[Maria-developers] Progress (by Sanja): Avoid flushing keycache to disk in case of DROPping table (53)
--- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...: Avoid flushing keycache to disk in case of DROPping table CREATION DATE..: Thu, 27 Aug 2009, 11:12 SUPERVISOR.: Monty IMPLEMENTOR: Sanja COPIES TO..: CATEGORY...: Server-Sprint TASK ID: 53 (http://askmonty.org/worklog/?tid=53) VERSION: Server-5.0 STATUS.: Assigned PRIORITY...: 60 WORKED HOURS...: 13 ESTIMATE...: 11 (hours remain) ORIG. ESTIMATE.: 24 PROGRESS NOTES: -=-=(Sanja - Thu, 03 Sep 2009, 19:43)=-=- New patch and its testing Worked 8 hours and estimate 11 hours remain (original estimate unchanged). -=-=(Sanja - Tue, 01 Sep 2009, 15:20)=-=- moving to 5.0 part1 Worked 2 hours and estimate 19 hours remain (original estimate unchanged). -=-=(Sanja - Tue, 01 Sep 2009, 15:19)=-=- first patch Worked 3 hours and estimate 21 hours remain (original estimate unchanged). -=-=(Sanja - Thu, 27 Aug 2009, 11:15)=-=- High-Level Specification modified. --- /tmp/wklog.53.old.23490 2009-08-27 11:15:24.0 +0300 +++ /tmp/wklog.53.new.23490 2009-08-27 11:15:24.0 +0300 @@ -1 +1,3 @@ - +To make special handler call 'prepare to drop' which mark table as changed (in +case of server crash just before removing table it force table to be repaired) +then flush keycache without writing on disk. DESCRIPTION: When DROP TABLE executed keycache flushed to disk before removing table which is wasting of resources. So better to have table removed from keycache without writing changes on disk. HIGH-LEVEL SPECIFICATION: To make special handler call 'prepare to drop' which mark table as changed (in case of server crash just before removing table it force table to be repaired) then flush keycache without writing on disk. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE --- WorkLog (v3.5.9) ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
[Maria-developers] Progress (by Sanja): Avoid flushing keycache to disk in case of DROPping table (53)
--- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...: Avoid flushing keycache to disk in case of DROPping table CREATION DATE..: Thu, 27 Aug 2009, 11:12 SUPERVISOR.: Monty IMPLEMENTOR: Sanja COPIES TO..: CATEGORY...: Server-Sprint TASK ID: 53 (http://askmonty.org/worklog/?tid=53) VERSION: Server-5.0 STATUS.: Assigned PRIORITY...: 60 WORKED HOURS...: 13 ESTIMATE...: 11 (hours remain) ORIG. ESTIMATE.: 24 PROGRESS NOTES: -=-=(Sanja - Thu, 03 Sep 2009, 19:43)=-=- New patch and its testing Worked 8 hours and estimate 11 hours remain (original estimate unchanged). -=-=(Sanja - Tue, 01 Sep 2009, 15:20)=-=- moving to 5.0 part1 Worked 2 hours and estimate 19 hours remain (original estimate unchanged). -=-=(Sanja - Tue, 01 Sep 2009, 15:19)=-=- first patch Worked 3 hours and estimate 21 hours remain (original estimate unchanged). -=-=(Sanja - Thu, 27 Aug 2009, 11:15)=-=- High-Level Specification modified. --- /tmp/wklog.53.old.23490 2009-08-27 11:15:24.0 +0300 +++ /tmp/wklog.53.new.23490 2009-08-27 11:15:24.0 +0300 @@ -1 +1,3 @@ - +To make special handler call 'prepare to drop' which mark table as changed (in +case of server crash just before removing table it force table to be repaired) +then flush keycache without writing on disk. DESCRIPTION: When DROP TABLE executed keycache flushed to disk before removing table which is wasting of resources. So better to have table removed from keycache without writing changes on disk. HIGH-LEVEL SPECIFICATION: To make special handler call 'prepare to drop' which mark table as changed (in case of server crash just before removing table it force table to be repaired) then flush keycache without writing on disk. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE --- WorkLog (v3.5.9) ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
[Maria-developers] Rev 2720: Fix for in file:///Users/hakan/work/monty_program/maria/
