versions of gbak - <br>remember the system tables are hard coded there - and other tools. <br><br>Don't break things just because they're not coded to your standards. <br><br>Regards, <br><br>Ann <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26628311 Re: Internal identifier encoding 2009-12-03T07:53:17Z 2009-12-03T07:53:17Z Ann W. Harrison Claudio Valderrama C. wrote: <br>> <br>> I seem to remember that Ann once said ASCII exception messages had to do <br>> with the way BLR works with them, but it would be easier that Ann explains <br>> it again than having me looking into archives with ancient messages. <br>> <br><br>Your memory is amazing.  There's a discussion on 9/9/01 among Dave <br>Schnepper, Sean Lyne, and me with the subject "Transliteration error" <br><br>Dave did the original character set work for InterBase at <br>Borland.  Here's his message... the rest are in the archives. <br>The complaint he addresses is included at the bottom. <br><br><br>  Non-ASCII characters have never been supported in meta-data. <br><br>  Yes, the character set for (almost) all meta data names is <br>  UNICODE_FSS, but that was done when we were making massive ODS changes <br>  (v4) - and since eventually we would be doing it, it seems like a <br>  "good idea" (TM) at the time. <br><br>  Reasons why non-ASCII wasn't supported: <br>  a) We were out of time in the v4.0 schedule -- after that it never had <br>  support from marketing (eg: not enough customers complained about the <br>  lack). <br>  b) In those days, I was really hung up on making sure the product was <br>  polyglot.  And I couldn't think of a good solution to the access and <br>  extract problem. <br>     Access:  Any user, using any character set, should be able to <br>          reference any metadata object. <br>     Extract: You should be able to extract the metadata for an entire <br>         database, regardless of current character set, and execute it to <br>         recreate the database schema. <br><br>  Now, in retrospect (hindsight 9 years later is great, eh?) - I should <br>  have gone ahead with mono-language support -- and leave the polyglot <br>  issues as "bugs", or misfeatures. <br><br>  Now, as for why your statements aren't working now. <br><br>  I'll have to assume someone fixed the first line of defense -- the SQL <br>  Parser didn't allow non-ASCII to be part of a SYMBOL.  You're getting <br>  a character set transliteration error, not a unrecognized token error. <br><br>  The 2nd implementation problem was having the sqlparser know the <br>  attachment character set - so it could transliterate into Unicode. <br>  All that information is stored in the dbb, which the dsql parser <br>  didn't have access to.  Getting from the engine to dsql took (at one <br>  time) about 8 layers of function calls -- so that info has to be <br>  passed along through each layer. <br><br>  (problem 2a: How does DSQL get back to the engine to access the intl <br>   modules to perform the transliteration). <br><br>  (aside: for years there was talk of binding dsql into the engine <br>  better, so there was a single metadata cache, for instance -- and all <br>  this work would have been easier with dsql in the engine -- but I <br>  digress...) <br><br>  3rd problem: DYN.  How does dsql submit symbols back to the engine? <br>   There were 4 possibilities I came up with: <br>      a) redefine each existing metadata symbol dyn so the following <br>   sequence of bytes defined a symbol in UNICODE_FSS (instead of NONE). <br>      b) Allow an escape symbol at start of symbol to define the <br>   character set. <br>      c) Make new dyn for each item. <br>      d) Make an "escape" dyn.  eg:   dyn_charsetis <charsetid> <br>   dyn_relation_name <relation name in charset charsetid> <br><br>   c is almost too ugly to mention - I think there were 50 or so dyn that <br>   defined symbols (forgive me, this is all from memory, I haven't gone <br>   to check the sources... don't byte if there are only 49! or 10 or 1!) <br><br>   a & b both had backwards compatibility problems <br>   d isn't particularly elegant. <br><br>Anyway, I had sketched out an approach - but "the project never got <br>funded"... <br><br>To be more specific to the problem below: <br><br>  When dyn grabs a symbol off of the input dyn stream, it's putting it <br>   nto a DSC with charset ASCII.  Ascii does not define the upper 8 bit <br>  characters, so when the engine goes to transliterate from Ascii to <br>  Unicode it complains. <br><br>  Dave <br><br><br> > -----Original Message----- <br> > From: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26628311&i=0" target="_top" rel="nofollow">firebird-devel-admin@...