There are a fair few issues against the SQL Server db adapter that I''d like to see fixed. There are also a couple of features I''d like to add, but don''t really want to until these bugs have been fixed. It''s time to show SQL Server some love! Many of these outstanding issues have patches against that need reviewing, testing or just generally cleaning up. I''m more than willing to go through these patches and do this work, but what''s the best way to get them actually committed to trunk? Applying each patch individually causes conflicts as they (mostly) work against the same file, so once one or two have been committed, the rest would need rewriting. I''d prefer to create a larger combined bugfix patch, but don''t want to waste my time if it''s not going to be useful. Thoughts? Tom
> Many of these outstanding issues have patches against that need > reviewing, testing or just generally cleaning up. I''m more than > willing to go through these patches and do this work, but what''s the > best way to get them actually committed to trunk?Create one new aggregated patch and mark the others as duplicate. We can then apply the aggregate. We''ve done that in the past for both SQL Server and Oracle too. Thanks for taking this on. -- David Heinemeier Hansson http://www.loudthinking.com -- Broadcasting Brain http://www.basecamphq.com -- Online project management http://www.backpackit.com -- Personal information manager http://www.rubyonrails.com -- Web-application framework
On 7/6/06, Tom Ward <tom@popdog.net> wrote:> There are a fair few issues against the SQL Server db adapter that I''d > like to see fixed. There are also a couple of features I''d like to > add, but don''t really want to until these bugs have been fixed. It''s > time to show SQL Server some love!I''d like to apologize for letting the adapter atrophy. I''ve been extremely neglectful of a number of commitments recently ("it''s not you, it''s me" :) I think it would be in the best interest of the project for a new maintainer to take over regular testing and patch verification duties. My time is becoming unreliable and I''m currently in the process of making sure I never have to deal with SQL Server ever again so I hardly make for an ideal maintainer. Tom has arguably contributed every bit as much if not more to the apapter as myself so I''d like to nominate him. He was probably the right guy for the job months ago. Thanks, Ryan
On Jul 7, 2006, at 6:51 PM, Ryan Tomayko wrote:> On 7/6/06, Tom Ward <tom@popdog.net> wrote: >> There are a fair few issues against the SQL Server db adapter that >> I''d >> like to see fixed. There are also a couple of features I''d like to >> add, but don''t really want to until these bugs have been fixed. It''s >> time to show SQL Server some love! > > Tom has arguably contributed every bit as much if not more to the > apapter as myself so I''d like to nominate him. He was probably the > right guy for the job months ago.Thanks for all the work you''ve put into the adapter, Ryan. Good luck migrating away. So, Tom, you looking to be a bug magnet? ;) jeremy
Jeremy Kemper wrote:> So, Tom, you looking to be a bug magnet? ;)Tom: let me know if you''d like a copy of the script I use to test against the Oracle adapter after every checkin. It''s not very fancy, but it does a good job spamming rails-core when stuff breaks. ;-)
On 7/8/06, Michael A. Schoen <schoenm@earthlink.net> wrote:> Jeremy Kemper wrote: > > So, Tom, you looking to be a bug magnet? ;) > > Tom: let me know if you''d like a copy of the script I use to test > against the Oracle adapter after every checkin. It''s not very fancy, but > it does a good job spamming rails-core when stuff breaks. ;-)That''d be great. I''ll enjoy flooding this list with SQL Server test failures! Tom
On 7/8/06, Jeremy Kemper <jeremy@bitsweat.net> wrote:> On Jul 7, 2006, at 6:51 PM, Ryan Tomayko wrote: > > On 7/6/06, Tom Ward <tom@popdog.net> wrote: > >> There are a fair few issues against the SQL Server db adapter that > >> I''d > >> like to see fixed. There are also a couple of features I''d like to > >> add, but don''t really want to until these bugs have been fixed. It''s > >> time to show SQL Server some love! > > > > Tom has arguably contributed every bit as much if not more to the > > apapter as myself so I''d like to nominate him. He was probably the > > right guy for the job months ago. > > Thanks for all the work you''ve put into the adapter, Ryan. Good luck > migrating away. > > So, Tom, you looking to be a bug magnet? ;)Sure, I''m happy to take on the job. I''m going to be stuck using SQL Server at work for the forseeable future, so it''s in my best interests to keep the adapter in good shape. Tom
On 7/8/06, Ryan Tomayko <rtomayko@gmail.com> wrote:> Tom has arguably contributed every bit as much if not more to the > apapter as myself so I''d like to nominate him. He was probably the > right guy for the job months ago.That''s very gracious of you Ryan, but don''t underestimate your own contribution. Thanks for all the work you''ve put in. Tom
* Michael A. Schoen (schoenm@earthlink.net) [060707 22:18]:> Tom: let me know if you''d like a copy of the script I use to test > against the Oracle adapter after every checkin. It''s not very fancy, but > it does a good job spamming rails-core when stuff breaks. ;-)We have some interest in providing the same service for Postgres -- would you be so kind as to forward me a copy of your script and we''ll see if we can get a Postgres test server set up (unless there are some objections by people on the core list). Rick -- http://www.rickbradley.com MUPRN: 673 | result in civil random email haiku | penalties and claims of PER | OCCURRENCE USC 227[c]).
