2007-06-23T00:08:42 <grzywacz> This is going to be the most broken push ever...
2007-06-23T00:09:13 <grzywacz> (sorry)
2007-06-23T00:09:15 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2257:463dfc178337 1.7-jabber-knowak/MoinMoin/events/ (notification.py messages.py): Rename messages.pu to notification.py
2007-06-23T00:09:16 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2258:062b76cf8d48 1.7-jabber-knowak/MoinMoin/ (6 files in 4 dirs):
2007-06-23T00:09:16 <CIA-27> moin: Introduce an base EventResult class for event handler return values.
2007-06-23T00:09:16 <CIA-27> moin: Make send_notification use a list of jids, instead of one.
2007-06-23T00:09:17 <CIA-27> moin: I'll remove prints tomorrow.
2007-06-23T00:09:19 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2259:10ec0e8d35e1 1.7-jabber-knowak/jabberbot/ (commands.py xmlrpcbot.py xmppbot.py): Make send_notification accept a list of JIDs, instead of one.
2007-06-23T00:09:22 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2260:55196f5e7000 1.7-jabber-knowak/MoinMoin/events/ (notification.py messages.py): messages.py -> notification.py; a broken commit
2007-06-23T00:09:25 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2261:c2309607a996 1.7-jabber-knowak/MoinMoin/events/notification.py: Merge with devel...
2007-06-23T00:09:52 <xorAxAx> ???
2007-06-23T00:10:12 <xorAxAx> why do you push broken commits? you can fix them locally first :)
2007-06-23T00:10:40 <grzywacz> I don't know how to rollback more than one.
2007-06-23T00:10:49 <grzywacz> And I'm falling asleep atm.
2007-06-23T00:10:50 <zenhase> grzywacz: look in mercurial tips
2007-06-23T00:11:07 <zenhase> grzywacz: there is something to merge several changesets into one
2007-06-23T00:11:15 <grzywacz> Ah, those. Ok. I'd fsck it up in this state of mind, though...
2007-06-23T00:11:32 <zenhase> it's basically hg update <old-rev>; hg revert -r tip; hg ci
2007-06-23T00:12:10 <zenhase> and then clone it to get rid of the dangling changesets if you want
2007-06-23T00:12:43 <grzywacz> Hm, oh well. :\
2007-06-23T00:14:02 <grzywacz> Good night. :\
2007-06-23T00:14:09 <zenhase> hey, cheer up
2007-06-23T00:14:16 <zenhase> have a good night :>
2007-06-23T00:55:02 <lanius> xorAxAx: i wonder what the best way is to define constants in python
2007-06-23T00:55:15 <lanius> xorAxAx: add them as class variable to StorageBackend
2007-06-23T00:55:25 <lanius> xorAxAx: or just as module variables in interfaces.py
2007-06-23T00:55:30 <lanius> xorAxAx: or create a new class for them
2007-06-23T00:55:51 <xorAxAx> module vars or class vars
2007-06-23T00:55:56 <xorAxAx> choose one
2007-06-23T00:56:03 <xorAxAx> but no new class please :)
2007-06-23T00:56:15 <lanius> what's the better name metadata_deleted or just deleted
2007-06-23T00:58:19 <xorAxAx> hmm, no idea
2007-06-23T00:58:55 <lanius> ThomasWaldmann: any preference?
2007-06-23T01:00:23 <ThomasWaldmann> hm?
2007-06-23T01:01:30 <ThomasWaldmann> DELETED maybe?
2007-06-23T01:02:05 <ThomasWaldmann> otoh, having a constant DELETED for the string 'deleted' is a bit pointless
2007-06-23T01:04:05 <xorAxAx> ThomasWaldmann: the idea is that this avoids checks like "if 'deletteeeedd' in metadata:" and gives an attributeerror instead
2007-06-23T01:04:26 <xorAxAx> also, its a strong plead for standardised key names
2007-06-23T01:22:10 <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2092:b3abb39a15bc 1.7-storage-hwendel/MoinMoin/storage/ (backends.py external.py fs_moin16.py interfaces.py): add constant for deleted flag, has_item returns None instead of False
2007-06-23T02:26:34 <CIA-27> moin: Florian Krupicka <florian.krupicka@googlemail.com> * 2065:e4bfd2e69c2d 1.7-template-fkrupicka/MoinMoin/templating/legacy.py: Put the clocks in LegacyWrapper too ...
