| 00:00.06 | oubiwann | heh |
| 00:00.15 | *** join/#storm wallflower (n=wallflow@ip-205-246-113-216.pool.grokthis.net) |
| 00:00.19 | radix | oubiwann: I've been getting pinged a lot lately :) |
| 00:00.24 | radix | oubiwann: and I'm sorry I haven't gotten to that review |
| 00:00.33 | oubiwann | I'll keep reminding you :-) |
| 00:00.37 | radix | oubiwann: I'm working on Twisted regression reviews/fixes now, trying to get 8.0.2 out |
| 00:00.50 | radix | oubiwann: because I'm trying to get it into hardy on an exception, but need the regressions fixed first |
| 00:01.17 | radix | exarkun and I are having a Twisted pico-sprint at the office |
| 00:01.29 | oubiwann | you're like a superperson, man |
| 00:01.31 | oubiwann | heh |
| 00:01.37 | oubiwann | femto-sprint? |
| 00:02.25 | radix | :-) |
| 00:43.24 | *** join/#storm jml_ (n=jml@ppp121-44-21-53.lns10.syd7.internode.on.net) |
| 02:33.08 | *** join/#storm harnak (n=harnak@bilbo.netconcepts.com) |
| 04:34.08 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 04:54.17 | *** join/#storm Brend (n=brendonh@220-131-224-198.HINET-IP.hinet.net) |
| 05:07.02 | *** join/#storm jamesh (n=james@canonical/launchpad/jamesh) |
| 05:43.17 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 05:46.45 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 05:53.51 | *** join/#storm dobee (n=dobeee@lsfw01.lovelysystems.com) |
| 07:32.55 | *** join/#storm mcella (n=mcella@ip-176-123.sn2.eutelia.it) |
| 10:53.45 | *** join/#storm harnak (n=harnak@ip-118-90-75-216.xdsl.xnet.co.nz) |
| 11:28.11 | *** join/#storm mcella_ (n=mcella@ip-176-123.sn2.eutelia.it) |
| 12:16.34 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 12:31.51 | *** join/#storm niemeyer (n=niemeyer@200-103-245-169.ctame705.dsl.brasiltelecom.net.br) |
| 12:32.11 | *** join/#storm aa_ (n=ali@pida/aa) |
| 12:33.08 | aa_ | Hi, I've got a weird thing I am probably doing something wrong. I get an ordered ResultSet and it has len(2), and when I create a list with the items I can see them, but then when I attempt to index it, or use first() or any other way of getting the object it returns None |
| 12:34.15 | aa_ | store.find(Episode, patient=self).order_by(Episode.date).first() |
| 12:34.22 | aa_ | is the code I am using |
| 12:40.18 | aa_ | ok, now I am convinced that I have gone insane |
| 12:41.08 | aa_ | so, I do list(res) and get a list, with length two, and appears to have the items in it. Then doing l[0] returns None |
| 12:41.22 | aa_ | is that me or is there some pretty dark magic going on here? |
| 12:48.45 | jamesh | aa_: try doing "from storm.tracer import debug; debug(True)" |
| 12:48.49 | jamesh | and see if the SQL looks sane |
| 12:48.55 | aa_ | jamesh: ok |
| 12:49.26 | aa_ | oh hmm, no tracer module |
| 12:49.28 | jamesh | maybe try issuing some of the problem queries manually to check the results |
| 12:49.36 | aa_ | version 0.12 from setuptools |
| 12:50.01 | aa_ | jamesh: so that symptom definitely indicates a problem? |
| 12:50.08 | aa_ | returning None like that |
| 12:50.55 | jamesh | storm.tracer is post-0.12 |
| 12:51.03 | aa_ | ok, branching etc. |
| 12:51.10 | jamesh | aa_: try "import storm.database; storm.database.DEBUG = True" instead |
| 12:54.41 | aa_ | query looks sane enough, does episode."date" look right? |
| 12:55.17 | jamesh | storm quotes table and column names that also happen to be SQL identifiers |
| 12:55.24 | jamesh | shouldn't be a problem |
| 12:55.25 | aa_ | yeah, that makes sense |
| 12:55.37 | jamesh | SQL keywords, I mean |
| 12:56.24 | aa_ | does this seem ok: WHERE episode.patient_id = ?', [<storm.variables.IntVariable object at 0x867652c>] |
| 12:56.44 | aa_ | I mean it does seem to work |
| 12:56.52 | aa_ | the resulting set has 2 items |
| 12:57.59 | aa_ | which then magically become None on realization |
| 13:24.27 | aa_ | ok, well I have tracked this down to a call to class.__new__(cls) |
| 13:25.09 | aa_ | I think it has to be a bug in python its so weird |
| 13:26.23 | aa_ | that call returns None |
| 13:35.54 | niemeyer | aa_: If you can create a small self-contained test case which demonstrates the issue, we can debug it |
| 13:37.36 | aa_ | yes |
| 13:37.52 | *** join/#storm goschtl (n=goschtl@p5B0BE278.dip.t-dialin.net) |
| 13:41.09 | aa_ | which I just can't do because in normal situations it works fine |
| 13:42.56 | aa_ | niemeyer: does __new__ ever get overridden? |
| 13:45.54 | aa_ | nevermind, I'll dig around |
| 13:49.11 | niemeyer | aa_: Nope |
| 13:49.50 | niemeyer | aa_: Smart doesn't really do any black magic |
| 13:49.59 | niemeyer | aa_: Storm |
| 13:50.04 | aa_ | then this is weird |
| 13:50.08 | niemeyer | is crazy |
