MoinMoin: MoinMoinChat/Logs/moin-dev/2006-07-26

2006-07-26T11:59:05  <Kepplar> ThomasWaldmann: Just to let you know I can't seem to get a route through from my server to the outside world so I havent been able to push up changes
2006-07-26T11:59:38  <Kepplar> then again the only change i have to push is my log, i've been working pen-and-paper style (yesterday, as irc logs support) on the specialised item classes
2006-07-26T12:02:02  <Kepplar> (someone reset the wireless router btw, and its some cheap thing so I'm not suprised half the networks broken)
2006-07-26T12:10:54  *** irc.freenode.net sets mode: +o ChanServ
2006-07-26T12:19:29  <xorAxAx> ThomasWaldmann: are you refactoring the request classes?
2006-07-26T12:33:17  <ThomasWaldmann> see repo
2006-07-26T12:38:16  <xorAxAx> ThomasWaldmann: "_emit_http_headers" has an underscore, but is called from the base class ...
2006-07-26T12:38:49  <xorAxAx> i wonder if that might be confusing and let people think that is a private method of the derived class
2006-07-26T12:38:54  <xorAxAx> should i rename it?
2006-07-26T12:41:37  <ThomasWaldmann> if just dont want anybody calling _emit_http_headers
2006-07-26T12:42:22  <Kepplar> do we have an automatic epydoc generated output of our branches?
2006-07-26T12:42:27  <ThomasWaldmann> it only has to get called from the base class
2006-07-26T12:42:34  <xorAxAx> sure
2006-07-26T12:42:48  <xorAxAx> but there isnt even an "abstract" method in the base class where this could be documented
2006-07-26T12:42:52  <xorAxAx> --> looks fragile
2006-07-26T12:43:10  <Kepplar> also I'm not too sure on "return" and "rtype" tags
2006-07-26T12:43:22  <Kepplar> i cant see a reason to use return really
2006-07-26T12:43:35  <xorAxAx> Kepplar: apt-get install python-epydoc # if you have debian
2006-07-26T12:43:43  <xorAxAx> Kepplar: and then you can generate it for yourself :)
2006-07-26T12:44:22  <ThomasWaldmann> xorAxAx: adding same method in base class is no problem
2006-07-26T12:44:34  * ThomasWaldmann adds it
2006-07-26T12:45:33  <Kepplar> @return for an object implies its str() value?
2006-07-26T12:45:33  <moinBot`> Kepplar: Error: "return" is not a valid command.
2006-07-26T12:47:10  <xorAxAx> Kepplar: @return should be followed by an explanation of the return value
2006-07-26T12:47:19  <xorAxAx> Kepplar: and @rtype should describe its type
2006-07-26T12:47:35  <Kepplar> Oh
2006-07-26T12:47:40  <Kepplar> so I can waffle on about what it is
2006-07-26T12:47:51  <Kepplar> instead of something definitive for example for ints
2006-07-26T12:47:58  <Kepplar>  @return: int
2006-07-26T12:48:00  <Kepplar> er
2006-07-26T12:48:04  <Kepplar>  @rtype: int
2006-07-26T12:48:07  <xorAxAx> yes
2006-07-26T12:48:08  <Kepplar>  @return 0
2006-07-26T12:48:14  <Kepplar> so for example
2006-07-26T12:48:16  <xorAxAx> and @return the revision number
2006-07-26T12:48:20  <xorAxAx> no, not the latter one
2006-07-26T12:48:22  <Kepplar>  @rtype Item
2006-07-26T12:48:29  <Kepplar>  @return New item object
2006-07-26T12:48:34  <xorAxAx> yes
2006-07-26T12:48:42  <Kepplar> cool
2006-07-26T12:48:42  <Kepplar> ta
2006-07-26T12:54:14  * ThomasWaldmann committed
2006-07-26T12:55:01  <ThomasWaldmann> anyone using something except CGI for 1.6?