Having both Postgres and SQL server in addition to Oracle autotest results on the list would spam it pretty much, I think. What do others think? On 7/8/06, Rick Bradley <rick@rickbradley.com> wrote:> > We have some interest in providing the same service for Postgres -- > would you be so kind as to forward me a copy of your script and we''ll > see if we can get a Postgres test server set up (unless there are some > objections by people on the core list)._______________________________________________ Rails-core mailing list Rails-core@lists.rubyonrails.org http://lists.rubyonrails.org/mailman/listinfo/rails-core
Ideally, if you could have the same person manage Oracle and other databases, then you could have the Oracle tests run, and if they pass, run the Postgresql tests. While this would delay the report of Postgresql specific problems when there is also an Oracle specific problem, it would reduce spam. We have Postgresql and Oracle at work, or perhaps Michael, who reports the Oracle tests, could set it up for Postgresql? Regards, Blair On Jul 8, 2006, at 9:03 AM, Mislav Marohnić wrote:> Having both Postgres and SQL server in addition to Oracle autotest > results on the list would spam it pretty much, I think. What do > others think? > > On 7/8/06, Rick Bradley <rick@rickbradley.com> wrote: We have some > interest in providing the same service for Postgres -- > would you be so kind as to forward me a copy of your script and we''ll > see if we can get a Postgres test server set up (unless there are some > objections by people on the core list).
I don't think we should reduce that sort of spam. The point is that it's annoying so it gets fixed. Maybe we can put something in the subject though so those of us who don't care about oracle and pgsql breaking can filter it out? Kev On 7/8/06, Blair Zajac <blair@orcaware.com> wrote:> Ideally, if you could have the same person manage Oracle and other > databases, then you could have the Oracle tests run, and if they > pass, run the Postgresql tests. While this would delay the report of > Postgresql specific problems when there is also an Oracle specific > problem, it would reduce spam. > > We have Postgresql and Oracle at work, or perhaps Michael, who > reports the Oracle tests, could set it up for Postgresql? > > Regards, > Blair > > On Jul 8, 2006, at 9:03 AM, Mislav Marohnić wrote: > > > Having both Postgres and SQL server in addition to Oracle autotest > > results on the list would spam it pretty much, I think. What do > > others think? > > > > On 7/8/06, Rick Bradley <rick@rickbradley.com> wrote: We have some > > interest in providing the same service for Postgres -- > > would you be so kind as to forward me a copy of your script and we'll > > see if we can get a Postgres test server set up (unless there are some > > objections by people on the core list). > > _______________________________________________ > Rails-core mailing list > Rails-core@lists.rubyonrails.org > http://lists.rubyonrails.org/mailman/listinfo/rails-core >-- Kevin Clark http://glu.ttono.us _______________________________________________ Rails-core mailing list Rails-core@lists.rubyonrails.org http://lists.rubyonrails.org/mailman/listinfo/rails-core
In an ideal world, code should never be checked into the repository that breaks the unit tests, one of the points of continuous integration is to encourage people not to check in broken code. I develop Rails apps (with Site5) on my Powerbook at home, but they are always deployed to Linux boxes, which, due to the wide ranges of libraries we use, sometimes causes incompatabilities and build breakages. Although my tests often run okay on my pbook, they may break on our Linux CIA box - to prevent that, I have a script that rsync's my working copy to an account on our test Linux box and runs the tests to confirm that nothing is broken. Though there may be some effort involved, perhaps it is worthwhile for the comitters to be set up with accounts