At file:///Users/hakan/work/monty_program/maria/ revno: 2720 revision-id: ha...@askmonty.org-20090903163555-hds72helhcy5htqq parent: kniel...@knielsen-hq.org-20090805072137-wg97dcem1cxnzt3p committer: Hakan Kuecuekyilmaz branch nick: maria timestamp: Thu 2009-09-03 18:35:55 +0200 message: Fix for mysqlslap: setting --engine does not get replicated http://bugs.mysql.com/bug.php?id=46967 Problem was that a "set storage_engine = " is not replicated. Fixed by adding an explicit Engine = to the build_table_string() function. Open question: Should MySQL's replication actually replicate the set storage_engine command or not? Answer: no it should not (documented). === modified file 'client/mysqlslap.c' --- a/client/mysqlslap.c2009-04-25 10:05:32 + +++ b/client/mysqlslap.c2009-09-03 16:35:55 + @@ -880,6 +880,13 @@ } dynstr_append(&table_string, ")"); + + if (default_engine) + { +dynstr_append(&table_string, " Engine = "); +dynstr_append(&table_string, default_engine); + } + ptr= (statement *)my_malloc(sizeof(statement), MYF(MY_ZEROFILL|MY_FAE|MY_WME)); ptr->string = (char *)my_malloc(table_string.length+1, ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
On Thu, Sep 3, 2009 at 5:18 AM, Henrik Ingo wrote: > Hi Giuseppe > > I'm cross-posting to maria-discuss since this is now a non-technical > discussion. Please remove maria-developers when replying. > > On Thu, Sep 3, 2009 at 3:01 PM, Giuseppe Maxia wrote: >> Why it's perfectly OK that MariaDB gets MySQL bugfixes for free and >> instead it's a rip off if MySQL gets MariaDB bug fixes? > > Of course, all of the MariaDB source code is available as GPL, just > like for MySQL. Sun is not prohibited from using it under the GPL, > just like we are using MySQL code. I look forward to using improvements published by MariaDB. > >> MariaDB requires the same kind of legal grants that Sun asks, and the >> FAQ even states "The MCA is based on Sun's Contributor Agreement. We >> think it is one of the best contributor agreements around! " >> http://askmonty.org/wiki/index.php/MCA_FAQ > > We've made some adjustments, but yes, apparently it is one of the best > contributor agreements around, the community seems to think. That's > why we use it. Will MariaDB incorporate BSD-licensed features published by the community? -- Mark Callaghan mdcal...@gmail.com ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
[Maria-developers] Rev 2721: Fix for coding style. in file:///Users/hakan/work/monty_program/maria/
At file:///Users/hakan/work/monty_program/maria/ revno: 2721 revision-id: ha...@askmonty.org-20090903150639-y5s9y2ylowfesyr4 parent: ha...@askmonty.org-20090902214617-5aljfm0c229hzdpu committer: Hakan Kuecuekyilmaz branch nick: maria timestamp: Thu 2009-09-03 17:06:39 +0200 message: Fix for coding style. === modified file 'client/mysqlslap.c' --- a/client/mysqlslap.c2009-09-02 21:46:17 + +++ b/client/mysqlslap.c2009-09-03 15:06:39 + @@ -881,7 +881,8 @@ dynstr_append(&table_string, ")"); - if (default_engine) { + if (default_engine) + { dynstr_append(&table_string, " Engine = "); dynstr_append(&table_string, default_engine); } ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2726)
#At lp:maria 2726 kniel...@knielsen-hq.org 2009-09-03 Fix most Compiler warnings seen in buildbot. Add suppressions for a few warnings that cannot be meaningfully fixed by MariaDB developers. Changes for XtraDB, PBXT, and YaSSL also submitted upstream. Also add a `ccfilter` wrapper that can be used to filter out suppressed warnings in a local build (to check that new warnings are not introduced). added: support-files/ccfilter modified: client/mysqlbinlog.cc config/ac-macros/misc.m4 extra/yassl/include/yassl_int.hpp extra/yassl/src/handshake.cpp extra/yassl/src/yassl_imp.cpp extra/yassl/src/yassl_int.cpp extra/yassl/taocrypt/include/modes.hpp extra/yassl/taocrypt/src/asn.cpp mysys/my_compress.c sql/mysqld.cc sql/strfunc.cc storage/pbxt/src/discover_xt.cc storage/xtradb/fil/fil0fil.c storage/xtradb/mtr/mtr0mtr.c storage/xtradb/srv/srv0srv.c storage/xtradb/srv/srv0start.c strings/decimal.c support-files/compiler_warnings.supp per-file messages: client/mysqlbinlog.cc Fix compiler warnings. config/ac-macros/misc.m4 Fix wrong naming, autoconfig requires _cv_ in cached names. extra/yassl/include/yassl_int.hpp Fix compiler warnings. extra/yassl/src/handshake.cpp Fix compiler warnings. extra/yassl/src/yassl_imp.cpp Fix compiler warnings. extra/yassl/src/yassl_int.cpp Fix compiler warnings. extra/yassl/taocrypt/include/modes.hpp Fix compiler warnings. extra/yassl/taocrypt/src/asn.cpp Fix compiler warnings. mysys/my_compress.c Fix compiler