MoinMoin: MoinMoinChat/Logs/moin-dev/2008-04-01

2008-04-01T01:21:42  <ThomasWaldmann> Student Applications Open - New Deadline April 7th: http://tinyurl.com/2czy6o - See FAQ for updated timeline
2008-04-01T05:23:26  <HongJun> ThomasWaldmann: Hi
2008-04-01T05:25:24  <HongJun> ThomasWaldmann:  has  the Semantic Moin been implemented
2008-04-01T05:45:27  <xqj37> the fabric of human society is the individual
2008-04-01T05:45:32  <xqj37> all hail the individual ;)
2008-04-01T05:57:31  <xqj37> and I would NOT say that there are few advancements in artificial intelligence
2008-04-01T05:58:17  <xqj37> there are many
2008-04-01T05:58:31  <xqj37> but I guess I should sleep now
2008-04-01T06:04:24  <HongJun> gn
2008-04-01T08:30:46  <gizmach> moin
2008-04-01T08:49:19  <ThomasWaldmann> moin
2008-04-01T08:54:05  <ThomasWaldmann> gizmach: i suggest you don't touch user profile storage in your project
2008-04-01T09:23:05  <dreimark> why ?
2008-04-01T09:29:10  <gizmach> ThomasWaldmann: morning and why?
2008-04-01T09:29:14  <gizmach> dreimark: morning
2008-04-01T09:33:55  * gizmach will be back in a half an hour all comments and sugestions for the project are wellcome 
2008-04-01T09:35:13  <dreimark> moin
2008-04-01T09:36:57  * dreimark added a comment
2008-04-01T09:46:57  <dreimark> btw. now we have 20 applications
2008-04-01T09:49:42  <gizmach> dreimark: where you added a comment ?
2008-04-01T09:49:58  <dreimark> gizmach: for ThomasWaldmann
2008-04-01T09:50:53  <gizmach> dreimark: oh, pardon, I taught it's on my project proposal, so I go and I haven't seen anything and I taught it's april fools joke or something
2008-04-01T09:51:09  <dreimark> summary of my ideas i got this night based on the discussion we had yesterday
2008-04-01T09:51:31  <gizmach> dreimark: :) thank you
2008-04-01T09:51:48  <dreimark> gizmach: bad you mentioned this day so i can't make a joke here
2008-04-01T09:52:27  <gizmach> dreimark: sorry :)
2008-04-01T09:58:07  * dreimark is back after lunch in some hours
2008-04-01T10:50:35  <zenhase> moin
2008-04-01T10:52:02  <zenhase> wow, april fools all over my feed reader
2008-04-01T10:55:26  <dreimark> http://www.pro-linux.de/news/2008/12513.html
2008-04-01T11:03:27  <gizmach> back
2008-04-01T12:06:13  <ThomasWaldmann> gizmach: the storage of the user profiles will be done by the new storage backend
2008-04-01T12:07:25  <ThomasWaldmann> but as this is not merged yet, you can't base your work on it
2008-04-01T12:46:19  <gizmach> ThomasWaldmann: then the only thing I can actually do is to implement mapper api for reading users from LDAP and all databases I plan to support. But wouldn't it be nice to implement storring user profiles to e.g LDAP and to add support to my implemented api for the new storrage backend when it will be implemented and to make it work but while it's not merged to enable this. I'm probbably wrong and this would be a double job, but I'll have already working
2008-04-01T12:53:02  <ThomasWaldmann> your text was truncated after "have already working"
2008-04-01T12:56:11  <ThomasWaldmann> gizmach: tasks should not depend on each other. if they do, you are in trouble if the other one fails or takes longer than expected.
2008-04-01T12:57:52  <ThomasWaldmann> but I guess refactoring all the groups/dicts/lists code to a more generic approach is also quite some work
2008-04-01T13:03:54  <gizmach> ThomasWaldmann: I triend to say something like that. Beacue I will already have an api for mapping in one way (ldap-moin) I could refactor all groups and other user data to make the mapping in other way with a new backend storrage easier. I know that the dependance of the task is bad.
2008-04-01T13:05:28  <gizmach> but in some case that the new storrage backend will have some problems it would be easier to just add some methods to my api(mapper) to add support for exporting user data to some other suitable storrage backend
2008-04-01T13:08:11  <gizmach> I don't know how it was done before with the gsoc, because it's my first time to apply for it, but I will probbably consider that part of code (api) as my obligation after gsoc. I don't know what is your policy in those cases.
2008-04-01T13:08:58  <gizmach> so if the new storage backend will  be done during gsoc I will probably add to my generic model new methods for exporting user data.