on a 'test box' that is capable of running the entire suite of unit tests. Obviously, to handle SQL Server it would have to be a windows box, but luckily, nearly all (perhaps all?) of the databases supported are free, or have free versions available - I know that crippled (but adequate) versions of SQL Server and Oracle are available, not sure about Sybase (is that free now?) and DB2. Having a single activerecord test box would (IMO) go a long way towards maintaining continuous integrity of the code, and it would also have the added side effect of minimising the build breakage spam on this list :) Cheers, -David Felstead On 7/9/06, Kevin Clark <kevin.clark@gmail.com> wrote:> I don't think we should reduce that sort of spam. The point is that > it's annoying so it gets fixed. Maybe we can put something in the > subject though so those of us who don't care about oracle and pgsql > breaking can filter it out? > > Kev > > On 7/8/06, Blair Zajac <blair@orcaware.com> wrote: > > Ideally, if you could have the same person manage Oracle and other > > databases, then you could have the Oracle tests run, and if they > > pass, run the Postgresql tests. While this would delay the report of > > Postgresql specific problems when there is also an Oracle specific > > problem, it would reduce spam. > > > > We have Postgresql and Oracle at work, or perhaps Michael, who > > reports the Oracle tests, could set it up for Postgresql? > > > > Regards, > > Blair > > > > On Jul 8, 2006, at 9:03 AM, Mislav Marohnić wrote: > > > > > Having both Postgres and SQL server in addition to Oracle autotest > > > results on the list would spam it pretty much, I think. What do > > > others think? > > > > > > On 7/8/06, Rick Bradley <rick@rickbradley.com> wrote: We have some > > > interest in providing the same service for Postgres -- > > > would you be so kind as to forward me a copy of your script and we'll > > > see if we can get a Postgres test server set up (unless there are some > > > objections by people on the core list). > > > > _______________________________________________ > > Rails-core mailing list > > Rails-core@lists.rubyonrails.org > > http://lists.rubyonrails.org/mailman/listinfo/rails-core > > > > > -- > Kevin Clark > http://glu.ttono.us > > _______________________________________________ > Rails-core mailing list > Rails-core@lists.rubyonrails.org > http://lists.rubyonrails.org/mailman/listinfo/rails-core > > >_______________________________________________ Rails-core mailing list Rails-core@lists.rubyonrails.org http://lists.rubyonrails.org/mailman/listinfo/rails-core
On 7/8/06, Kevin Clark <kevin.clark@gmail.com> wrote:> I don''t think we should reduce that sort of spam. The point is that > it''s annoying so it gets fixed. Maybe we can put something in the > subject though so those of us who don''t care about oracle and pgsql > breaking can filter it out?I think it''d be best if as many tests as possible run on wrath (the rails server), some databases like sql server and oracle will have to be run elsewhere. But most of the open source databases should be able to be run on wrath. -- Cheers Koz
David, If you''d like to setup every database and build a system that everyone who does core development can rsync against, PDI. Everyone on this list understands that broken code shouldn''t be checked in, but you''re right, access is an issue. If you depend on Oracle or DB2 or one of the other non-free SQL databases then feel free to do what you can to make sure new changes don''t break that database. Kevin
On 7/6/06, David Heinemeier Hansson <david.heinemeier@gmail.com> wrote:> Create one new aggregated patch and mark the others as duplicate. We > can then apply the aggregate. We''ve done that in the past for both SQL > Server and Oracle too.I''ve created a patch against #5670 which fixes many minor problems. Hopefully I''ll have some time at the weekend to try and sort out some of the more complex adapter problems. Tom