MoinMoin: MoinMoinChat/Logs/moin-dev/2008-07-07

2008-07-07T00:00:56  <mmihaljevic> well I need to take some sleep
2008-07-07T00:01:14  <mmihaljevic> night
2008-07-07T00:05:47  <xorAxAx> good night :)
2008-07-07T00:06:02  <xorAxAx> have sweet dreams about moin :)
2008-07-07T03:18:34  <CIA-53> Byeongweon [tasyblue@gmail.com] default * 3665:cec23df5a1cf 1.8-guieditor-mbyeongweon/wiki/htdocs/common/js/common.js: disable safari because of html tidy problem on fckeditor
2008-07-07T03:18:35  <CIA-53> Byeongweon [tasyblue@gmail.com] default * 3666:97dd5fb0c75b 1.8-guieditor-mbyeongweon/wiki/htdocs/applets/moinFCKplugins/ (restrict_actions/fckplugin.js selection/fckplugin.js): add comment to select plugin
2008-07-07T07:40:34  <ThomasWaldmann> moin
2008-07-07T07:50:54  <dreimark> moin
2008-07-07T08:25:45  <lesterc> is there a worm that spams moinmoin wikis out there?
2008-07-07T09:22:23  * mmihaljevic is away: bbl
2008-07-07T09:27:59  <dreimark> the reportlab pdf stuff looks nice. it would be interesting to see if we can use it for a formatter
2008-07-07T09:29:17  <TheSheep> I want text justification, with variable character width, leading, and visual aligning of margins :)
2008-07-07T10:22:50  <xorAxAx> dreimark: there are a few reprotlab driven formatters
2008-07-07T15:46:20  <dreimark> xorAxAx: thx,  will look later on it
2008-07-07T15:59:00  <CIA-53> Pawel Pacana <pawel.pacana@gmail.com> default * 4311:1bd0e8e4f5f4 1.8-mercurialbackend-ppacana/MoinMoin/ (12 files in 8 dirs): Merge with 1.8-storage-cdenter
2008-07-07T16:07:48  * PawelPacana invites for freesbie play outside Reval hotel :>
2008-07-07T16:08:27  <dennda> Oh boy
2008-07-07T16:08:40  <dennda> I am teh frisbee master
2008-07-07T16:27:41  <ThomasWaldmann> mmihaljevic: +        return (group != [])
2008-07-07T16:27:52  <ThomasWaldmann> return bool(group)
2008-07-07T16:29:18  <ThomasWaldmann> mmihaljevic: +        group = ldap_object.search_st(self.base_dn, self.scope, search_filter, timeout=self.timeout)
2008-07-07T16:29:37  <ThomasWaldmann> how many attributes is this query returning?
2008-07-07T16:31:35  <waldi> indefinite, not defined maps to '*'
2008-07-07T16:32:38  <waldi> and may not find anything on an openldap server
2008-07-07T16:32:44  <ThomasWaldmann> mmihaljevic: and it is NOT "group" ("group" is a bad name, maybe rather use results = ...)
2008-07-07T16:33:00  <ThomasWaldmann> waldi: why not?
2008-07-07T16:33:32  <waldi> ThomasWaldmann: there exists overlays which creates group entries on-the-fly. this attributes are operational and not included in '*'
2008-07-07T16:33:50  <waldi> this is handy if you want a group "all"
2008-07-07T16:54:08  <ThomasWaldmann> ok
2008-07-07T17:02:40  <waldi> its similar to sql: always specify what you want
2008-07-07T17:21:16  <dennda> Ok guys
2008-07-07T17:21:28  <dennda> Looks I can make it to this years ubucon (hopefully)
2008-07-07T17:22:08  <dennda> And I asked one of my friends of the organizational team; he said there'd be no problem at all for us to get a seperate room / table / whatever
2008-07-07T17:22:17  <dennda> I'll create a wikipage for that
2008-07-07T17:23:35  <xorAxAx> cool
2008-07-07T17:30:30  <dennda> johill: so, if you want to talk about wireless there, you are very welcome to do so :)
2008-07-07T18:04:01  <PawelPacana> dennda: johill: is it stated somewhere that revno is int?