| 13:50.12 | niemeyer | aa_: Hmm |
| 13:51.30 | niemeyer | aa_: Well.. yeah, it's hard to tell what's going on |
| 13:51.50 | niemeyer | aa_: If you can reproduce it somehow it a way we can see, we'll happily take a look |
| 13:52.03 | aa_ | right, I am thinking about that |
| 14:12.42 | aa_ | oh my god |
| 14:13.09 | aa_ | so I noticed that this none type getting returned had a load of methods |
| 14:13.31 | aa_ | and then I realised that it was an object who's __str__ was returning str(None) |
| 14:15.15 | aa_ | is that not the dumbest thing anyone ever ehard |
| 14:18.04 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 14:35.16 | radix | aa_: was that a __str__ defined by storm? |
| 14:35.26 | radix | aa_: or something in your application's code? |
| 14:45.04 | *** part/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 14:54.02 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 15:41.16 | *** join/#storm johan (n=jdahlin@201-68-7-115.dsl.telesp.net.br) |
| 15:52.40 | vinsci | niemeyer, how's your review schedule? Still waiting for feedback on http://paste.ubuntu.com/5787/ |
| 16:13.22 | niemeyer | vinsci: I haven't had time to look deeply into it, but I did notice that it's incompatible with how Storm currently work |
| 16:13.23 | niemeyer | s |
| 16:14.34 | vinsci | well, if you set the flag for a column that's true. If you don't set it, nothing has changed and the unit tests pass. |
| 16:15.29 | *** join/#storm festizio (n=greg@69.74.217.43) |
| 16:15.49 | niemeyer | vinsci: Which flag? |
| 16:15.52 | niemeyer | vinsci: |
| 16:15.53 | niemeyer | + # Define the PropertyColumn using the class where the property is |
| 16:15.54 | niemeyer | + # defined, rather than the class where it has been inherited. |
| 16:16.03 | niemeyer | vinsci: This is changing the behavior, without any flags as far as I could see |
| 16:16.39 | vinsci | niemeyer, hm, I added the flag after posting that patch - hold on |
| 16:17.56 | niemeyer | vinsci: So what's the benefit of the patch again? |
| 16:19.39 | vinsci | it allows you to map inherited classes to normalized tables, ie. the fields from the base class are stored in the base class table |
| 16:20.26 | vinsci | as Storm works now, you can only handle inheritance if you're willing to live with a non-normalized DB schema (I'm not ;-) |
| 16:20.43 | niemeyer | vinsci: That's certainly not true, as we discussed extensively |
| 16:20.59 | niemeyer | vinsci: Storm may be used with whatever schema is in the database |
| 16:21.22 | niemeyer | vinsci: Now, I don't follow what it really does |
| 16:21.59 | niemeyer | vinsci: You want Company.person_id to actually mean person.person_id? |
| 16:22.04 | vinsci | you have the example program in the patch, see what it does and how it behaves |
| 16:22.17 | niemeyer | vinsci: I've read it |
| 16:22.30 | vinsci | run it and look at the resulting database |
| 16:22.34 | niemeyer | vinsci: Ok |
| 16:24.09 | niemeyer | vinsci: Employee inherits from Person, and has a Person attribute? |
| 16:24.16 | niemeyer | vinsci: Which refers to Person |
| 16:26.13 | vinsci | it acts as the FK, so that the Person fields can be stored in the Person table |
| 16:28.02 | vinsci | I seem to have the updated patch on the laptop - bbiab |
| 16:28.10 | niemeyer | vinsci: The Person fields can be stored in the person table without any patching (!?) |
| 16:28.18 | niemeyer | vinsci: http://paste.ubuntu.com/6672/ |
| 16:28.31 | niemeyer | vinsci: Here is your example working with trunk |
| 16:28.46 | niemeyer | vinsci: I haven't touched the schema, nor any lines using the code |
| 16:29.02 | niemeyer | vinsci: Only changed the definition of Employee |
| 16:48.10 | vinsci | niemeyer, ah, thanks - will try that. Was there any changes between 0.12 and trunk that makes that work, or should it work in 0.12 as well? |
| 16:48.31 | vinsci | found the newer patch version on the laptop as well |
| 16:54.45 | *** join/#storm marianom (n=marianom@ubuntu/member/marianom) |
| 17:00.04 | niemeyer | vinsci: It should work with any public version, I think |
| 17:10.27 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 17:42.06 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 17:42.59 | *** join/#storm dobee (n=dobeee@85-124-200-100.static.xdsl-line.inode.at) |
| 21:00.36 | thumper | morning people |
| 21:00.45 | thumper | niemeyer: still around? |
| 21:00.55 | *** part/#storm johan (n=jdahlin@201-68-7-115.dsl.telesp.net.br) |
| 21:01.41 | niemeyer | thumper: Heya |
| 21:01.44 | niemeyer | thumper: Yep, still here |
| 21:48.43 | *** join/#storm harnak (n=harnak@bilbo.netconcepts.com) |
| 23:09.51 | *** part/#storm marianom (n=marianom@ubuntu/member/marianom) |