2007-06-23T02:26:37 <CIA-27> moin: Florian Krupicka <florian.krupicka@googlemail.com> * 2066:eac00a1520df 1.7-template-fkrupicka/MoinMoin/Page.py: even more clocks, since they are unexpensive
2007-06-23T02:26:38 <CIA-27> moin: Florian Krupicka <florian.krupicka@googlemail.com> * 2067:73f17af4be7e 1.7-template-fkrupicka/MoinMoin/ (5 files in 4 dirs): Moved _genshitheme to seperate theme loading code (no more plugin-meddling)
2007-06-23T03:17:52 <zenhase> i put cProfile into the the Theme now, directly around the call to .generate
2007-06-23T03:18:08 <zenhase> and store away the profile data in a directory
2007-06-23T03:18:25 <zenhase> let's see what accumlated data says
2007-06-23T11:53:40 <grzywacz> moin
2007-06-23T12:06:58 <dreimark> moin
2007-06-23T13:01:19 <zenhase> moin
2007-06-23T13:04:47 <dreimark> bbl
2007-06-23T15:10:35 <ThomasWaldmann> moin
2007-06-23T19:37:51 <lanius> ThomasWaldmann: is defining the domain (underlay, standard, all) in some methods really required?
2007-06-23T19:49:35 <zenhase> hi fabi
2007-06-23T19:49:40 <Fabi> moin
2007-06-23T19:49:54 <xorAxAx> lanius: what do you mean by that?
2007-06-23T19:50:12 <Fabi> zenhase, do you have an idea now where the performace issues are?
2007-06-23T19:51:22 <zenhase> Fabi: in the old design or in the genshi-code?
2007-06-23T19:51:29 <Fabi> both
2007-06-23T19:51:43 <lanius> xorAxAx: e.g. page_exists(domain = ...)
2007-06-23T19:51:48 <lanius> page.exists(
2007-06-23T19:52:08 <lanius> if we really need that or if i can kick that argument
2007-06-23T19:52:15 <xorAxAx> lanius: domain would be a layering violation
2007-06-23T19:52:22 <xorAxAx> hmm
2007-06-23T19:52:29 <xorAxAx> or how is it defined generally?
2007-06-23T19:52:54 <lanius> i only found it used one time, in RecentChanges, have to look closer if it is needed there
2007-06-23T19:54:49 <lanius> i don't know why it is used there, i think it can be removed
2007-06-23T19:55:43 <xorAxAx> generally, it would be nice if the upper levels could differentiate if a page is on a specific level, though
2007-06-23T19:55:56 <xorAxAx> but that would be a large change
2007-06-23T19:56:31 <lanius> i introduced a name argument for the backend, so you can differentiate for each item where it comes from
2007-06-23T19:56:51 <lanius> but you cannot only query for a item in one backend
2007-06-23T19:56:53 <xorAxAx> but an item might be on multiple levels
2007-06-23T19:56:58 <lanius> rihgt
2007-06-23T19:57:06 <lanius> that's the problem
2007-06-23T19:57:07 <xorAxAx> which is the pain the current design has
2007-06-23T19:57:25 <lanius> which design do you mean?
2007-06-23T19:57:30 <lanius> with current
2007-06-23T19:59:32 <zenhase> Fabi: in the old design much time is eaten by theme.send_title it seems
2007-06-23T19:59:52 <zenhase> in genshi it's obviously the .generate step
2007-06-23T20:00:35 <zenhase> old design: total: 0.033s, send_title: 0.021s
2007-06-23T20:01:39 <xorAxAx> lanius: moin 1.5
2007-06-23T20:01:44 <zenhase> regarding genshis .generate times i can't really give meaningful numbers right now, since my test-wiki runs c-profile around the .generate-call to accumulate some data
2007-06-23T20:02:21 <zenhase> i still try to think about how to speed up generate here
2007-06-23T20:02:31 <lanius> xorAxAx: so you think it is necessary to have such functionality?
2007-06-23T20:03:10 <xorAxAx> lanius: if one wanted to add UI to make the current experience saner, then ye
2007-06-23T20:03:11 <xorAxAx> s
2007-06-23T20:08:04 <zenhase> i am currently working on the proxy objects for the templates
2007-06-23T20:08:10 <zenhase> page, user and the like
2007-06-23T20:08:37 <zenhase> i would like to do them first and then make the base-template more sophisticated
2007-06-23T20:09:03 <zenhase> then i can do better profiling what exactly eats time in genshis .generate
2007-06-23T20:10:19 <zenhase> currently the cProfile readings don't give me real impressions
2007-06-23T20:34:54 <Fabi> don't waste too much time there
2007-06-23T20:35:13 <zenhase> Fabi: you mean profiling?
2007-06-23T20:35:16 <Fabi> If you got an idea keep that in mind and then focus on the integration work
2007-06-23T20:35:19 <Fabi> yup
2007-06-23T20:35:24 <zenhase> i don't
2007-06-23T20:35:40 <zenhase> i installed in the the theme-code and let it write profile-data into a directory
2007-06-23T20:35:47 <zenhase> so can look at it later :>
2007-06-23T20:35:57 <Fabi> oki
2007-06-23T20:36:09 <zenhase> and while the design evolves delete old profile data and keep just the new
2007-06-23T21:04:41 <grzywacz> xorAxAx, is it ok to add a module-level Logger objects for MoinMoin.events?
