| 00:00.19 | *** join/#storm wallflower (n=wallflow@ip-205-246-113-216.pool.grokthis.net) |
| 01:41.06 | pyCube | hmm.. so get_insert_identity is not implemented for mysql.. heh |
| 01:45.22 | jamesh | pyCube: didn't we have this same discussion before? Your database current workflow makes it impossible to know the identity of newly inserted rows. |
| 01:46.05 | jamesh | if MySQL implemented the PostgreSQL "INSERT ... RETURNING ..." statement it'd be workable, but MySQL doesn't provide the needed information. |
| 01:46.18 | pyCube | i know.. i am just trying to figure out what to do.. |
| 01:46.57 | jamesh | work within MySQL's limitations or switch to a database that doesn't have those limitations. |
| 01:47.08 | pyCube | heh |
| 01:47.11 | pyCube | right |
| 01:48.10 | pyCube | i dont really get the problem though... i mean, the tables have an on before insert trigger that fills in the data.. |
| 01:49.06 | jamesh | pyCube: right. But how does the client find out what data was filled in? |
| 01:49.24 | jamesh | especially when the information needed to identify the row is what the trigger is filling in |
| 01:49.36 | pyCube | i dunno.. but it seems to only be a problem when using storm |
| 01:49.56 | jamesh | are you saying that other ORMs handle it? |
| 01:50.12 | pyCube | no |
| 01:50.43 | jamesh | because it sounds like it will be a problem for any program that wants to insert data and be able to identify the newly inserted row |
| 01:50.58 | jamesh | which is pretty much any ORM |
| 01:51.06 | pyCube | right.. |
| 01:51.38 | pyCube | how is it different than an auto inc column |
| 01:51.39 | pyCube | ? |
| 01:52.16 | jamesh | MySQL provides an API for determining the value assigned to the auto increment column. |
| 01:53.27 | jamesh | it is very specific to auto increment columns though, so won't cover other sorts of server side primary key allocation |
| 01:54.48 | jamesh | http://dev.mysql.com/doc/refman/5.0/en/mysql-insert-id.html |
| 01:54.56 | pyCube | sure |
| 01:55.57 | pyCube | well.. hehe.. this is an auto inc column.. just on a different table |
| 01:56.03 | pyCube | heheh |
| 01:57.35 | jamesh | well, if it doesn't cover the cases mentioned in that page, then MySQL isn't sending the ID to us |
| 01:58.03 | pyCube | might it be possible to maintain my own tweaking of the sotrm/mysql code to do what i need, or should i just press for giving each table its own id? |
| 01:58.45 | jamesh | With PostgreSQL, you can do things like "INSERT INTO foo (name) VALUES ('bar') RETURNING foo.id", which will work however foo.id is allocated |
| 01:59.06 | pyCube | right |
| 01:59.26 | pyCube | i WISH this db was postgresql, but it was all determined before i started here |
| 01:59.43 | jamesh | if you can do what the trigger does on the application side, you could set the primary key explicitly |
| 02:01.13 | pyCube | i played with that idea.. asking the toc table for the id that it would next assign and passing that in explicitly for inserts |
| 02:01.20 | pyCube | but that scares me |
| 02:05.05 | jamesh | what is so scary about doing that app-side? |
| 02:05.22 | jamesh | you've still got transactional isolation, right? |
| 02:05.58 | pyCube | yeah |
| 02:58.42 | jml | hi |
| 02:59.09 | jml | how do you do a 'foo, count(bar)' / group by thingy with storm? |
| 03:01.32 | jamesh | result = store.find((Foo, Count(bar)), where_conditions); result.group_by(Foo) |
| 03:01.52 | jml | jamesh: thank you. |
| 03:02.12 | jamesh | you'll get (foo_instance, integer) pairs in the result set |
| 03:09.38 | jml | jamesh: cool. that's what I expected _and_ wanted :) |
| 04:32.22 | *** join/#storm kripken (n=alon@217.132.18.11) |
| 04:40.55 | jml | jamesh: how would I order that by the integer? |
| 04:41.15 | jamesh | jml: result.order_by(Count(bar)) |
| 04:42.01 | jml | that easy, eh? |
| 04:42.29 | jamesh | if there is an obvious answer to the question, it is often the right one with storm :) |
| 04:43.11 | jml | heh |
| 05:25.45 | mwhudson | that should be the quote of the release, or something |
| 05:50.40 | *** join/#storm jukart (n=jukart@d91-128-122-97.cust.tele2.at) |
| 06:06.16 | *** join/#storm kov (n=kov@debian/developer/kov) |
| 06:44.56 | *** join/#storm jukart (i=lovely@81.189.156.94) |
| 06:48.46 | *** join/#storm mcella (n=mcella@ip-142-69.sn3.eutelia.it) |
| 08:03.12 | *** join/#storm kov (n=kov@debian/developer/kov) |
| 08:06.13 | *** join/#storm maze (i=wijnand@CAcert/Assurer/maze) |
| 08:32.23 | *** join/#storm goschtl (n=goschtl@p5B0BC293.dip.t-dialin.net) |
| 09:09.29 | *** join/#storm mcella (n=mcella@ip-142-69.sn3.eutelia.it) |
| 09:55.06 | *** join/#storm niemeyer (n=niemeyer@201-89-10-224.ctame705.dsl.brasiltelecom.net.br) |
| 10:16.36 | *** join/#storm sidnei (n=sidnei@plone/dreamcatcher) |
| 12:21.56 | *** join/#storm jukart (n=jukart@d91-128-122-97.cust.tele2.at) |
| 12:50.09 | *** join/#storm kov (n=kov@debian/developer/kov) |
| 13:25.40 | *** join/#storm vvinet (n=vince@132.210.76.203) |
| 13:47.36 | *** join/#storm jukart (n=jukart@d91-128-122-97.cust.tele2.at) |
| 13:58.51 | *** join/#storm jukart_ (n=jukart@d91-128-122-97.cust.tele2.at) |
| 15:01.08 | *** join/#storm oubiwann (n=oubiwann@97-119-87-190.omah.qwest.net) |
| 16:35.45 | *** join/#storm sidnei (n=sidnei@plone/dreamcatcher) |
| 16:44.54 | *** join/#storm andrea-bs (n=andrea-b@ubuntu/member/beeseek.developer.andrea-bs) |
| 17:18.45 | *** join/#storm bac (n=bac@cpe-065-190-187-178.nc.res.rr.com) |
| 18:48.51 | *** join/#storm jukart (n=jukart@d91-128-122-97.cust.tele2.at) |
| 18:52.54 | *** join/#storm andrea-bs (n=andrea-b@ubuntu/member/beeseek.developer.andrea-bs) |
| 22:48.17 | *** join/#storm bigdog3 (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |