| 02:12.23 | *** join/#storm tjs (n=tjs@203.206.162.25) |
| 02:12.26 | tjs | G'day |
| 02:12.55 | tjs | is there a hook for just before save? |
| 02:48.36 | jamesh | tjs: you mean before commit? |
| 02:49.09 | jamesh | or before a particular object gets flushed? |
| 02:54.23 | tjs | I just need to update some stuff for a given type before it gets saved |
| 02:59.52 | jamesh | again, what do you mean by saved? |
| 03:00.27 | jamesh | do you mean flushed (data sent to the database) or committed (database transaction committed and the results visible to other transactions) |
| 03:02.28 | tjs | flushed |
| 03:03.07 | jamesh | add a __storm_pre_flush__() method to your class |
| 03:03.12 | tjs | I'm persisting marker interfaces on objects |
| 03:03.22 | tjs | that have been added with alsoProvides() |
| 03:03.29 | tjs | aah cool |
| 03:03.34 | tjs | ty :) |
| 03:04.11 | jamesh | note that only dirty objects get flushed though |
| 03:04.22 | tjs | ah |
| 03:04.24 | tjs | hm |
| 03:05.01 | jamesh | there is a Store._set_dirty(object) method |
| 03:05.31 | jamesh | but it'd be best to bring up your use case on the mailing list first rather than using Storm internals |
| 03:06.09 | tjs | yeh.. but Id have to monkey patch z.i.alsoProvides to set the dirty flag |
| 03:06.51 | jamesh | how do other systems know to persist this info? |
| 03:07.35 | tjs | erm.. I imagine its checked when the zodb goes to pickle an object explicitly |
| 03:08.51 | tjs | just looking it up now |
| 03:46.53 | tjs | ok this is fairly disgusting.. |
| 03:47.15 | tjs | but alsoProvides sticks a descriptor on object.__provides__ |
| 03:48.05 | tjs | so I'll make sure my object already has a __provides__ descriptor, which takes the z.i descriptor and assigns it to __provides2__ or something, and relays to it |
| 03:48.21 | tjs | but at the same time sets the dirty flag on object |
| 03:49.08 | tjs | then any object with alsoProvides called on it will be dirty, and then the rest of this hideous hack will work |
| 05:11.20 | *** join/#storm jukart (n=jukart@85-124-221-45.static.xdsl-line.inode.at) |
| 05:48.46 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 06:33.47 | *** join/#storm dobee (n=dobeee@194.183.146.186) |
| 08:25.00 | *** join/#storm bozzo (n=chatzill@85.10.14.223) |
| 08:41.45 | *** join/#storm dobee (n=dobeee@194.183.146.186) |
| 09:17.29 | *** join/#storm bozzo (n=chatzill@85.10.14.223) |
| 09:31.36 | *** join/#storm jamesh (n=james@canonical/launchpad/jamesh) |
| 09:42.17 | *** join/#storm mcella (n=michele@ip-176-22.sn2.eutelia.it) |
| 10:07.26 | *** join/#storm Fujitsu (n=fujitsu@ubuntu/member/fujitsu) |
| 12:28.22 | *** join/#storm niemeyer (n=niemeyer@200-163-194-205.ctame705.dsl.brasiltelecom.net.br) |
| 13:25.28 | bozzo | radix: I can emulate subclassing perfectly by liberal use of Proxy() on a fake base class that just handles the primary_FK and contains the base class |
| 13:52.04 | radix | hooray |
| 13:57.03 | *** join/#storm EnTeQuAk (n=EnTeQuAk@p57B9BC16.dip0.t-ipconnect.de) |
| 14:00.29 | bozzo | Hmm it seems I cannot do a Reference() against a Proxy(), how hard would it be to implement ProxyReference() ? |
| 14:01.20 | bozzo | Proxy() does not work on a Reference() either |
| 14:06.34 | *** join/#storm niemeyer (n=niemeyer@200-163-194-205.ctame705.dsl.brasiltelecom.net.br) |
| 14:21.10 | *** join/#storm brosner (n=brian@63.150.173.39) |
| 15:41.21 | *** join/#storm hagna (n=hagna@70.98.162.62) |
| 15:45.16 | hagna | to use storm synchronously in twisted do you just use it like normal |
| 15:45.28 | radix | synchronously? sure |
| 15:45.43 | hagna | radix: and for async spawnProcess maybe |
| 15:45.52 | hagna | that's why I asked that in #twisted |
| 15:45.55 | radix | well, that's not really a solution for "using storm asynchronously" |
| 15:46.05 | hagna | it's a practice |
| 15:46.13 | hagna | not the best one |
| 15:46.15 | radix | if you're putting it off into another process, you're just writing a second program which uses storm synchronously |
| 15:46.19 | radix | and calling it from the asynchronous program |
| 15:47.34 | hagna | radix: right and what's wrong with that? |
| 15:47.55 | hagna | are you wearing a computer? |
| 15:48.01 | radix | nothing, but it's not "using storm asynchronously" |
| 15:48.29 | radix | no |
| 15:48.34 | radix | I'm at home, using a laptop |
| 15:49.30 | hagna | radix: ok just checking. I imagined you in that show about wearable computers at MIT |
| 15:49.39 | radix | heh |
| 15:59.41 | *** join/#storm rafael (n=rafael@moinmoin/fan/rafael) |
| 16:31.58 | *** join/#storm jamesh_ (n=james@canonical/launchpad/jamesh) |
| 16:35.29 | *** join/#storm jukart (n=jukart@lsfw01.lovelysystems.com) |
| 16:47.51 | *** join/#storm metatheorem (n=metatheo@cpe-66-66-196-45.rochester.res.rr.com) |
| 17:58.03 | *** join/#storm dobee (n=dobeee@81-223-53-162.dornbirn.xdsl-line.inode.at) |
| 18:27.48 | *** join/#storm dobee (n=dobeee@81-223-53-162.dornbirn.xdsl-line.inode.at) |
| 19:59.52 | *** join/#storm dobee (n=dobeee@81-223-53-162.dornbirn.xdsl-line.inode.at) |
| 20:01.29 | *** join/#storm dobee (n=dobeee@81-223-53-162.dornbirn.xdsl-line.inode.at) |
| 20:34.03 | *** join/#storm bozzo (n=chatzill@BSN-210-221-20.dial-up.dsl.siol.net) |
| 23:59.52 | *** join/#storm wallflower (n=wallflow@216.113.246.205) |