2008-04-01T13:11:49  <ThomasWaldmann> well, help is always welcome :)
2008-04-01T13:12:35  <ThomasWaldmann> gizmach: did you already have a look at wikidicts.py?
2008-04-01T13:17:40  <gizmach> I only see that for a short while but I didn't try to get familliar to source code.
2008-04-01T13:18:09  * gizmach just opened wikidicts.py again and browsing code
2008-04-01T13:18:24  <ThomasWaldmann> it's a bit "twisted" X)
2008-04-01T13:19:16  <ThomasWaldmann> you also have to have some thoughts about performance
2008-04-01T13:19:23  <gizmach> ThomasWaldmann: oh it's your source code
2008-04-01T13:19:41  <ThomasWaldmann> doing sql and ldap queries takes time (like reading and parsing wiki pages)
2008-04-01T13:19:41  <gizmach> please can you explain me the last sentence
2008-04-01T13:19:46  <gizmach> yes
2008-04-01T13:19:48  <gizmach> I see
2008-04-01T13:19:59  <gizmach> forget my proposal for explanation
2008-04-01T13:20:30  <ThomasWaldmann> acl processing happens often in moin, so you need some way to cache group information
2008-04-01T13:20:54  <ThomasWaldmann> while detecting when the authoritaive source changes groups
2008-04-01T13:21:34  * ThomasWaldmann has no idea how that would work for sql/ldap
2008-04-01T13:25:57  <gizmach> well for the performance problem I saw django built web sites working very good with all those sql querries all the time, at the moment I'm not sure if that would work just fine but i something can work just fine , we could implement something with simmilar performance. Ok my sentences probably sounds not understandable.
2008-04-01T13:27:18  <ThomasWaldmann> well, we first could try without caching and see
2008-04-01T13:29:37  <gizmach> ThomasWaldmann: yes that would be easier because with caching I'll need something like detecting querry while something changed, because temporarily checking will not work good
2008-04-01T13:31:22  <ThomasWaldmann> maybe thing about dropping stage 2 and adding stage 0 "refactor wikidicts"
2008-04-01T13:31:39  <ThomasWaldmann> s/thing/think/
2008-04-01T13:33:12  <gizmach> yes that's probbably a good idea, and to leave stage 2 for some other time or if new sotorrage backen will be done
2008-04-01T13:33:28  <ThomasWaldmann> (refactoring it in a similar way than we are doing for the storage code: have some api to some generic groups/dicts/lists code that has another backend api to more simple backend plugins for misc group sources)
2008-04-01T13:35:45  <ThomasWaldmann> btw, there is already some mysql_group auth plugin
2008-04-01T13:36:23  <gizmach> So you think I could try my code working with it ?
2008-04-01T13:36:36  <gizmach> to work wit it*
2008-04-01T13:36:47  <ThomasWaldmann> it's quite special and there maybe should be also some more generic auth plugin authenticating a user again a db server (similar as ldap does)
2008-04-01T13:36:57  <ThomasWaldmann> +st
2008-04-01T13:38:04  <ThomasWaldmann> and I guess you will have to implement that. having group code check against sql db without having auth also check against sql db doesn't make that much sense.
2008-04-01T13:38:19  <gizmach> so your point is to make as much things as I can to de everything more generic
2008-04-01T13:38:24  <ThomasWaldmann> (for ldap this is already done)
2008-04-01T13:38:34  <gizmach> yes I saw that
2008-04-01T13:39:11  <gizmach> well at least I see there is a little more work to do to put it in my project prpoposal
2008-04-01T13:39:16  <ThomasWaldmann> yes, you'll have to refactor and cleanup that code (or your additions will just make a bigger mess)
2008-04-01T13:39:39  <gizmach> :) well yes :)
2008-04-01T13:40:02  * gizmach just immagining how it will look like just to add her own code
2008-04-01T13:43:27  <ThomasWaldmann> btw, what's the difference between gizmo and gizmach?
2008-04-01T13:44:02  <ThomasWaldmann> hah, i see tabs in your code X)
2008-04-01T13:44:27  <gizmach> well I tried to register my nickname as gizmo
2008-04-01T13:44:47  <gizmach> but it croatian it will sound like a little gizmo - gizmach :D
2008-04-01T13:45:17  <ThomasWaldmann> ah :)
2008-04-01T13:45:19  <gizmach> well I'll try to write nice code :)
2008-04-01T13:45:50  <gizmach> and what's wrong with tabs? :)
2008-04-01T13:46:28  <ThomasWaldmann> for python they would be a constant source of trouble
2008-04-01T13:47:06  <ThomasWaldmann> (that's not true for a single developer ONLY using tabs, but as you will work in a group and everyone else is using 4 blanks, it is true)
2008-04-01T13:47:44  <ThomasWaldmann> and we have a unit test that will fail for source with tabs
2008-04-01T13:48:06  <ThomasWaldmann> which editor do you use for python?