warnings. sql/mysqld.cc Fix compiler warnings. sql/strfunc.cc Fix compiler warnings. storage/pbxt/src/discover_xt.cc Fix compiler warnings. storage/xtradb/fil/fil0fil.c Fix compiler warnings. storage/xtradb/mtr/mtr0mtr.c Fix compiler warnings. storage/xtradb/srv/srv0srv.c Fix compiler warnings. storage/xtradb/srv/srv0start.c Fix compiler warnings. strings/decimal.c Fix compiler warnings. support-files/ccfilter Add helper for suppressing compiler warnings in local developer source tree. Allows to check for not introducing new warnings into Buildbot without having to actually run the build through Buildbot. support-files/compiler_warnings.supp Suppress a few warnings that cannot be meaningfully fixed in source code. === modified file 'client/mysqlbinlog.cc' --- a/client/mysqlbinlog.cc 2009-02-13 16:41:47 + +++ b/client/mysqlbinlog.cc 2009-09-03 13:20:22 + @@ -689,8 +689,8 @@ Exit_status process_event(PRINT_EVENT_IN Format events are not concerned by --offset and such, we always need to read them to be able to process the wanted events. */ - if ((rec_count >= offset) && - ((my_time_t)(ev->when) >= start_datetime) || + if (((rec_count >= offset) && + ((my_time_t)(ev->when) >= start_datetime)) || (ev_type == FORMAT_DESCRIPTION_EVENT)) { if (ev_type != FORMAT_DESCRIPTION_EVENT) === modified file 'config/ac-macros/misc.m4' --- a/config/ac-macros/misc.m4 2008-02-26 17:38:43 + +++ b/config/ac-macros/misc.m4 2009-09-03 13:20:22 + @@ -594,15 +594,15 @@ dnl dnl MYSQL_NEEDS_MYSYS_NEW AC_DEFUN([MYSQL_NEEDS_MYSYS_NEW], -[AC_CACHE_CHECK([needs mysys_new helpers], mysql_use_mysys_new, +[AC_CACHE_CHECK([needs mysys_new helpers], mysql_cv_use_mysys_new, [ AC_LANG_PUSH(C++) AC_TRY_LINK([], [ class A { public: int b; }; A *a=new A; a->b=10; delete a; -], mysql_use_mysys_new=no, mysql_use_mysys_new=yes) +], mysql_cv_use_mysys_new=no, mysql_cv_use_mysys_new=yes) AC_LANG_POP(C++) ]) -if test "$mysql_use_mysys_new" = "yes" +if test "$mysql_cv_use_mysys_new" = "yes" then AC_DEFINE([USE_MYSYS_NEW], [1], [Needs to use mysys_new helpers]) fi === modified file 'extra/yassl/include/yassl_int.hpp' --- a/extra/yassl/include/yassl_int.hpp 2008-11-18 16:45:44 + +++ b/extra/yassl/include/yassl_int.hpp 2009-09-03 13:20:22 + @@ -441,7 +441,7 @@ public: const Ciphers&GetCiphers() const; const DH_Parms& GetDH_Parms() const; const Stats& GetStats()const; -const VerifyCallback getVerifyCallback() const; +VerifyCallback getVerifyCallback() const; pem_password_cb GetPasswordCb() const; void* GetUserData() const; boolGetSessionCacheOff() const; === modified file 'extra/yassl/src/handshake.cpp' --- a/extra/yassl/src/handshake.cpp 2008-04-28 16:24:05 + +++ b/extra/yassl/src/handshake.cpp 2009-09-03 13:20:22 + @@ -789,7 +789,7 @@ void processReply(SSL& ssl) { if (ssl.GetError()) return; -if (DoProcessReply(ssl)) +if (DoProcessReply(ssl)) { // didn't complete process if (!ssl.getSocket().IsNonBlocking()) {
[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2725)
#At lp:maria 2725 kniel...@knielsen-hq.org 2009-09-03 MBug#423035: error in parsing enum value for plugin variable in mysqld command-line option Fix parsing of invalid plugin enum option value. Previous patch to fix plugin enum option parsing on big-endian introduced another bug due to incorrect comparison of unsigned value. This would cause an incorrect value to be parsed as value 0. See also MySQL Bug#41010 and Bug#32034. added: mysql-test/r/mysqld_option_err.result mysql-test/t/mysqld_option_err.test modified: mysql-test/mysql-test-run.pl mysys/my_getopt.c per-file messages: mysql-test/mysql-test-run.pl Add a facility for test case to run the mysqld binary (to test that invalid startup options are rejected correctly). mysql-test/r/mysqld_option_err.result Add a test case to check that invalid startup options for mysqld are rejected. This is needed to test MBug#423035. Also add a few other similar tests, as this was completely untested before this patch. mysql-test/t/mysqld_option_err.test Add a test case to check that invalid startup options for mysqld are rejected. This is needed to test MBug#423035. Also add a few other similar tests, as this was completely untested before this patch. mysys/my_getopt.c Fix parsing of invalid plugin enum option value. === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-06-22 08:06:35 + +++ b/mysql-test/mysql-test-run.pl 2009-09-03 13:05:02 + @@ -1771,6 +1771,20 @@ sub tool_arguments ($$) { return mtr_args2str($exe, @$args); } +# This is not used to actually start a mysqld server, just to allow test +# scripts to run the mysqld binary to test invalid server startup options. +sub mysqld_client_arguments () { + my $default_mysqld= default_mysqld(); + my $exe = find_mysqld($basedir); + my $args; + mtr_init_args(\$args); + mtr_add_arg($args, "--no-defaults"); + mtr_add_arg($args, "--basedir=%s", $basedir); + mtr_add_arg($args, "--character-sets-dir=%s", $default_mysqld->value("character-sets-dir")); + mtr_add_arg($args, "--language=%s", $default_mysqld->value("language")); + return mtr_args2str($exe, @$args); +} + sub have_maria_support () { my $maria_var= $mysqld_variables{'maria'}; @@ -1968,6 +1982,7 @@ sub environment_setup { $ENV{'MYSQLADMIN'}= native_path($exe_mysqladmin); $ENV{'MYSQL_CLIENT_TEST'}=mysql_client_test_arguments(); $ENV{'MYSQL_FIX_SYSTEM_TABLES'}= mysql_fix_arguments(); + $ENV{'MYSQLD'}= mysqld_client_arguments(); $ENV{'EXE_MYSQL'}=$exe_mysql; # === added file 'mysql-test/r/mysqld_option_err.result' --- a/mysql-test/r/mysqld_option_err.result 1970-01-01 00:00:00 + +++ b/mysql-test/r/mysqld_option_err.result 2009-09-03 13:05:02 + @@ -0,0 +1,6 @@ +Test that unknown option is not silently ignored. +Test bad binlog format. +Test bad default storage engine. +Test non-numeric value passed to number option. +Test that bad value for plugin enum option is rejected correctly. +Done. === added file 'mysql-test/t/mysqld_option_err.test' --- a/mysql-test/t/mysqld_option_err.test 1970-01-01 00:00:00 + +++ b/mysql-test/t/mysqld_option_err.test 2009-09-03 13:05:02 + @@ -0,0 +1,47 @@ +# +# Test error checks on mysqld command line option parsing. +# +# Call mysqld with different invalid options, and check that it fails in each case. +# +# This means that a test failure results in mysqld starting up, which is only +# caught when the test case times out. This is not ideal, but I did not find an +# easy way to have the server shut down after a successful startup. +# + +--source include/not_embedded.inc + +# We have not run (and do not need) bootstrap of the server. We just +# give it a dummy data directory (for log files etc). + +mkdir $MYSQLTEST_VARDIR/tmp/mysqld_option_err; + + +--echo Test that unknown option is not silently ignored. +--error 2 +--exec $MYSQLD --skip-networking --datadir=$MYSQLTEST_VARDIR/tmp/mysqld_option_err --skip-grant-tables --nonexistentoption >$MYSQLTEST_VARDIR/tmp/mysqld_option_err/mysqltest.log 2>&1 + + +--echo Test bad binlog format. +--error 1 +--exec $MYSQLD --skip-networking --datadir=$MYSQLTEST_VARDIR/tmp/mysqld_option_err --skip-grant-tables --log-bin --binlog-format=badformat >>$MYSQLTEST_VARDIR/tmp/mysqld_option_err/mysqltest.log 2>&1 + + +--echo Test bad default storage engine. +--error 1 +--exec $MYSQLD --skip-networking --datadir=$MYSQLTEST_VARDIR/tmp/mysqld_option_err --skip-grant-tables --default-storage-engine=nonexistentengine >>$MYSQLTEST_VARDIR/tmp/mysqld_option_err/mysqltest.log 2>&1 + + +--echo Test non-numeric value passed to number option. +--error 1 +--exec $MYSQLD --skip-networking --datadir=$MYSQLTEST_VARDIR/tmp/mysqld_op
Re: [Maria-developers] MariaDB and the SCA
Hi Giuseppe I'm cross-posting to maria-discuss since this is now a non-technical discussion. Please remove maria-developers when replying. On Thu, Sep 3, 2009 at 3:01 PM, Giuseppe Maxia wrote: > Why it's perfectly OK that MariaDB gets MySQL bugfixes for free and > instead it's a rip off if MySQL gets MariaDB bug fixes? Of course, all of the MariaDB source code is available as GPL, just like for MySQL. Sun is not prohibited from using it under the GPL, just like we are using MySQL code. > MariaDB requires the same kind of legal grants that Sun asks, and the > FAQ even states "The MCA is based on Sun's Contributor Agreement. We > think it is one of the best contributor agreements around! " > http://askmonty.org/wiki/index.php/MCA_FAQ We've made some adjustments, but yes, apparently it is one of the best contributor agreements around, the community seems to think. That's why we use it. > So, where is