</a> <br> > [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26628311&i=1" target="_top" rel="nofollow">firebird-devel-admin@...</a>]On <br> > Behalf Of Sergey Mereutsa <br> > Sent: Saturday, September 08, 2001 2:25 AM <br> > To: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26628311&i=2" target="_top" rel="nofollow">firebird-devel@...</a> <br> > Subject: Re: [Firebird-devel] Transliteration error <br> > <br> > <br> > Hi, Claudio ! <br> > I`v got the following result: <br> > <br> > [root@virgo db]# isql -user sysdba -password masterkey ibtest.gdb <br> > Database:  ibtest.gdb, User: sysdba <br> > SQL> create table "K?L"(a int); <br> > Statement failed, SQLCODE = -607 <br> > <br> > unsuccessful metadata update <br> > -STORE RDB$USER_PRIVILEGES failed defining a table <br> > -arithmetic exception, numeric overflow, or string truncation <br> > -Cannot transliterate character between character sets <br> > SQL> show version; <br> > ISQL Version: LI-V6.0.0.624 <br> > InterBase/linux Intel (access method), version "LI-V6.0.0.624" <br> > InterBase/linux Intel (remote server), version "LI-V6.0.0.624/tcp <br> > (virgo.protv.md)/P10" <br> > InterBase/linux Intel (remote interface), version "LI-V6.0.0.624/tcp <br> > (virgo.protv.md)/P10" <br> > on disk structure version 10.0 <br> > SQL> <br> > <br> > The same situation when I try to create table with cyrillic <br> > letters (NOTE I do <br> > not indicated NAMES or default charset for database): <br> > <br> > SQL>  create table "????"(a int); <br> > Statement failed, SQLCODE = -607 <br> > <br> > unsuccessful metadata update <br> > -STORE RDB$USER_PRIVILEGES failed defining a table <br> > -arithmetic exception, numeric overflow, or string truncation <br> > -Cannot transliterate character between character sets <br> > SQL> <br> > <br> > Best regards, Serj. <br> > <br><br>I didn't find anything specific to exceptions messages.  And, <br>unless I'm very mistaken, we did do some translation of error <br>messages that used eight bit characters at one time in InterBase. <br>And, in the very dark ages (1991) TwinSun, a Japanese company <br>working with Ashton-Tate, did a Japanses version with Kanji <br>metadata and errors. <br><br>So, that's what I've found and remember <br><br><br>Cheers, <br><br>Ann <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26627862 Re: Remove not used system fields 2009-12-03T07:29:53Z 2009-12-03T07:29:53Z Paul Beach <div class='shrinkable-quote'>>> Are we going to remove system fields supposed to never being used? <br>>> <br>>> For example: <br>>> <br>>> RDB$COLLATIONS <br>>> - RDB$FUNCION_NAME <br>>> <br>>> RDB$CHARACTER_SETS <br>>> - RDB$FORM_OF_USE <br>>> - RDB$FUNCION_NAME </div><br><br>Are you sure this is necessary - its liley to break a lot of tools/applications <br>that access the metadata <br><br>Paul Beach <br>Tel (France): +33 (0) 2 47 58 30 43 <br>Mob (France): +33 (0) 6 79 24 32 32 <br><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26627750 Fbguard - was FB3 and ... 2009-12-03T07:22:00Z 2009-12-03T07:22:00Z Dunbar, Norman <br>Afternoon all, <br><br>following on from discussions about the deprecation of fbmgr from 2.1 <br>and it's eventual replacement by fbguard, I have a couple of questions. <br><br>When will fbguard take over all the duties of fbmgr - version 3.x or <br>before that in 2.5? <br><br>Will (does) fbguard have an option to stop the database engine as in <br>fbmgr -shut? I'm running 2.0 on Windows and 2.1.3 on Linux and in the <br>Linux version at least, it doesn't do anything to stop the engine. <br><br>What will the -daemon option (from 2.5 on I believe) do? I assume it <br>will start the engine as a daemon. (Pretty obvious really, but best to <br>be sure). I also assume that running "fbguard -daemon -pidfile ..." will <br>not require me to use "startproc" or "nohup fbguard ... &" as I need to <br>do now if I wish to use fbguard directly. <br><br>I'm writing the fbguard manual and I need to know this information. <br><br><br>Thanks. <br>Norman. <br><br><br>Norman Dunbar <br>Dunbar IT Consultants Ltd <br>Tel: 0779 3292 984 <br>Tel: 0773 4531 439 <br><br>Registered address: <br>Thorpe House <br>61 Richardshaw Lane <br>Pudsey <br>West Yorkshire <br>United Kingdom <br>LS28 7EL <br><br>Company Number: 05132767 <br><br><br>Information in this message may be confidential and may be legally privileged. If you have received this message by mistake, please notify the sender immediately, delete it and do not copy it to anyone else.   