| 01:27.01 | *** join/#storm wgrant (n=wgrant@ubuntu/member/wgrant) [NETSPLIT VICTIM] |
| 01:27.01 | *** join/#storm munin (n=munin@209.9.227.50) [NETSPLIT VICTIM] |
| 01:27.01 | *** join/#storm spats (n=spats@mail.fitness2live.com.au) [NETSPLIT VICTIM] |
| 01:28.01 | *** join/#storm vvinet (n=vince@modemcable188.10-70-69.static.videotron.ca) [NETSPLIT VICTIM] |
| 01:28.01 | *** join/#storm radix (n=r@wordeology.com) [NETSPLIT VICTIM] |
| 02:55.23 | jamesh | fcorrea: calling flush() would be enough. |
| 02:56.22 | jamesh | fcorrea: another alternative is to define id as Int(primary=True, default=AutoReload) |
| 02:56.52 | jamesh | that way accessing the id attribute will cause the object to be flushed automatically to retrieve the real ID |
| 02:59.02 | fcorrea | jamesh: right |
| 03:00.23 | fcorrea | I ended up on trial and error and flush() was enough. Didn't know about AutoReload |
| 03:01.05 | jamesh | fcorrea: if you just want to set up a foreign key relationship though, just use a Reference() |
| 03:02.49 | fcorrea | mmm....so if you have a A class with a reference to B as in A.id = Reference(B.id), setting A.id = 2 would set in both A and B? |
| 03:03.15 | fcorrea | wooops, set in B but when you access A.id you get it from B |
| 03:03.56 | *** join/#storm shaunm (n=shaunm@c-98-212-133-244.hsd1.il.comcast.net) |
| 03:05.01 | jamesh | fcorrea: if you have class A with primary key "id", and define class B with primary key "id" and "a = Reference(id, A.id)" |
| 03:05.18 | jamesh | then setting some_b.a = some_a will create the relationship |
| 03:05.51 | fcorrea | that's nice |
| 03:05.54 | jamesh | Storm will flush the two objects in the right order and set some_b.id to some_a.id |
| 03:07.39 | fcorrea | Great. Now i just need to make this wsgi middleware to be thread safe. Currently fighting threading.local |
| 03:08.55 | fcorrea | Anyways, tks for you help jamesh |
| 03:10.50 | jamesh | fcorrea: one of our users wrote up a bit about using the storm.zope thread local store code in WSGI apps here: https://storm.canonical.com/PylonsRepoze.tm2 |
| 03:11.28 | jamesh | you might find it helpful in managing your DB connections |
| 03:12.20 | fcorrea | jamesh: yeah, I am using storm+repoze.tm as well. The problem is that I am trying to get my store on a thread local basis |
| 03:13.05 | jamesh | fcorrea: the storm.zope code (specifically the ZStorm bits) implement that for you |
| 03:13.06 | fcorrea | then the other data manager (which is just responsible for committing or rolling back transactions) gets a store and commits it but nothing gets to the database |
| 03:13.44 | jamesh | it implements thread local stores, and transaction manager integration |
| 03:14.32 | fcorrea | yeah, already read that code. This is a repoze.zope2 app and I which I could use the zstorm utility but it won't work because of the version of ZODB |
| 03:14.56 | fcorrea | s/which/wish/ |
| 03:15.11 | jamesh | what does the ZODB version have to do with it? |
| 03:15.52 | fcorrea | jamesh: there's code from the transaction that is used in the z3 integration that just doesn't work in zope2 |
| 03:16.49 | fcorrea | so if you create a store like <store name="foo" uri"something" /> it wont work and last time I checked it was related to code from the transaction module |
| 03:16.55 | fcorrea | but I might be wrong |
| 03:19.33 | jamesh | if you're configuring via ZCML, make sure you're including meta.zcml from storm.zope |