2008-07-07T18:04:41  <dennda> PawelPacana: Revno is int, yes
2008-07-07T18:09:14  <dennda> johill: http://tinyurl.com/62plnl
2008-07-07T19:14:15  * mmihaljevic is back (gone 09:51:53)
2008-07-07T19:14:24  <mmihaljevic> re and pass exam
2008-07-07T19:15:08  <mmihaljevic> ThomasWaldmann: it is returning one result
2008-07-07T19:15:59  <mmihaljevic> because I only search for a group name
2008-07-07T19:45:49  * mmihaljevic goes to play with cache
2008-07-07T20:21:30  <mmihaljevic> xorAxAx: r u there I have some questions about caching
2008-07-07T21:38:53  <mmihaljevic> xorAxAx: about pulling informations you talked to me yesterday
2008-07-07T21:39:47  <mmihaljevic> I'm not sure what you think for it -> I need to track mtime changes?
2008-07-07T21:44:23  <mmihaljevic> brb ~10min food
2008-07-07T21:50:54  <xorAxAx> mmihaljevic: yes, you need to do it like the old code - have you tried to understand it?
2008-07-07T21:59:33  <mmihaljevic> xorAxAx: I'm doing that now (playing with caching code)
2008-07-07T22:00:30  <mmihaljevic> needsUpdate is what I need to look at it if I'm right
2008-07-07T22:00:35  <mmihaljevic> btw storm again
2008-07-07T22:00:59  <xorAxAx> what is needsupdate?
2008-07-07T22:01:08  <xorAxAx> byegonweon: hi
2008-07-07T22:01:26  <byegonweon> xorAxAx: hi~
2008-07-07T22:01:35  <xorAxAx> byegonweon: i am sorry but http://hg.moinmo.in/moin/1.8-guieditor-mbyeongweon/rev/97dd5fb0c75b is full of changes that are unrelated to the adding of a comment
2008-07-07T22:01:44  <xorAxAx> byegonweon: this is the third or fourth time i have done this remark
2008-07-07T22:02:00  <xorAxAx> and you still seem to be violating this basic rule, i am not amused ...
2008-07-07T22:02:02  <mmihaljevic> xorAxAx: function in cache that checks mtime
2008-07-07T22:02:13  <xorAxAx> mmihaljevic: "in cache"?
2008-07-07T22:02:27  <mmihaljevic> pardon my fault caching.py
2008-07-07T22:02:37  <mmihaljevic> need to be more precise
2008-07-07T22:02:43  <byegonweon> xorAxAx: I'm sorry for that.
2008-07-07T22:03:02  <byegonweon> xorAxAx: but I add comment and reformating same file too
2008-07-07T22:03:25  <byegonweon> xorAxAx: hmm. should I that kind of works separated?
2008-07-07T22:03:37  <xorAxAx> mmihaljevic: well, that function simply compares the mtime, right?
2008-07-07T22:03:42  <mmihaljevic> yes
2008-07-07T22:03:45  <xorAxAx> byegonweon: yes, and mention it in any case
2008-07-07T22:04:00  <xorAxAx> byegonweon: if you just commit a reformat changeset, i can simply skip it while reviewing
2008-07-07T22:04:15  <xorAxAx> byegonweon: but if you do changes and reformat in the same step, thats not possible anymore
2008-07-07T22:04:34  <xorAxAx> because there are many blocks of changes, called hunks, of unrelated stuff
2008-07-07T22:04:45  <byegonweon> xorAxAx: oh.. I understand why that's not so good.
2008-07-07T22:04:56  <xorAxAx> code is not only written to be understood by a computer but also for humans :)
2008-07-07T22:05:00  <xorAxAx> like you and me :)
2008-07-07T22:05:19  <byegonweon> xorAxAx: hehe. thanks for teaching me.
2008-07-07T22:05:35  <xorAxAx> hehe
2008-07-07T22:07:45  <mmihaljevic> xorAxAx: if I understanded I will not need to compare mtime I will got notification that the file changed
2008-07-07T22:08:26  <mmihaljevic> sorry for stupd questions but I need to be clear that I understand what I need to do
2008-07-07T22:10:00  <xorAxAx> no, the question is not stupid :)
2008-07-07T22:10:15  <xorAxAx> mmihaljevic: well
2008-07-07T22:10:31  <xorAxAx> mmihaljevic: depends on who "you" are :) in fact, you have a backend and the groupmanager stuff.