We have checked this email and its attachments for viruses. But you should still check any attachment before opening it. We may have to make this message and any reply to it public if asked to under the Freedom of Information Act, Data Protection Act or for litigation.  Email messages and attachments sent to or from any Environment Agency address may also be accessed by someone other than the sender or recipient, for business purposes.  If we have sent you information and you wish to use it please read our terms and conditions which you can get by calling us on 08708 506 506.  Find out more about the Environment Agency at www.environment-agency.gov.uk <br><br>Information in this message may be confidential and may be legally privileged. If you have received this message by mistake, please notify the sender immediately, delete it and do not copy it to anyone else. <br><br>We have checked this email and its attachments for viruses. But you should still check any attachment before opening it. <br>We may have to make this message and any reply to it public if asked to under the Freedom of Information Act, Data Protection Act or for litigation.  Email messages and attachments sent to or from any Environment Agency address may also be accessed by someone other than the sender or recipient, for business purposes. <br><br>If we have sent you information and you wish to use it please read our terms and conditions which you can get by calling us on 08708 506 506.  Find out more about the Environment Agency at www.environment-agency.gov.uk <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26626768 GPRE -raw 2009-12-03T06:26:25Z 2009-12-03T06:26:25Z Adriano dos Santos Fernandes-3 Is anyone against removing -raw from the GPRE calls of our build scripts? <br><br>It's recurrent that I need to remove it to analyze generated BLR... <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26625594 Re: Remove not used system fields 2009-12-03T05:00:55Z 2009-12-03T05:00:55Z Adriano dos Santos Fernandes-3 Claudio Valderrama C. escreveu: <div class='shrinkable-quote'><br>>> -----Original Message----- <br>>> From: Adriano dos Santos Fernandes [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26625594&i=0" target="_top" rel="nofollow">adrianosf@...</a>] <br>>> Sent: Jueves, 03 de Diciembre de 2009 8:28 <br>>> <br>>> Are we going to remove system fields supposed to never being used? <br>>> <br>>> For example: <br>>> <br>>> RDB$COLLATIONS <br>>> - RDB$FUNCION_NAME <br>>> <br>>> RDB$CHARACTER_SETS <br>>> - RDB$FORM_OF_USE <br>>> - RDB$FUNCION_NAME <br>>>     <br>> <br>> Are these the only ones you want to remove or only an example? <br>>   </div>They are just an example. I haven't did a full look. <br><br>> The fields support user extensions that never were implemented nor <br>> documented. <br>> <br>> Now, about other fields, I think it's impossible to get rid of them: there <br>> could be some GDML based program that includes them. <br>>   <br>GDML BLR references only fields used in the query, so this is not GDML <br>specific problem. <br><br>There are two situations: <br>1) SELECT * in SQL <br>2) Referencing a field in GDML or SQL <br><br>Removing them does not direct break (1), but depends on what the <br>application do. <br><br>IMO, both cases could be considered as not good written application <br>reading non-useful fields. <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26625490 Re: Remove not used system fields 2009-12-03T04:51:22Z 2009-12-03T04:51:22Z Claudio Valderrama C. <div class='shrinkable-quote'>> -----Original Message----- <br>> From: Adriano dos Santos Fernandes [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26625490&i=0" target="_top" rel="nofollow">adrianosf@...</a>] <br>> Sent: Jueves, 03 de Diciembre de 2009 8:28 <br>> <br>> Are we going to remove system fields supposed to never being used? <br>> <br>> For example: <br>> <br>> RDB$COLLATIONS <br>> - RDB$FUNCION_NAME <br>> <br>> RDB$CHARACTER_SETS <br>> - RDB$FORM_OF_USE <br>> - RDB$FUNCION_NAME </div><br>Are these the only ones you want to remove or only an example? <br>The fields support user extensions that never were implemented nor <br>documented. <br><br>Now, about other fields, I think it's impossible to get rid of them: there <br>could be some GDML based program that includes them. <br><br>But I agree we have several fields in different tables that currently serve <br>no purpose. If we can be sure some fields are related to pyxis only, I would <br>wipe them out. <br><br>C. <br><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26624606 Re: Remove not used system fields 2009-12-03T03:34:41Z 2009-12-03T03:34:41Z Roman Rokytskyy-2 <div class='shrinkable-quote'>> Are we going to remove