| 03:19.51 | fcorrea | sure |
| 03:23.42 | jamesh | if you can reproduce the error, it'd be helpful if you file a bug report |
| 03:24.10 | jamesh | I can't promise we'll know how to fix it, but without the information we almost certainly won't be able to. |
| 03:25.06 | fcorrea | jamesh: np, will play with that code tomorrow morning and will report if something bad happens |
| 04:33.33 | *** join/#storm oubiwann (n=oubiwann@97-119-81-240.omah.qwest.net) [NETSPLIT VICTIM] |
| 04:37.23 | *** join/#storm shaunm (n=shaunm@c-98-212-133-244.hsd1.il.comcast.net) [NETSPLIT VICTIM] |
| 04:37.23 | *** join/#storm sidnei (n=sidnei@plone/dreamcatcher) [NETSPLIT VICTIM] |
| 04:37.23 | *** join/#storm rockstar (n=rockstar@canonical/launchpad/rockstar) [NETSPLIT VICTIM] |
| 05:40.00 | *** join/#storm jukart (i=lovely@81.189.156.94) |
| 06:48.58 | *** join/#storm jukart (i=lovely@81.189.156.94) |
| 06:49.55 | *** join/#storm kanika_vats (n=kanika_v@59.180.103.159) |
| 06:56.09 | *** join/#storm fcorrea (n=fcorrea@c953f0f2.virtua.com.br) |
| 08:54.18 | *** join/#storm shaunm (n=shaunm@c-98-212-133-244.hsd1.il.comcast.net) [NETSPLIT VICTIM] |
| 08:54.18 | *** join/#storm sidnei (n=sidnei@plone/dreamcatcher) [NETSPLIT VICTIM] |
| 08:54.18 | *** join/#storm rockstar (n=rockstar@canonical/launchpad/rockstar) [NETSPLIT VICTIM] |
| 11:02.11 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 11:33.41 | *** join/#storm kanika_vats (n=kanika_v@59.180.104.165) |
| 11:57.11 | *** join/#storm kanika_vats (n=kanika_v@59.180.28.58) |
| 12:26.38 | *** join/#storm kanika_vats (n=kanika_v@59.180.70.149) |
| 12:53.35 | *** join/#storm niemeyer (n=niemeyer@200.181.169.207) |
| 14:01.04 | *** join/#storm shaunm (n=shaunm@proxyserver.wolfram.com) |
| 14:05.57 | fcorrea | jamesh: Hello there. Just tried to reproduce what I said about using the storm utility in zope2. The problem is related to the z3 configuration inside storm which uses a unknown zcml parameter for z2 (like_class) |
| 14:06.14 | fcorrea | ConfigurationError: like_class parameter not supported. |
| 14:06.38 | fcorrea | so, using the default storm utility in zope2 is a no go |
| 14:26.26 | jamesh | fcorrea: okay, thanks. Could you try changing that <require/> line to the same as the <allow/> line from the previous stanza? |
| 14:26.49 | fcorrea | jamesh: sure |
| 14:27.14 | jamesh | I believe that's what the like_class syntax is designed to do: copy the checkers from one class to another |
| 14:29.16 | jamesh | fcorrea: if that solves the problem, please file a bug saying so and include the version of Zope you're using |
| 14:36.00 | jamesh | hmm. I can see references to like_class going back til the earliest versions of zope.app.component though |
| 14:36.29 | jamesh | well, at least to the start of zope/app/component/metadirectives.py's history |
| 14:37.01 | jamesh | so versions would be very useful |
| 14:39.35 | fcorrea | jamesh: right. In zope2, metadirectives.py is provided by Five |
| 14:39.52 | fcorrea | but it seems to be working now tho |
| 14:40.08 | fcorrea | will file a bug report |
| 14:42.43 | fcorrea | jamesh: I mean, it seems to be working under the instructions you gave about copying the same <allow /> from the previous one |
| 14:43.28 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 14:43.50 | jamesh | gotcha. If you can include a trace back in the bug report, that'd be great. |