2008-04-01T13:48:12  <gizmach> I I see
2008-04-01T13:48:14  <gizmach> vim
2008-04-01T13:48:15  <gizmach> :D
2008-04-01T13:48:22  <gizmach> well I can set my tabs to 4
2008-04-01T13:48:30  <ThomasWaldmann> good, did you see that vim plugin on moinmo.in?
2008-04-01T13:49:00  <gizmach> no I haven't
2008-04-01T13:49:18  <gizmach> I'll installed it immidiately
2008-04-01T13:49:47  <ThomasWaldmann> http://moinmo.in/CodingStyle read that, until you can recite it while sleeping :D
2008-04-01T13:51:09  <gizmach> hehe ok :) I really need to improve and find my coding style
2008-04-01T13:51:09  <ThomasWaldmann> the vim pep8 plugin is really nice. it will make your code look quite RED, jumping in your face, for place not conforming to the styleguide
2008-04-01T13:51:48  * ThomasWaldmann uses vim with that plugin
2008-04-01T13:52:04  <gizmach> yes I wanted to ask why the version of moinmoin in ubuntu is 1.5? I taught it would have some newer version
2008-04-01T13:52:15  <ThomasWaldmann> i even found an indenting error in your salmostalni_OOP code :)
2008-04-01T13:52:33  <gizmach> probably
2008-04-01T13:52:36  <gizmach> :)
2008-04-01T13:52:45  <ThomasWaldmann> line 22, mixed tabs and blanks
2008-04-01T13:53:10  <gizmach> I done it in a hurry so that's one of my ugliest code ever I think
2008-04-01T13:53:18  <gizmach> or maybe I done some more ugly
2008-04-01T13:53:30  <ThomasWaldmann> hehe
2008-04-01T13:53:41  <ThomasWaldmann> ok, gtg. bbl.
2008-04-01T13:54:50  <gizmach> bye and thx
2008-04-01T14:13:56  <ThomasWaldmann> gizmach: distributions are always behind
2008-04-01T14:15:36  <xqj37> arch linux has 1.5.9
2008-04-01T14:16:18  <xqj37> I'm kinda disappointed ;)
2008-04-01T14:16:32  <xorAxAx> "09:51:10 < dreimark> summary of my ideas i got this night based on the discussion we had yesterday" -- which discussions?
2008-04-01T14:22:08  <gizmach> xqj37: ubuntu is 1.5.7
2008-04-01T14:24:50  <vpv> Fedora 8 has 1.5.8 and the upcoming Fedora 9 has 1.6.2 :)
2008-04-01T14:30:43  * gizmach went to programm in Prolog and learn for exam ... be back in few hours
2008-04-01T14:31:47  <dreimark> xorAxAx: gizmach and me in a query (we should have switched on some point to the dev channel, sorry for that mistake)
2008-04-01T14:34:34  <gizmach> dreimark: please take a look at my discussion with ThomasWaldmann
2008-04-01T14:35:25  <dreimark> gizmach_away: we would have get earlier to that point if we would have discussed it here
2008-04-01T14:38:17  <zenhase> re
2008-04-01T14:38:26  <zenhase> huhu dreimark
2008-04-01T14:38:34  <dreimark> hi zenhase
2008-04-01T14:38:54  <zenhase> sorry if we could not talk yesterday
2008-04-01T14:39:10  <zenhase> you still want to talk about this send_page refactoring?
2008-04-01T14:39:27  * dreimark finds it boring to have always to reread all applications to get if one comment was added
2008-04-01T14:40:40  <zenhase> hmm?
2008-04-01T14:41:17  <dreimark> zenhase: principle yes currently I do review all applications
2008-04-01T14:41:39  <zenhase> ok
2008-04-01T14:41:47  <zenhase> then i don't want to interrupt you :)
2008-04-01T14:41:58  * dreimark misses info or recent changes on the applications
2008-04-01T14:43:38  * zenhase files another application: Make MoinMoin into a webapp for GSoC-like contests
2008-04-01T14:43:48  <zenhase> ;)
2008-04-01T14:44:05  <zenhase> this would be funny on a real metaphysical level
2008-04-01T14:51:04  <dreimark> yep or they do think it't an april fault
2008-04-01T14:51:52  <dreimark> zenhase: I think in  2002 compared today it would have been easier to refactor send_page
2008-04-01T14:54:31  <zenhase> dreimark: definitely
2008-04-01T14:54:42  <zenhase> tripled in code-size iirc
2008-04-01T14:56:25  <zenhase> further delay while constantly adding new features will make it increasingly harder
2008-04-01T14:56:56  <zenhase> regarding that this is a refactoring task which i think is big enough to fill 3 months of work :)
2008-04-01T14:57:55  <dreimark> in some points it is quite similiar to the storage refactoring
2008-04-01T14:58:05  <zenhase> hmm?