system fields supposed to never being used? <br>> <br>> For example: <br>> <br>> RDB$COLLATIONS <br>> - RDB$FUNCION_NAME <br>> <br>> RDB$CHARACTER_SETS <br>> - RDB$FORM_OF_USE <br>> - RDB$FUNCION_NAME </div><br>Better not. If some tool included them in the select list, it will <br>suddenly stop working. We could deprecate them, but removing them, I <br>think, is not good idea. <br><br>Roman <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26624482 Remove not used system fields 2009-12-03T03:26:20Z 2009-12-03T03:26:20Z Adriano dos Santos Fernandes-3 Are we going to remove system fields supposed to never being used? <br><br>For example: <br><br>RDB$COLLATIONS <br>- RDB$FUNCION_NAME <br><br>RDB$CHARACTER_SETS <br>- RDB$FORM_OF_USE <br>- RDB$FUNCION_NAME <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26624409 Re: Backporting data 2009-12-03T03:18:13Z 2009-12-03T03:18:13Z Dmitry Yemanov Claudio, <br><br> > There seems to be a developers gathering in Eris <br> > <a href="http://en.wikipedia.org/wiki/Eris_%28dwarf_planet%29" target="_top" rel="nofollow">http://en.wikipedia.org/wiki/Eris_%28dwarf_planet%29</a><br> > because even my question about changing the ODS for <br> > the system flag went unanswered :-) <br><br>It's hard to answer something you don't see ;-) The newsserver is <br>offline and it's not a big pleasure to follow the SF archives... <br><br><br>Dmitry <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26624365 Re: Sys flag 2009-12-03T03:15:07Z 2009-12-03T03:15:07Z Dmitry Yemanov Hi Claudio, <br><br> > Hello, do people have patience to go through another full build? <br><br>Why not? ;-) <br><br> > I made rdb$system_flag not null as I said some time ago. <br><br>No objections here. <br><br><br>Dmitry <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26622207 Firebird_ODBC_2.0.0.148 and ASP on 2003 server 2009-12-02T23:57:07Z 2009-12-02T23:57:07Z tom interface <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META content="text/html; charset=iso-8859-1" http-equiv=Content-Type> <META name=GENERATOR content="MSHTML 8.00.6001.18852"> </HEAD> <BODY bgColor=#ffffff> <DIV><FONT size=2 face=Arial>Hi,</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>i'm french so excuse me for my english.</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>I have a classic ASP application with Firebird server (1.5.5.4926) with ODBC 2.0.0.148</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>There's only one server for IIS and Firebird and i've 60 users, and each have his connection string (so i can't use pooling)</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>My connection string is : </FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>"Driver=Firebird/InterBase(r) driver;Uid=mylogin;Pwd=mypass;DbName=127.0.0.1:d:\database\xxxxxx.fdb"</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>Sometimes, when the connection is opened i have <FONT size=3 face="Times New Roman">intermittent 800a01fb error and by day it represents hundreds of error 500. I look in the IIS log and when i try the page that cause error, no problem for me. Users told me that when they refresh sometimes it works sometimes not. </FONT></FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>I'm using lots of response.redirect but always close and set nothing for connection and recordset before. </FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>I use debugdiag and have always this message (sometimes very quickly when i open connection to firebird): </FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV>In w3wp__PID__2620__Date__12_02_2009__Time_01_19_58PM__499__First Chance Access Violation.dmp the assembly instruction at <B>OdbcFb!SQLSetDescFieldW+5478</B> in <B>C:\WINDOWS\system32\OdbcFb.dll</B> from <B>IBPhoenix Inc</B> has caused an <B>access violation exception (0xC0000005)</B> when trying to <B>read from</B></B> memory location <B>0x0c641000</B> on thread <A href="mhtml:file://C:\Program Files\DebugDiag\Reports\CrashHang_Report__PID_2620__PID_5852(1_dumps)__120220091506084.mht#2620:25382Thread4468" target="_top" rel="nofollow"><B>67</B></A></DIV> <DIV> </DIV> <DIV><FONT size=2 face=Arial>I look for a solution for a long time but can't find a solution to solve this </FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>Can anyone help me ?? security rights problem ? Too many connections ?</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>I hope someone can help</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>Thanks</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV> <DIV><FONT size=2 face=Arial>Tom</FONT></DIV> <DIV><FONT size=2 face=Arial></FONT> </DIV><BR> <BR> __________ Information from ESET Smart Security, version of virus signature database 4656 (20091202) __________<BR> <BR> The message was checked by ESET Smart Security.