2006-07-26T12:56:23  <ThomasWaldmann> btw, I wont be here this evening
2006-07-26T12:56:30  * xorAxAx uses standalone
2006-07-26T12:58:49  <Kepplar> i use whatever make test uses
2006-07-26T12:58:54  <Kepplar> i use CGI on 1.5
2006-07-26T13:04:02  <ThomasWaldmann> standalone works (did a quick test run)
2006-07-26T13:07:16  <ThomasWaldmann> make test also works now
2006-07-26T17:04:17  <xorAxAx> ThomasWaldmann: it was broken?
2006-07-26T17:04:26  <xorAxAx> i mean, there was one failing test. do you mean that one?
2006-07-26T17:28:06  <Kepplar> I might be implimenting a new generic datatype for moin
2006-07-26T17:28:11  <Kepplar> i think it deserves it own module too
2006-07-26T17:28:14  <Kepplar> where shall I put it
2006-07-26T17:28:27  <Kepplar> (its just a datatype like dict or list, its not tied to anything)
2006-07-26T17:34:46  <xorAxAx> Kepplar: what kind of data type?
2006-07-26T17:45:13  <Kepplar> effectively a multi-dimentional dict
2006-07-26T17:45:35  <Kepplar> which im calling rubix
2006-07-26T17:45:37  <Kepplar> k
2006-07-26T17:45:38  <Kepplar> even
2006-07-26T17:45:54  <xorAxAx> what do you want to store in it?
2006-07-26T17:46:12  <Kepplar> mime to itemtype translation
2006-07-26T17:46:23  <Kepplar> and itemtype to storagelocation translation
2006-07-26T17:46:38  <Kepplar> and POSSIBLY (i havent thought it entirely through) it might be helpful for locating subpagesd
2006-07-26T17:46:47  <xorAxAx> i wonder where the multiple dimensions are in that use case :)
2006-07-26T17:46:57  <Kepplar> hm?
2006-07-26T17:48:02  <Kepplar> basically would work like this
2006-07-26T17:48:37  <Kepplar> myspecialiseditemrevisiontype = myrubiks["text/x-moin-wiki"] for example
2006-07-26T17:48:44  <Kepplar> takes first stanza
2006-07-26T17:48:52  <Kepplar> checks against dictionary (possibly even with regex)
2006-07-26T17:49:12  <Kepplar> finds the rubik thats related to the key the first entry matches
2006-07-26T17:49:19  <xorAxAx> umm, why would you want a dictionary if you are doing regex searches?
2006-07-26T17:49:20  <Kepplar> and runs the second against the rubik
2006-07-26T17:49:36  <Kepplar> no this is just a serialised way of storing it
2006-07-26T17:49:37  <Kepplar> plus
2006-07-26T17:49:45  <Kepplar> it does use regex
2006-07-26T17:50:27  <Kepplar> what would you do, first have a regex for text and then another if successful (nested if) for x-moin-wiki?
2006-07-26T17:50:34  <Kepplar> then extending it will be via putting another line it?
2006-07-26T17:50:40  <Kepplar> thats what im avoiding
2006-07-26T17:51:14  <xorAxAx> i think a dictionary that maps strings with regexes to classes is equivalent, isnt it?
2006-07-26T17:51:46  <Kepplar> kinda
2006-07-26T17:51:51  <Kepplar> this is generic type
2006-07-26T17:52:00  <Kepplar> I'm using something already which does most of it
2006-07-26T17:52:03  <Kepplar> in RevisionBuilder
2006-07-26T17:52:31  <Kepplar> at the moment I'm debating how to construct the rubik
2006-07-26T17:52:57  <Kepplar> possibly there will be multiple arrays including feeding it a giant serialised representation of All the data and it building it up on __init__
2006-07-26T17:53:14  <Kepplar> this would be slow, especially if ones building it in an algor
2006-07-26T18:10:29  <Kepplar> Hmm
2006-07-26T18:10:45  <Kepplar> afk for a while
2006-07-26T18:49:09  <xorAxAx> ThomasWaldmann: action=info is broken in the test wiki
2006-07-26T20:10:06  <Kepplar> yaya, network problems solved :)
2006-07-26T20:10:20  <Kepplar> a switches arp table was mashed