2008-07-07T22:10:42  <mmihaljevic> yes
2008-07-07T22:11:07  <xorAxAx> the backend will do it like the old code ... but it might use some observer pattern (thats to be designed, not sure if its the best choice) to signal the groupmanager what to do
2008-07-07T22:12:01  <mmihaljevic> "the backend will do it like the old code" - like the old wikidicts code or?
2008-07-07T22:13:45  <xorAxAx> yes. have you understood the basic pattern?
2008-07-07T22:13:51  <xorAxAx> of that "old code"
2008-07-07T22:14:05  <xorAxAx> i only mean the updating code ... not the whole wikidicts structure
2008-07-07T22:14:17  <mmihaljevic> observer pattern ?
2008-07-07T22:14:34  <xorAxAx> (that is gone for good ... as i said that one doesnt need to keep it and you already started to use the new api)
2008-07-07T22:14:41  <mmihaljevic> xorAxAx: yes I understand (u mean updating backends code)
2008-07-07T22:14:44  <xorAxAx> mmihaljevic: no, there is no observer pattern in the old code
2008-07-07T22:15:08  <mmihaljevic> I think you mean the pattern for the new code
2008-07-07T22:15:09  <xorAxAx> have you realised how the old code noticed that a group page changed?
2008-07-07T22:15:16  <xorAxAx> no, i want you to nuderstand the old code
2008-07-07T22:15:23  <mmihaljevic> by checking the mtime
2008-07-07T22:15:44  <mmihaljevic> and comparing to the cache mtime
2008-07-07T22:15:51  <xorAxAx> checking the mtime of what?
2008-07-07T22:16:05  <mmihaljevic> mtime of the group page file
2008-07-07T22:16:19  <mmihaljevic> and comparing them
2008-07-07T22:17:37  <mmihaljevic> well I did that the same in the group_wiki.py code
2008-07-07T22:17:44  <mmihaljevic> if you have that in mind
2008-07-07T22:18:14  <xorAxAx> what is the group page file? :)
2008-07-07T22:18:58  <mmihaljevic> http://paste.pocoo.org/show/78847/
2008-07-07T22:20:19  <xorAxAx> pasting an url does not let me recognize that you have actually understood the idea and concept behind the code that the url refers to
2008-07-07T22:20:27  <xorAxAx> :)
2008-07-07T22:20:34  <mmihaljevic> ok :)
2008-07-07T22:20:43  <mmihaljevic> let me try to explain :)
2008-07-07T22:25:53  <mmihaljevic> ok at first I have a Page object creted
2008-07-07T22:27:13  <xorAxAx> that is irrelevant to the concept
2008-07-07T22:35:33  <xorAxAx> byegonweon: but i like the comments :)
2008-07-07T22:36:42  <byegonweon> xorAxAx: thanks a lot :)
2008-07-07T22:37:23  <mmihaljevic> ok, if a page exists it is stored for example /home/gizmo/1.8-ldapgroups-mmihaljevic/tests/wiki/underlay/pages/FrontPage/revidions/<revision_number> and if a page is changed the new revision is created
2008-07-07T22:38:25  <xorAxAx> yes but how is that related to the above mtime check?
2008-07-07T22:39:00  <mmihaljevic> the current file is changed with the new revision number?
2008-07-07T22:39:42  <xorAxAx> thats completly correct but not your issue. the question is - which file is checked for mtime changes in this case?
2008-07-07T22:39:51  <xorAxAx> i dont know where you copied the above code from
2008-07-07T22:40:00  <xorAxAx> because you didnt supply /any/ context
2008-07-07T22:40:12  <mmihaljevic> ah the pastebin code?
2008-07-07T22:41:27  <mmihaljevic> edit_log ?
2008-07-07T22:43:01  <xorAxAx> edit log of which page?
2008-07-07T22:43:54  <mmihaljevic> of a page that I'm looking if it changed ?
2008-07-07T22:44:50  <xorAxAx> which page?