2008-04-01T14:58:21  <zenhase> well yes
2008-04-01T14:58:32  <dreimark> it can take longer as one soc project has time
2008-04-01T14:58:46  <zenhase> similar to storage there is much functionality in Page-class, which should be decoupled
2008-04-01T14:59:21  <zenhase> well, i would try my best and would certainly stick around after SoC to finish this
2008-04-01T14:59:45  <zenhase> also since i would like to see my template code integrated in the far future ;)
2008-04-01T14:59:53  <dreimark> yeah
2008-04-01T15:00:13  <zenhase> or better: get the template-project up and running again
2008-04-01T15:00:42  <zenhase> i have to admit, in regard to my 2 proposals, i deem send_page the really hard one
2008-04-01T15:01:20  <zenhase> WSGI may have some small issues that one will run into, but i think it's ultimately more straight-forward
2008-04-01T15:01:48  <zenhase> but the benefits with send_page will be greater in the end imo
2008-04-01T15:02:34  * dreimark has a similiar opninion
2008-04-01T15:03:00  <zenhase> coolest thing would of course be someone refactoring request-layer (WSGI), someone refactoring parser/formatter-layer (like the DOM?) and then send_page refactoring can leverage both those efforts to make it easier
2008-04-01T15:04:20  <zenhase> since much of the problems stem from request and formatting permeating through all Page-code
2008-04-01T15:05:22  <dreimark> yeah we have to kerep in mind that the order of the application task do play a role too
2008-04-01T15:05:35  <zenhase> the order?
2008-04-01T15:05:49  <zenhase> you mean that some have to be finished first?
2008-04-01T15:05:56  <zenhase> before others can benefit?
2008-04-01T15:06:07  <dreimark> right
2008-04-01T15:06:34  <dreimark> it is similiar too to write backends for storage
2008-04-01T15:06:48  <zenhase> yeah
2008-04-01T15:07:21  <zenhase> without the backend you can't start refactoring the old code
2008-04-01T15:07:30  <zenhase> because the backends dictates the new API
2008-04-01T15:08:28  <dreimark> johill: has pointed on that point too.
2008-04-01T15:08:59  <zenhase> hmm
2008-04-01T15:09:50  <zenhase> i am thinking what i should possibly do next
2008-04-01T15:09:50  <johill> well we have a filesys and sql backend now
2008-04-01T15:10:26  <johill> like right now or in your soc work?
2008-04-01T15:10:43  <zenhase> like right now, to flesh out the ideas in my proposals
2008-04-01T15:13:31  <johill> oh. dunno. don't think I read the latest version
2008-04-01T15:13:53  <zenhase> hmm i think you did :)
2008-04-01T15:14:03  <johill> heh ok
2008-04-01T15:14:06  <zenhase> i did not change anything yet on my proposals
2008-04-01T15:14:11  <zenhase> hmm wait
2008-04-01T15:14:21  <zenhase> i had an empty detail-text on WSGI first
2008-04-01T15:14:29  <zenhase> that one is there now, of course
2008-04-01T15:14:37  <zenhase> but otherwise there are no additions
2008-04-01T15:35:43  <xorAxAx> dreimark: ah, no problem :)
2008-04-01T15:40:42  <kikka1> moin
2008-04-01T16:31:05  <zenhase> anyone want to apply for virgle?
2008-04-01T16:38:46  <ThomasWaldmann> johill: maybe check the soc application whether they conflict with what you have already done recently
2008-04-01T17:11:16  <gizmach> re
2008-04-01T19:16:43  <grzywacz> moin
2008-04-01T19:18:35  <dreimark> bbl dinner
2008-04-01T20:04:54  <ThomasWaldmann> re
2008-04-01T20:10:52  <gizmach> wb :)
2008-04-01T21:20:23  <kikka1> re
2008-04-01T22:12:35  <dennda> Hey there
2008-04-01T22:13:02  <dennda> I am sorry for being a bit unresponsive the last few days. I am still doing quite some stuff for university
2008-04-01T22:13:39  <dennda> I'll hopefully post some more code tomorrow, ThomasWaldmann