<BR> <BR> <A HREF="http://www.eset.com" target="_top" rel="nofollow">http://www.eset.com</A><BR> </BODY></HTML> <br />------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br />_______________________________________________ <br>Firebird-odbc-devel mailing list <br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26622207&i=0" target="_top" rel="nofollow">Firebird-odbc-devel@...</a> <br><a href="https://lists.sourceforge.net/lists/listinfo/firebird-odbc-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-odbc-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-odbc-devel-f4398.html" embed="fixTarget[4398]" target="_top" >firebird-odbc-devel</a></p> tag:old.nabble.com,2006:post-26621194 Re: Backporting data 2009-12-02T21:20:17Z 2009-12-02T21:20:17Z Claudio Valderrama C. > -----Original Message----- <br>> From: Lester Caine [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26621194&i=0" target="_top" rel="nofollow">lester@...</a>] <br>> Sent: Miércoles, 02 de Diciembre de 2009 12:28 <br><br>> <br>> No comments? <br>> Ann HAS now confirmed that rather than being an 'unsupported <br>> bodge' it was <br>> actually by design from Interbase1 and I have now found the <br>> notes about in in <br>> Interbase5 days. It was the recommended route back a version .... <br><br>I was going to reply but I'm still stuck in some debugging (I have something <br>to say because I changed gbak many times looking for downgrade paths). <br><br>There seems to be a developers gathering in Eris <br><a href="http://en.wikipedia.org/wiki/Eris_%28dwarf_planet%29" target="_top" rel="nofollow">http://en.wikipedia.org/wiki/Eris_%28dwarf_planet%29</a><br>because even my question about changing the ODS for the system flag went <br>unanswered. <br>:-) <br><br>C. <br><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26620930 Re: Internal identifier encoding 2009-12-02T20:49:43Z 2009-12-02T20:49:43Z Claudio Valderrama C. > -----Original Message----- <br>> From: Adriano dos Santos Fernandes [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26620930&i=0" target="_top" rel="nofollow">adrianosf@...</a>] <br>> Sent: Miércoles, 02 de Diciembre de 2009 11:47 <br><br>> that <br>> reminds the exception text being stored as NONE. So we can <br>> also discuss <br>> how to change that, with may involve some new GBAK option to <br>> understand <br>> the encoding of this field and/or replacement of non-ascii <br>> characters to <br>> question marks. <br><br>I seem to remember that Ann once said ASCII exception messages had to do <br>with the way BLR works with them, but it would be easier that Ann explains <br>it again than having me looking into archives with ancient messages. <br><br>C. <br><br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26612978 Re: Вопрос о сертификации Firebird 2009-12-02T09:33:54Z 2009-12-02T09:33:54Z Алексей Вишняков Думаю, что очень вряд ли, поскольку для того, чтобы его сертифицировать, надо как минимум иметь право на производство средств защиты информации, а это как минимум лицензия ФСТЭК на ТЗКИ (техническую защиту конфиденциальной информации), а то и на производство средств шифрования, на их обслуживание (лицензии ФСБ) и т.п., а это, как вы, наверное, догадываетесь, весьма и весьма недешево.<br> <br>Вишнякова Татьяна.<br><br>2 декабря 2009 г. 18:37 пользователь Dmitri Kuzmenko <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26612978&i=0" target="_top" rel="nofollow">kdv@...</a>> написал:<br>><br>> Hello, plasmorf!<br>><br>> plasmorf wrote<br>><br> >> В свете нового закона "о персональных данных" (152 ФЗ) возник вопрос<br>>> к разработчикам - будет ли проводится сертификация Firebird по какому-<br>>> либо классу защищенности и доступа? (Например как oracle) Или ну его<br> >> нах?<br>><br>> 1. не будет<br>> 2. "как оракл" - я бы советовал ознакомиться со списком сертифицированных<br>> СУБД. Оракл там представлен конкретными версиями, причем очень скудным<br>> списком.<br> ><br>><br>> --<br>> Dmitri Kouzmenko, <a href="http://www.ibase.ru" target="_top" rel="nofollow">www.ibase.ru</a>, (495) 953-13-34<br>><br>><br>><br><br><br><br>-- <br>-- <br>Norritt, mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=26612978&i=1" target="_top" rel="nofollow">NorrittMobile@...</a><br> <br> <p>From forum: <a href="http://old.nabble.com/firebird-russian-f18839.html" embed="fixTarget[18839]" target="_top" >firebird-russian</a></p> tag:old.nabble.com,2006:post-26628796 [FB-Tracker] Reopened: (CORE-2785) Transliteration error with COMMENT statement 2009-12-02T08:09:34Z 2009-12-02T08:09:34Z JIRA tracker@firebirdsql.org <br>     [ <a href="http://tracker.firebirdsql.org/browse/CORE-2785?