2008-07-07T22:44:57  <xorAxAx> are you looking for
2008-07-07T22:45:24  <mmihaljevic> the one that is cached
2008-07-07T22:45:29  * mmihaljevic is confused
2008-07-07T22:45:59  <xorAxAx> huh
2008-07-07T22:46:13  <xorAxAx> "cached"? we are talking about the group information caching process
2008-07-07T22:46:20  <xorAxAx> group information is spread on a lot of pages
2008-07-07T22:49:54  <mmihaljevic> than I need to check all edit_logs of pages that contain group informations?
2008-07-07T22:51:40  <xorAxAx> hmm
2008-07-07T22:51:42  <xorAxAx> "you need"?
2008-07-07T22:51:48  <xorAxAx> the point is that the code is there ...
2008-07-07T22:51:56  <xorAxAx> you only need to understand how it works
2008-07-07T22:52:12  <xorAxAx> but i guess the code does something like that. but when does it do that?
2008-07-07T22:54:12  <mmihaljevic> ok if I understanded the question, it does that when there is a request for a grouppage, it checks if the page is uptodate
2008-07-07T22:56:53  <xorAxAx> no
2008-07-07T22:57:17  <xorAxAx> the question is - how does the old caching code from wikidicts work
2008-07-07T22:57:43  <xorAxAx> depending on the amount of modifications that happened in the last months, i actually know the answer to this question
2008-07-07T22:57:59  <xorAxAx> but you need to know it as well to successfully continue to implement the related parts
2008-07-07T22:59:01  <xorAxAx> also i dont see the caching code in any datastructs.backends module on your branch
2008-07-07T22:59:14  <mmihaljevic> xorAxAx: yes it is
2008-07-07T22:59:17  <xorAxAx> so you might need to have another look at it http://hg.moinmo.in/moin/1.8/file/b3b04ffbecd8/MoinMoin/wikidicts.py
2008-07-07T22:59:29  <xorAxAx> excuse me?
2008-07-07T22:59:45  <mmihaljevic> xorAxAx: caching code
2008-07-07T22:59:47  <mmihaljevic> ?
2008-07-07T22:59:52  <mmihaljevic> u asked for it
2008-07-07T23:00:27  <ThomasWaldmann> re
2008-07-07T23:00:29  <mmihaljevic> only thng related to caching is in datastruct/backends/group_wiki.py
2008-07-07T23:00:29  <xorAxAx> i mean, what do you mean by "it is"?
2008-07-07T23:00:41  <xorAxAx> i cant see the "old caching code" there
2008-07-07T23:00:47  <xorAxAx> and we are talking about the old code
2008-07-07T23:00:50  <xorAxAx> and how to understand it
2008-07-07T23:00:54  <mmihaljevic> ah pardon
2008-07-07T23:02:23  * mmihaljevic looks at the wikidict code
2008-07-07T23:03:59  <ThomasWaldmann> mmihaljevic: it is still (a) results (list),
2008-07-07T23:04:14  <ThomasWaldmann> even if there is only one result in it
2008-07-07T23:04:47  <mmihaljevic> ThomasWaldmann: ah yes
2008-07-07T23:05:00  <ThomasWaldmann> and as you are searching for multiple group patterns ORed,
2008-07-07T23:05:30  <xorAxAx> note that the code is still not thread-safe
2008-07-07T23:05:32  <ThomasWaldmann> I guess in some pathological case, you could even get multiple e
2008-07-07T23:05:37  <ThomasWaldmann> results
2008-07-07T23:05:43  <xorAxAx> but it should give you a basic understanding of how it could be implemented at least
2008-07-07T23:06:08  <mmihaljevic> ThomasWaldmann: you mean to have posixgroup and grouOfnames with the same name
2008-07-07T23:06:11  <mmihaljevic> ?
2008-07-07T23:06:27  <mmihaljevic> xorAxAx: ok
2008-07-07T23:07:43  <mmihaljevic> xorAxAx: you mean on scan_begin and scan_end time?
2008-07-07T23:08:15  <xorAxAx> mmihaljevic: i dont see how that is used
2008-07-07T23:08:31  <mmihaljevic> ThomasWaldmann: than I will need also to check how many results I have
2008-07-07T23:08:43  <xorAxAx> no, why do you think that i am talking about those (mysterious to me) attributes?