the catch? If Sun asks for the SCA, it's "greedy > salesmen". If MariaDB asks the same thing, it's "protecting the > community". Actually, it is very important that MariaDB copyrights are pooled to one owner, we could get into trouble if we didn't do that. (I won't give the details, so as not to give any ideas.) Same is true for XtraDB, where we have advised Percona to pool the copryights from any non-Percona contributors. It's not like we (MP) want to own everything, but somebody has to do it. Personally my first motivation in favor of pooling copyrights was however entirely benevolent towards MySQL and its users. By having the copyrights pooled in one place (rather than just going all GPL), we are able to assign the code back to Sun (if we so agree), and I'd really like to do that so that all MySQL users, also the commercial licensees would benefit from the MariaDB work. henrik -- email: henrik.i...@avoinelama.fi tel: +358-40-5697354 www: www.avoinelama.fi/~hingo book: www.openlife.cc ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
Hi Arjen, (Also posted to Maria-Discuss, so postings to Maria-Developers can be dropped on follow-up). On Sep 3, 2009, at 1:21 PM, Arjen Lentz wrote: Hi Henrik, Kristian, On 03/09/2009, at 9:05 PM, Henrik Ingo wrote: On Thu, Sep 3, 2009 at 12:51 PM, Kristian Nielsen wrote: Kristian Nielsen writes: Davi Arnaut writes: I've been monitoring the maria-developers list and there are some patches that we would like to gradually incorporate into our branches. For a practical example, i would like to merge into 5.1 parts of this (http://tinyurl.com/q2ulgt) patch by Kristian. Negotiations between Monty Program and Sun have started some time ago on a deal which would cover how Sun will get the rights it needs (ie. SCA) to include the MariaDB code into the MySQL code, and how Monty Program will receive compensation for giving these rights. As I understand it, these negotiations are however still in the early phase. Apparently, this is moving nowhere :-(. Davi, I'm sorry there has been no progress on this. We discussed this again in MariaDB, and would like to get things working better for now. What Monty suggested was that it would work for Sun if we submit specific patches under a BSD-new license. I realise that the http://tinyurl.com/q2ulgt patch request is now so old that it may be of no interest to you any longer. But if you have a request for this or other patches, please send them. All developers at Monty Program have the ability to submit their work to Sun at their own discretion, so we should be able to make this work on a low-overhead technician-to-technician level. Thanks for picking this up. I just wanted to confirm this is true and we are happy to cooperate. It's how it works with Drizzle. However, I've blogged and otherwise written and spoken about this before, it does not make me particularly happy. Here's the layout: - the GPL parts of the code are owned by Sun. - BSD can be incorporated non-OSS derivatives. And the consequenec of this is that Sun is able to perpetuate the dual licensing model - a model which perhaps worked once upon a type and did well, but which is definitely outdated, and only abused by greedy salespeople. As far as I know, with the advent of the BSD licensed libdrizzle (https://launchpad.net/libdrizzle ), which also runs with MySQL (and I assume MariaDB), the dual licensing model is now irrelevant. My reading of the GPL license (http://www.gnu.org/copyleft/gpl.html) is that only programs that link GPL components become part of the greater work. From the GPL section "1. Source Code.": "For example, Corresponding Source includes interface definition files associated with source files for the work, and the source code for shared libraries and dynamically linked subprograms that the work is specifically designed to require, such as by intimate data communication or control flow between those subprograms and other parts of the work." So linked sub-programs are included. But: "However, it does not include the work's System Libraries, or general- purpose tools or generally available free programs which are used unmodified in performing those activities but which are not part of the work." So an unmodified free program like mysqld (the MySQL daemon), are not part of the work. This means that a programming linking libdrizzle, and then accessing mysqld over a communications _channel_ (like TCP, socket, pipe, etc), does not need to be GPL. By agreeing to this arrangement, contributors ensure that Sun is able to continue that. Do you want that? "yes" might be a valid choice, on the basis that it might be more important to get the changes upstream. But I think the question needs to be asked explicitly. Hence. -- Paul McCullagh PrimeBase Technologies www.primebase.org www.blobstreaming.org pbxt.blogspot.com ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