2007-06-23T21:06:23 <xorAxAx> why not?
2007-06-23T21:07:02 <grzywacz> No idea, I though there might have been a different preferred solution.
2007-06-23T21:07:06 <grzywacz> *thought
2007-06-23T21:07:15 <grzywacz> Going with a module-wide one, then.
2007-06-23T21:07:21 <xorAxAx> i think thats how its meant to be used
2007-06-23T21:07:23 <xorAxAx> AFAIK
2007-06-23T21:07:43 <grzywacz> No experience on my part yet. ;)
2007-06-23T21:12:44 <grzywacz> Hm.
2007-06-23T21:12:45 <ThomasWaldmann> re
2007-06-23T21:13:09 <grzywacz> re
2007-06-23T21:13:20 <grzywacz> File "/usr/lib64/python2.3/logging/__init__.py", line 1036, in callHandlers
2007-06-23T21:13:21 <grzywacz> if record.levelno >= hdlr.level:
2007-06-23T21:13:21 <grzywacz> AttributeError: class StreamHandler has no attribute 'level'
2007-06-23T21:13:23 <grzywacz> Broken. 8)
2007-06-23T21:14:05 <grzywacz> (by me :D)
2007-06-23T21:15:06 <grzywacz> I think that's the only thing I'm often getting trapped by in Python - no compile-time type checking. ;)
2007-06-23T21:17:14 <xorAxAx> you could use hungarian notation ...
2007-06-23T21:17:21 <xorAxAx> ... and automatically fail SOC :-)
2007-06-23T21:17:42 <ThomasWaldmann> <eg>
2007-06-23T21:17:47 <grzywacz> lol
2007-06-23T21:17:55 <grzywacz> No, thank you for your kind advice. ;)
2007-06-23T21:18:23 <xorAxAx> rpython has even type inference ... so you get an obfuscated error message if the type doesnt match at compile time :)
2007-06-23T21:18:53 <grzywacz> Zope interfaces? ;)
2007-06-23T21:19:26 <xorAxAx> thats unrelated
2007-06-23T23:28:46 <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2094:c8ab29158abf 1.7-storage-hwendel/MoinMoin/ (Page.py config/multiconfig.py): exists() uses new storage engine
2007-06-23T23:28:48 <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2095:df8d97109eb4 1.7-storage-hwendel/MoinMoin/ (Page.py PageEditor.py action/Load.py storage/fs_moin16.py): more storage integration; locking currently does not work
2007-06-23T23:35:55 <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2096:922be3113175 1.7-storage-hwendel/MoinMoin/ (Page.py storage/fs_moin16.py storage/interfaces.py): add size() to DataBackend and call it from Page.size
2007-06-23T23:38:00 <lanius> xorAxAx: ThomasWaldmann: i'd like to remove isWriteable from Page, i don't think it is really neede
2007-06-23T23:38:23 <xorAxAx> lanius: what is using it?
2007-06-23T23:38:32 <xorAxAx> it never really occured to me
2007-06-23T23:38:39 <lanius> newpage, PageEditor and PageGraphicalEditor
2007-06-23T23:38:54 <lanius> return os.access(self._text_filename(), os.W_OK) or not self.exists()
2007-06-23T23:38:56 <lanius> that's all it does
2007-06-23T23:39:05 <xorAxAx> ok, then its fine
2007-06-23T23:39:12 <lanius> to remove?
2007-06-23T23:39:22 <xorAxAx> yes
2007-06-23T23:39:32 <lanius> good
2007-06-23T23:39:51 <lanius> i think if the file itself is not writable that's a configuration error
2007-06-23T23:40:34 <xorAxAx> yes
2007-06-23T23:42:42 <CIA-27> moin: Heinrich Wendel <h_wendel@cojobo.net> * 2097:6b4ba5245512 1.7-storage-hwendel/MoinMoin/ (5 files in 3 dirs): remove isWritable()
2007-06-23T23:44:44 <lanius> already cut Page.py 300 lines or 16% down :-)
2007-06-23T23:45:03 <xorAxAx> thats not much :)
2007-06-23T23:45:24 <xorAxAx> i guess that you will be able to cut another few hundred lines :)
2007-06-23T23:45:28 <zenhase> every line helps! :>
2007-06-23T23:45:55 <lanius> sure, i think i am only through 1/3 of the Page.py code, that would make another 600 lines ;)
2007-06-23T23:46:04 <xorAxAx> hehe
2007-06-23T23:46:41 <lanius> which then would be a bit under 1000, sounds like a nice goal
MoinMoin: MoinMoinChat/Logs/moin-dev/2007-06-23 (last edited 2007-10-29 19:13:01 by localhost)