2008-07-07T23:10:03  <ThomasWaldmann> mmihaljevic: cpng
2008-07-07T23:10:08  <ThomasWaldmann> oops
2008-07-07T23:10:29  <ThomasWaldmann> mmihaljevic: congrats to your exam. what was it about?
2008-07-07T23:10:33  <mmihaljevic> xorAxAx: nvm I see that they aren't used
2008-07-07T23:11:00  <mmihaljevic> ThomasWaldmann: economics :( I failed it first time (needed to know samuelson book )
2008-07-07T23:11:20  <mmihaljevic> xorAxAx: ok about the caching
2008-07-07T23:11:45  <ThomasWaldmann> what the problem about caching? /me has some kind of deja vue...
2008-07-07T23:13:39  <mmihaljevic> ThomasWaldmann: uh my understanding :| ahm...
2008-07-07T23:14:55  <mmihaljevic> xorAxAx: well it inits a cache entry and update data
2008-07-07T23:15:09  <ThomasWaldmann> i had a quick look through the backlog, but not sure about what / why we are discussing
2008-07-07T23:15:18  <ThomasWaldmann> some short summary maybe?
2008-07-07T23:15:33  <xorAxAx> ThomasWaldmann: we are discussing about the learning process, just on a different level :)
2008-07-07T23:15:44  <ThomasWaldmann> heh
2008-07-07T23:15:51  <xorAxAx> ThomasWaldmann: melita is trying to implement the page backend
2008-07-07T23:15:59  <xorAxAx> and for that, she needs to understand how the old code worked
2008-07-07T23:16:20  <xorAxAx> and we need to understand why understanding the code doesnt work out fine  yet
2008-07-07T23:16:26  <ThomasWaldmann> well, maybe rather how the new code should work
2008-07-07T23:16:37  <mmihaljevic> :)
2008-07-07T23:16:39  <ThomasWaldmann> that's not necessarily the same
2008-07-07T23:17:32  <xorAxAx> ThomasWaldmann: its essential to understand the old code and its deficies
2008-07-07T23:17:43  <xorAxAx> also to understand all requirements
2008-07-07T23:18:46  <ThomasWaldmann> i am not sure we should invest the next few days into this
2008-07-07T23:19:09  <mmihaljevic> well, the old code is updating data and not checking if it needs to be updated?
2008-07-07T23:19:21  <xorAxAx> mmihaljevic: no
2008-07-07T23:19:35  <ThomasWaldmann> if melita does and does not get any code done, she maybe will understand it, but the repo will still look rather bad
2008-07-07T23:21:23  <xorAxAx> well, i am not her mentor ... nevertheless i currently see large problems related to understanding the goal (not only existing code), she also said this and asking for code does not fix these
2008-07-07T23:21:30  <xorAxAx> the mentor should therefore help
2008-07-07T23:21:49  <ThomasWaldmann> so i rather would delay implementing the ultimate caching thing to after midterm
2008-07-07T23:22:11  <ThomasWaldmann> and before get some code done that "basically works"
2008-07-07T23:22:51  <ThomasWaldmann> without addressing mtime granularity effects, in memory caching etc.
2008-07-07T23:22:59  <xorAxAx> yes, makes sense
2008-07-07T23:23:06  <xorAxAx> if there are no problems of understanding on that part
2008-07-07T23:23:24  <gizmach> sorry got disconnected
2008-07-07T23:23:42  <ThomasWaldmann> last thing you got?
2008-07-07T23:23:58  <gizmach> <ThomasWaldmann> so i rather would delay implementing the ultimate caching thing to after midterm
2008-07-07T23:24:05  <xorAxAx> 23:22:10 <  ThomasWaldmann > and before get some code done that "basically works"
2008-07-07T23:24:09  <xorAxAx> 23:22:51 <  ThomasWaldmann > without addressing mtime granularity effects, in memory caching etc.