On Thu, Sep 3, 2009 at 1:40 PM, Henrik Ingo wrote: > On Thu, Sep 3, 2009 at 2:21 PM, Arjen Lentz wrote: What Monty suggested was that it would work for Sun if we submit specific patches under a BSD-new license. I realise that the http://tinyurl.com/q2ulgt patch request is now so old that it may be of no interest to you any longer. But if you have a request for this or other patches, please send them. All developers at Monty Program have the ability to submit their work to Sun at their own discretion, so we should be able to make this work on a low-overhead technician-to-technician level. >>> >>> Thanks for picking this up. I just wanted to confirm this is true and >>> we are happy to cooperate. >> >> >> It's how it works with Drizzle. >> However, I've blogged and otherwise written and spoken about this before, it >> does not make me particularly happy. >> >> Here's the layout: >> - the GPL parts of the code are owned by Sun. >> - BSD can be incorporated non-OSS derivatives. >> >> And the consequenec of this is that Sun is able to perpetuate the dual >> licensing model - a model which perhaps worked once upon a type and did >> well, but which is definitely outdated, and only abused by greedy >> salespeople. >> By agreeing to this arrangement, contributors ensure that Sun is able to >> continue that. >> Do you want that? >> >> "yes" might be a valid choice, on the basis that it might be more important >> to get the changes upstream. > > For now, we will do this for all changes we'd like to get merged > upstream, but not everything we do. Basically this amounts to "all bug > fixes". > > You're absolutely right that it doesn't make sense that something our > company employees spent hundreds of hours producing, would just be > given for free to Sun to support their business of selling essentially > proprietary MySQL licensees. We are open to talking with Sun on how to > do that too, but for now we don't see a point in just giving it away. Why it's perfectly OK that MariaDB gets MySQL bugfixes for free and instead it's a rip off if MySQL gets MariaDB bug fixes? MariaDB requires the same kind of legal grants that Sun asks, and the FAQ even states "The MCA is based on Sun's Contributor Agreement. We think it is one of the best contributor agreements around! " http://askmonty.org/wiki/index.php/MCA_FAQ So, where is the catch? If Sun asks for the SCA, it's "greedy salesmen". If MariaDB asks the same thing, it's "protecting the community". Giuseppe ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
On Thu, Sep 3, 2009 at 2:21 PM, Arjen Lentz wrote: >>> What Monty suggested was that it would work for Sun if we submit specific >>> patches under a BSD-new license. >>> >>> I realise that the http://tinyurl.com/q2ulgt patch request is now so old >>> that >>> it may be of no interest to you any longer. But if you have a request for >>> this >>> or other patches, please send them. All developers at Monty Program have >>> the >>> ability to submit their work to Sun at their own discretion, so we should >>> be >>> able to make this work on a low-overhead technician-to-technician level. >> >> Thanks for picking this up. I just wanted to confirm this is true and >> we are happy to cooperate. > > > It's how it works with Drizzle. > However, I've blogged and otherwise written and spoken about this before, it > does not make me particularly happy. > > Here's the layout: > - the GPL parts of the code are owned by Sun. > - BSD can be incorporated non-OSS derivatives. > > And the consequenec of this is that Sun is able to perpetuate the dual > licensing model - a model which perhaps