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel" target="_top" rel="nofollow">http://tracker.firebirdsql.org/browse/CORE-2785?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel</a> ] <br><br>Adriano dos Santos Fernandes reopened CORE-2785: <br>------------------------------------------------ <br><br><br>This is not direct related to COMMENT ON. It's a blob transliteration problem in a specific case. <br><div class='shrinkable-quote'><br>> Transliteration error with COMMENT statement <br>> -------------------------------------------- <br>> <br>>                 Key: CORE-2785 <br>>                 URL: <a href="http://tracker.firebirdsql.org/browse/CORE-2785" target="_top" rel="nofollow">http://tracker.firebirdsql.org/browse/CORE-2785</a><br>>             Project: Firebird Core <br>>          Issue Type: Bug <br>>          Components: Charsets/Collation <br>>    Affects Versions: 2.5 Beta 2, 2.1.3, 2.5 RC1 <br>>         Environment: Windows, IBExpert, Client CharSet UTF8 <br>>            Reporter: GR <br>>            Assignee: Adriano dos Santos Fernandes <br>>         Attachments: 2785.sql, db.sql <br>> <br>> <br>> Transliteration error occurs with statement <br>> COMMENT ON table T IS <br>> 'параметры:                                          количество <br>>              если больше 0 подгоняет сумму товаров под данное значение пропорционально себестоимости <br>>                если меньше 0, на выходе в GSUMMA будет себестоимость за 1 ед. <br>>                если равно 0, делает только разложение на компоненты <br>>   Результат'; <br>> -- <br>> This operation is not defined for system tables. <br>> unsuccessful metadata update. <br>> Create metadata BLOB failed. <br>> Cannot transliterate character between character sets. <br>> -- <br>> Reproduces with IBExpert. With ISQL - or hangs up, or gets error too: <br>> Statement failed, SQLCODE = -607 <br>> unsuccessful metadata update <br>> -Create metadata BLOB failed <br>> -Cannot transliterate character between character sets <br>> With CharSet WIN1251 works fine. </div><br>-- <br>This message is automatically generated by JIRA. <br>- <br>If you think it was sent incorrectly contact one of the administrators: <a href="http://tracker.firebirdsql.org/secure/Administrators.jspa" target="_top" rel="nofollow">http://tracker.firebirdsql.org/secure/Administrators.jspa</a><br>- <br>For more information on JIRA, see: <a href="http://www.atlassian.com/software/jira" target="_top" rel="nofollow">http://www.atlassian.com/software/jira</a><br><br>        <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610789 Re: Backporting data 2009-12-02T07:27:32Z 2009-12-02T07:27:32Z Lester Caine Lester Caine wrote: <div class='shrinkable-quote'><br>> OK <br>> <br>> Having now been informed that the method I rely on for managing older sites <br>> until they can be safely upgrade is just an unsupported bodge. I'd like to stick <br>> my head up and ask if there is ANY chance that a 'supported' alternative can be <br>> provided. <br>> <br>> I have a number of sites currently that have FB1.5 on, and it will be some time <br>> before the machines are due for renewal, which is the only time they will be <br>> upgraded. FB2.1.2 is my current upgrade point having skipped FB2.0 and all the <br>> machines here run that, but I have a FB1.5 gbak on the system so that I can pull <br>> data from site - play with it - and then in an emergency - push it back to site. <br>> <br>> We are not talking about 'new features' since the database schema is always the <br>> older version, but is there anything stopping a version of gbak being made <br>> available which WOULD handle a backport from FB2.5 to something earlier? <br>> <br>> Personally I will not move forward from FB2.1.x until all of the sites are up to <br>> that level, and we will probably have FB3 by then, but this 'legacy bodge' has <br>> proven very useful in the past, and would be a convenient tool in the support <br>> arsenal? </div><br>No comments? <br>Ann HAS now confirmed that rather than being an 'unsupported bodge' it was <br>actually by design from Interbase1 and I have now found the notes about in in <br>Interbase5 days. It was the recommended route back a version .... <br><br>-- <br>Lester Caine - G8HFL <br>----------------------------- <br>Contact - <a href="http://lsces.co.uk/wiki/?page=contact" target="_top" rel="nofollow">http://lsces.co.uk/wiki/?page=contact</a><br>L.S.Caine