2008-07-07T23:24:22  <gizmach> :) thx
2008-07-07T23:25:21  <ThomasWaldmann> i talked with reimar a lot yesterday and we think 'concentrate just on groups'
2008-07-07T23:25:29  <gizmach> ok
2008-07-07T23:25:55  <ThomasWaldmann> i.e. make them work for config,page,ldap with tests proving it
2008-07-07T23:26:12  <ThomasWaldmann> (and simple caching, if needed)
2008-07-07T23:26:20  <gizmach> ok
2008-07-07T23:26:41  <ThomasWaldmann> once we have that, improve caching
2008-07-07T23:27:02  <ThomasWaldmann> then do dicts and lists (piece of cake)
2008-07-07T23:27:23  <gizmach> ok
2008-07-07T23:27:28  <ThomasWaldmann> then get rid of wikidicts
2008-07-07T23:28:37  <ThomasWaldmann> (if done differently, it would lead to lots of change duplication over groups,lists,dicts)
2008-07-07T23:28:52  <gizmach> ok
2008-07-07T23:28:55  <xorAxAx> whats the difference between groups and lists again?
2008-07-07T23:29:14  <ThomasWaldmann> there needn't to be one
2008-07-07T23:29:18  <xorAxAx> ok
2008-07-07T23:29:24  <xorAxAx> the memory datastructure is kind of immutable anyway
2008-07-07T23:29:27  <ThomasWaldmann> groups could be sets, lists are lists
2008-07-07T23:29:34  <xorAxAx> so having two objects (list/set) is fine
2008-07-07T23:32:12  <gizmach> ThomasWaldmann: than I have to stat with ldap code ?
2008-07-07T23:32:37  <ThomasWaldmann> are your wikipage group tests working?
2008-07-07T23:33:11  <gizmach> ThomasWaldmann: the one I wrote to chek if the methods wrok are working
2008-07-07T23:33:23  <gizmach> s/chek/check
2008-07-07T23:33:41  <ThomasWaldmann> check on acl with group level
2008-07-07T23:33:52  <ThomasWaldmann> like you did for config
2008-07-07T23:34:02  <gizmach> ok
2008-07-07T23:34:56  <dreimark> moin
2008-07-07T23:35:03  <gizmach> moin dreimark
2008-07-07T23:36:41  * ThomasWaldmann held a small moin session at europython today
2008-07-07T23:37:31  <ThomasWaldmann> ernestas (GHOP) was also at the conference
2008-07-07T23:38:56  <xorAxAx> cool
2008-07-07T23:39:41  <ThomasWaldmann> he's from vilnius :)
2008-07-07T23:42:49  * dreimark has read backlogs
2008-07-07T23:44:07  <dreimark> gizmach: why are you so fixated to work always again on ldap while other parts still broken or not completed
2008-07-07T23:44:28  <dreimark> investigating ldap does cost lots of time and will give later probably code
2008-07-07T23:44:36  <gizmach> dreimark: will complete the wikipage to work
2008-07-07T23:44:50  <gizmach> hehe also investigating cache does :)
2008-07-07T23:44:56  <dreimark> while you can do invesitigating and coding at the same time
2008-07-07T23:45:06  <gizmach> dreimark: true :)
2008-07-07T23:45:12  <gizmach> you re right there
2008-07-07T23:45:51  <ThomasWaldmann> and esp. doing something else while you wait for test runs to complete
2008-07-07T23:46:32  <gizmach> :)
2008-07-07T23:48:59  <dreimark> gizmach: please don't discuss in queries parts of MM with one of us, it is not possible to continue or help without repeating everything by one of the others
2008-07-07T23:49:30  <dreimark> this costs much time for everyone
2008-07-07T23:49:55  <dreimark> because sometimes someone else could smell why one has problemes to get something but can't help
2008-07-07T23:53:26  <dreimark> gizmach: xorAxAx is also right if one likes to implement a new concept it is necessary to understand the old one for preventing doing some mistakes or unwanted features
2008-07-07T23:53:54  <xorAxAx> dreimark: yes, but its correct that thats all caching related and we can postpone that for now
2008-07-07T23:56:34  <dreimark> xorAxAx: yeah dicussed with Thomas long about that is the current idea gizmach's learning process should/will continue
2008-07-07T23:57:28  <dreimark> but before she spent too much time on getting how caching works the group parts should be implemented
2008-07-07T23:58:07  <dreimark> some of the missing parts
2008-07-07T23:58:22  <gizmach> ok