worked once upon a type and did > well, but which is definitely outdated, and only abused by greedy > salespeople. > By agreeing to this arrangement, contributors ensure that Sun is able to > continue that. > Do you want that? > > "yes" might be a valid choice, on the basis that it might be more important > to get the changes upstream. For now, we will do this for all changes we'd like to get merged upstream, but not everything we do. Basically this amounts to "all bug fixes". You're absolutely right that it doesn't make sense that something our company employees spent hundreds of hours producing, would just be given for free to Sun to support their business of selling essentially proprietary MySQL licensees. We are open to talking with Sun on how to do that too, but for now we don't see a point in just giving it away. henrik ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
Hi Henrik, Kristian, On 03/09/2009, at 9:05 PM, Henrik Ingo wrote: On Thu, Sep 3, 2009 at 12:51 PM, Kristian Nielsen wrote: Kristian Nielsen writes: Davi Arnaut writes: I've been monitoring the maria-developers list and there are some patches that we would like to gradually incorporate into our branches. For a practical example, i would like to merge into 5.1 parts of this (http://tinyurl.com/q2ulgt) patch by Kristian. Negotiations between Monty Program and Sun have started some time ago on a deal which would cover how Sun will get the rights it needs (ie. SCA) to include the MariaDB code into the MySQL code, and how Monty Program will receive compensation for giving these rights. As I understand it, these negotiations are however still in the early phase. Apparently, this is moving nowhere :-(. Davi, I'm sorry there has been no progress on this. We discussed this again in MariaDB, and would like to get things working better for now. What Monty suggested was that it would work for Sun if we submit specific patches under a BSD-new license. I realise that the http://tinyurl.com/q2ulgt patch request is now so old that it may be of no interest to you any longer. But if you have a request for this or other patches, please send them. All developers at Monty Program have the ability to submit their work to Sun at their own discretion, so we should be able to make this work on a low-overhead technician-to-technician level. Thanks for picking this up. I just wanted to confirm this is true and we are happy to cooperate. It's how it works with Drizzle. However, I've blogged and otherwise written and spoken about this before, it does not make me particularly happy. Here's the layout: - the GPL parts of the code are owned by Sun. - BSD can be incorporated non-OSS derivatives. And the consequenec of this is that Sun is able to perpetuate the dual licensing model - a model which perhaps worked once upon a type and did well, but which is definitely outdated, and only abused by greedy salespeople. By agreeing to this arrangement, contributors ensure that Sun is able to continue that. Do you want that? "yes" might be a valid choice, on the basis that it might be more important to get the changes upstream. But I think the question needs to be asked explicitly. Hence. Cheers, Arjen. -- Arjen Lentz, Exec.Director @ Open Query (http://openquery.com) Exceptional Services for MySQL at a fixed budget. Follow our blog at http://openquery.com/blog/ OurDelta: enhanced builds for MySQL @ http://ourdelta.org ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
On Thu, Sep 3, 2009 at 12:51 PM, Kristian Nielsen wrote: > Kristian Nielsen writes: > >> Davi Arnaut writes: >> >>> I've been monitoring the maria-developers list and there are some >>> patches that we would like to gradually incorporate into our >>> branches. For a practical example, i would like to merge into 5.1 >>> parts of this (http://tinyurl.com/q2ulgt) patch by Kristian. > >> Negotiations between Monty Program and Sun have started some time ago on a >> deal which would cover how Sun will get the rights it needs (ie. SCA) to >> include the MariaDB code into the MySQL code, and how Monty Program will >> receive compensation for giving these rights. As I understand it, these >> negotiations are however still in the early phase. > > Apparently, this is moving nowhere :-(. > > Davi, I'm sorry there has been no progress on this. We discussed this again in > MariaDB, and would like to get things working better for now. > > What Monty suggested was that it would work for Sun if we submit specific > patches under a BSD-new license. > > I realise that the http://tinyurl.com/q2ulgt patch request is now so old that > it may be of no interest to you any longer. But if you have a request for this > or other patches, please send them. All developers at Monty Program have the > ability to submit their work to Sun at their own discretion, so we should be > able to make this work on a low-overhead technician-to-technician level. Kristian, Thanks for picking this up. I just wanted to confirm this is true and we are happy to cooperate. henrik -- Henrik Ingo Project Manager and COO, Monty Program Ab hi...@askmonty.org, skype:henrik.ingo, +358405697354 http://askmonty.org/wiki/index.php/About_Us What's up with MariaDB? http://askmonty.org/wiki/index.php/MariaDB ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] MariaDB and the SCA
Kristian Nielsen writes: > Davi Arnaut writes: > >> I've been monitoring the maria-developers list and there are some >> patches that we would like to gradually incorporate into our >> branches. For a practical example, i would like to merge into 5.1 >> parts of this (http://tinyurl.com/q2ulgt) patch by Kristian. > Negotiations between Monty Program and Sun have started some time ago on a > deal which would cover how Sun will get the rights it needs (ie. SCA) to > include the MariaDB code into the MySQL code, and how Monty Program will > receive compensation for giving these rights. As I understand it, these > negotiations are however still in the early phase. Apparently, this is moving nowhere :-(. Davi, I'm sorry there has been no progress on this. We discussed this again in MariaDB, and would like to get things working better for now. What Monty suggested was that it would work for Sun if we submit specific patches under a BSD-new license. I realise that the http://tinyurl.com/q2ulgt patch request is now so old that it may be of no interest to you any longer. But if you have a request for this or other patches, please send them. All developers at Monty Program have the ability to submit their work to Sun at their own discretion, so we should be able to make this work on a low-overhead technician-to-technician level. - Kristian. ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp
Re: [Maria-developers] Rev 2720: Fix for in file:///Users/hakan/work/monty_program/maria/
Hakan Kuecuekyilmaz пишет: > At file:///Users/hakan/work/monty_program/maria/ > > > revno: 2720 > revision-id: ha...@askmonty.org-20090902214617-5aljfm0c229hzdpu > parent: kniel...@knielsen-hq.org-20090805072137-wg97dcem1cxnzt3p > committer: Hakan Kuecuekyilmaz > branch nick: maria > timestamp: Wed 2009-09-02 23:46:17 +0200 > message: > Fix for > mysqlslap: setting --engine does not get replicated > http://bugs.mysql.com/bug.php?id=46967 > > Problem was that a "set storage_engine = " > is not replicated. > > Fixed by adding an explicit > Engine = > > to the build_table_string() function. > > Open question: > Should MySQL's replication actually replicate the > set storage_engine command or not? Answer: no it should > not (documented). The correct question is should be default engine be replicated as 'default' or as certain engine. I think it should be 'default' (it seems logical for me and gives more space for different setup of master/slave). > === modified file 'client/mysqlslap.c' > --- a/client/mysqlslap.c 2009-04-25 10:05:32 + > +++ b/client/mysqlslap.c 2009-09-02 21:46:17 + > @@ -880,6 +880,12 @@ > } > >dynstr_append(&table_string, ")"); > + > + if (default_engine) { Move above { on a new line, This program should follow mysql code style. > +dynstr_append(&table_string, " Engine = "); > +dynstr_append(&table_string, default_engine); > + } > + >ptr= (statement *)my_malloc(sizeof(statement), >MYF(MY_ZEROFILL|MY_FAE|MY_WME)); >ptr->string = (char *)my_malloc(table_string.length+1, ___ Mailing list: https://launchpad.net/~maria-developers Post to : maria-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~maria-developers More help : https://help.launchpad.net/ListHelp