| 14:44.10 | jamesh | feel free to censor the outer frames if it is private code |
| 14:49.45 | niemeyer | jamesh: Yo! |
| 14:49.49 | niemeyer | jamesh: Had a good trip back home? |
| 14:50.06 | jamesh | niemeyer: pretty good, as far as day-long trips can be. |
| 14:50.13 | niemeyer | jamesh: Nice :) |
| 14:50.19 | jamesh | it was 1:30am monday when I arrived home |
| 14:50.28 | niemeyer | jamesh: Ugh |
| 14:50.40 | jamesh | how about you? |
| 14:50.56 | niemeyer | jamesh: I got home at 8pm.. some bad weather reached me and there was some convoluted back and forth |
| 14:51.22 | niemeyer | jamesh: But all good otherwise |
| 14:52.19 | niemeyer | jamesh: I'm wondering about our conversations on Storm.. we should start some actions about it |
| 14:52.28 | jamesh | fcorrea: http://svn.zope.org/Zope/trunk/src/Products/Five/metaconfigure.py?rev=99145&view=auto this looks like the relevant Five code |
| 14:53.12 | jamesh | fcorrea: looks like they've recently changed the error to a warning but that would still leave the class without the checker, so we'd still need to modify the ZCML |
| 14:53.49 | jamesh | niemeyer: yeah. If you want, I can send an email to the list tomorrow with some of the ideas we discussed. |
| 14:53.57 | jamesh | if you're too busy, that is. |
| 14:54.33 | niemeyer | jamesh: Oh, that would be wonderful! |
| 14:54.46 | niemeyer | jamesh: Well, busy we're all always :-) |
| 14:55.12 | niemeyer | jamesh: But you had some nice perspective on the ideas, so it'd be wonderful if you could bootstrap some discussion |
| 15:02.36 | fcorrea | jamesh: reported. Thanks for the help |
| 15:21.40 | *** join/#storm kanika_vats (n=kanika_v@59.180.45.145) |
| 15:46.28 | *** join/#storm kanika__vats (n=kanika_v@59.180.54.103) |
| 16:06.58 | *** join/#storm spats (n=spats@mail.fitness2live.com.au) |
| 16:19.11 | *** join/#storm jukart (n=jukart@d86-32-163-25.cust.tele2.at) |
| 18:20.42 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 18:22.53 | *** join/#storm thumper (n=tim@canonical/launchpad/thumper) |
| 20:23.14 | *** join/#storm bac (n=bac@cpe-065-190-186-163.nc.res.rr.com) |
| 20:38.51 | *** join/#storm Zer2 (i=Zer@98.220.251.193) |
| 20:41.55 | Zer2 | Hey... I was trying to use Storm with SQLite on Windows. If I use / in the paths it doesn't seem to open the right file, but if I use \\, it assumes the : in the drive letter name is part of the URL. Any idea? |
| 20:44.19 | Zer2 | Hmm. Actually |
| 20:45.06 | Zer2 | The docs really should be updated to reflect this... unlike what the tutorial suggests, you need to just use sqlite: instead of sqlite:// (not scheme:// with scheme being sqlite, postgres, etc..) |
| 20:45.29 | sidnei | Zer2: i believe sqlite:/// works too |
| 20:45.53 | sidnei | (ie, scheme://host/ with an empty host) |
| 20:47.07 | Zer2 | Hmm. |
| 20:47.10 | Zer2 | So it'd be sqlite:///? |
| 20:47.35 | Zer2 | Let me try that. I was using sqlite://, just assuming that it'd figure it was a file |
| 20:48.25 | Zer2 | Yep, that works too |
| 20:50.55 | Zer2 | By the way, does Storm create transactions automatically? I may not need that always |
| 20:56.39 | mzz | it does |
| 21:00.04 | Zer2 | No way around it? There are some .finds where I really don't need it |
| 21:59.08 | *** join/#storm fcorrea (n=fcorrea@c953f0f2.virtua.com.br) |
| 22:37.30 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) [NETSPLIT VICTIM] |