Electronic Services - <a href="http://lsces.co.uk" target="_top" rel="nofollow">http://lsces.co.uk</a><br>EnquirySolve - <a href="http://enquirysolve.com/" target="_top" rel="nofollow">http://enquirysolve.com/</a><br>Model Engineers Digital Workshop - <a href="http://medw.co.uk//" target="_top" rel="nofollow">http://medw.co.uk//</a><br>Firebird - <a href="http://www.firebirdsql.org/index.php" target="_top" rel="nofollow">http://www.firebirdsql.org/index.php</a><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610712 Re: Testing Firebird-2.1.3 64bit on SPARC 2009-12-02T07:25:30Z 2009-12-02T07:25:30Z Adriano dos Santos Fernandes-3 Alexander Peshkoff escreveu: <div class='shrinkable-quote'><br>> On Wednesday 02 December 2009 17:41:58 Adriano dos Santos Fernandes wrote: <br>>   <br>>> Is WORDS_BIGENDIAN correctly defined? <br>>> <br>>> If not, sort (in opt.cpp) does not align the data. Alex, should not this <br>>> be changed to RISC_ALIGNMENT? <br>>>     <br>> <br>> It's in rse.cpp. And it has no WORDS_BIGENDIAN (or RISC_ALIGNMENT). <br>> Something makes me think it may be related with <br>>   </div>But it's in opt.cpp that the offsets of the "sort message" is setup. <br><br>> case SMB_DBKEY: <br>>       rpb->rpb_number.setValue(*reinterpret_cast<SINT64*>(from.dsc_address)); <br>>       break; <br>> <br>> There were some DBKEY alignment fixes in 2.5. But certainly I can't be sure:) <br>> <br>> BTW, on sparc both RISC_ALIGNMENT and WORDS_BIGENDIAN are defined. And can you <br>> be more exact - what place in opt.cpp seems problematic to you? This can be <br>> important for ia64. <br>Search for "#ifndef WORDS_BIGENDIAN". There are two. <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610556 Re: Testing Firebird-2.1.3 64bit on SPARC 2009-12-02T07:16:00Z 2009-12-02T07:16:00Z Alexander Peshkoff On Wednesday 02 December 2009 17:41:58 Adriano dos Santos Fernandes wrote: <br>> Is WORDS_BIGENDIAN correctly defined? <br>> <br>> If not, sort (in opt.cpp) does not align the data. Alex, should not this <br>> be changed to RISC_ALIGNMENT? <br><br>It's in rse.cpp. And it has no WORDS_BIGENDIAN (or RISC_ALIGNMENT). <br>Something makes me think it may be related with <br><br>case SMB_DBKEY: <br>      rpb->rpb_number.setValue(*reinterpret_cast<SINT64*>(from.dsc_address)); <br>      break; <br><br>There were some DBKEY alignment fixes in 2.5. But certainly I can't be sure:) <br><br>BTW, on sparc both RISC_ALIGNMENT and WORDS_BIGENDIAN are defined. And can you <br>be more exact - what place in opt.cpp seems problematic to you? This can be <br>important for ia64. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610467 Re: Internal identifier encoding 2009-12-02T07:10:48Z 2009-12-02T07:10:48Z Alexander Peshkoff On Wednesday 02 December 2009 17:48:17 Adriano dos Santos Fernandes wrote: <br>> Alexander Peshkoff escreveu: <br>> > context.putSpecific() can throw an error on some platforms in case of <br>> > system call failure. <br>> <br>> Bad. <br><br>The only 'good' case is when __thread attribute can be used. But it's <br>supported not at all platforms :-( <br><br>> > This is ceratinyl extremely rare case, but looks like netsed <br>> > try is needed. <br>> <br>> Or maybe some new RAII class could do the job too. <br><br>Maybe.,. <br>Though I have no idea what should it contain. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610407 Re: Internal identifier encoding 2009-12-02T07:04:57Z 2009-12-02T07:04:57Z Dimitry Sibiryakov-3 >>    Our non-ASCII exceptions are already crap and it is reflected in tracker. <br>> Although you didn't tried to add something useful to discuss, that <br>> reminds the exception text being stored as NONE. So we can also discuss <br>> how to change that, with may involve some new GBAK option to understand <br>> the encoding of this field and/or replacement of non-ascii characters to <br>> question marks. <br>> <br>> Comments? <br><br>   To add something more useful than CORE-2431? No. I just can say that <br>whatever you do in this area, it will be better than current garbage in <br>messages. <br><br>   SY, SD. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610068 Re: Internal identifier encoding 2009-12-02T06:46:49Z 2009-12-02T06:46:49Z Adriano dos Santos Fernandes-3 Alexander Peshkoff escreveu: <br>> context.putSpecific() can throw an error on some platforms in case of system <br>> call failure. <br>Bad. <br><br>> This is ceratinyl extremely rare case, but looks like netsed <br>> try is needed. <br>Or maybe some new RAII class could do the job too. <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610045 Re: Internal identifier encoding 2009-12-02T06:44:57Z 2009-12-02T06:44:57Z Adriano dos Santos Fernandes-3 Dimitry Sibiryakov escreveu: <br>>> This make the problem visible in other place: identifiers in exceptions. <br>>> But this is easy to fix, as our exception texts are already ASCII-based <br>>> (or we treat as UTF, considering translations). Exception strings should <br>>> be converted to user charset. <br>>> <br>>> Comments? <br>>>     <br>> <br>>    Our non-ASCII exceptions are already crap and it is reflected in tracker. <br>Although you didn't tried to add something useful to discuss, that <br>reminds the exception text being stored as NONE. So we can also discuss <br>how to change that, with may involve some new GBAK option to understand <br>the encoding of this field and/or replacement of non-ascii characters to <br>question marks. <br><br>Comments? <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26610003 Re: Internal identifier encoding 2009-12-02T06:42:14Z 2009-12-02T06:42:14Z Alexander Peshkoff On Wednesday 02 December 2009 17:34:08 Adriano dos Santos Fernandes wrote: <div class='shrinkable-quote'><br>> Hi! <br>> <br>> JRD operates with connection charset UNICODE_FSS, hence have its cache <br>> using this encoding. <br>> <br>> DSQL operates with user charset. This cause a lot of problems and code <br>> efforts for conversions, and is not future proof, as one day there will <br>> be no DSQL cache. <br>> <br>> And there is new problem in v3. External routines may operate with <br>> different character set. That would mix names encoded different in the <br>> DSQL metadata cache. <br>> <br>> It seems I have good solution for the problem. The lexer automatically <br>> convert (double-quoted) identifiers from user charset to metadata <br>> charset. So DSQL automatically starts to operate with metadata charset. <br>> Then, some conversions (for DYN, DdlNodes, PackageNodes) are removed. <br>> <br>> This make the problem visible in other place: identifiers in exceptions. <br>> But this is easy to fix, as our exception texts are already ASCII-based <br>> (or we treat as UTF, considering translations). Exception strings should <br>> be converted to user charset. <br>> <br>> Comments? <br>> <br>> To do that conversion, looks like better place is in engine entrypoints <br>> in jrd. An access to tdbb is needed, so is it a problem [it currently <br>> does: context.putSpecific()] to move its contruction to outside the "try"? <br>> <br>>     ThreadContextHolder tdbb(user_status); <br>>     try <br>>     { <br>>        ... <br>>     } <br>>     catch (const Exception& ex) <br>>     { <br>>         return abnormalCompletion(tdbb, ex, user_status); <br>>     } <br>> <br>>     return successful_completion(user_status); <br>> <br>> Otherwise, we'll need to nest another try. </div><br>context.putSpecific() can throw an error on some platforms in case of system <br>call failure. This is ceratinyl extremely rare case, but looks like netsed <br>try is needed. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26609970 Re: Testing Firebird-2.1.3 64bit on SPARC 2009-12-02T06:40:15Z 2009-12-02T06:40:15Z Adriano dos Santos Fernandes-3 Is WORDS_BIGENDIAN correctly defined? <br><br>If not, sort (in opt.cpp) does not align the data. Alex, should not this <br>be changed to RISC_ALIGNMENT? <br><br><br>Adriano <br><br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p> tag:old.nabble.com,2006:post-26609942 Re: Internal identifier encoding 2009-12-02T06:36:16Z 2009-12-02T06:36:16Z Dimitry Sibiryakov-3 > This make the problem visible in other place: identifiers in exceptions. <br>> But this is easy to fix, as our exception texts are already ASCII-based <br>> (or we treat as UTF, considering translations). Exception strings should <br>> be converted to user charset. <br>> <br>> Comments? <br><br>   Our non-ASCII exceptions are already crap and it is reflected in tracker. <br><br>   SY, SD. <br><br>------------------------------------------------------------------------------ <br>Join us December 9, 2009 for the Red Hat Virtual Experience, <br>a free event focused on virtualization and cloud computing. <br>Attend in-depth sessions from your desk. Your couch. Anywhere. <br><a href="http://p.sf.net/sfu/redhat-sfdev2dev" target="_top" rel="nofollow">http://p.sf.net/sfu/redhat-sfdev2dev</a><br>Firebird-Devel mailing list, web interface at <a href="https://lists.sourceforge.net/lists/listinfo/firebird-devel" target="_top" rel="nofollow">https://lists.sourceforge.net/lists/listinfo/firebird-devel</a><br><p>From forum: <a href="http://old.nabble.com/firebird-devel-f4395.html" embed="fixTarget[4395]" target="_top" >firebird-devel</a></p>