MoinMoin: MoinMoinChat/Logs/moin-dev/2010-05-18

   1 2010-05-18T01:21:46  *** gwork has quit IRC
   2 2010-05-18T05:03:09  *** ei-grad has joined #moin-dev
   3 2010-05-18T05:17:59  *** kursor has joined #moin-dev
   4 2010-05-18T05:18:59  *** kursor_ has joined #moin-dev
   5 2010-05-18T05:18:59  *** kursor has quit IRC
   6 2010-05-18T05:18:59  *** kursor_ is now known as kursor
   7 2010-05-18T05:33:47  *** TheSheep has quit IRC
   8 2010-05-18T05:34:36  *** TheSheep has joined #moin-dev
   9 2010-05-18T05:57:03  *** kursor has quit IRC
  10 2010-05-18T06:19:39  *** xjjk has quit IRC
  11 2010-05-18T06:23:55  *** ei-grad has quit IRC
  12 2010-05-18T06:28:13  *** ei-grad has joined #moin-dev
  13 2010-05-18T06:28:58  <ei-grad> moin
  14 2010-05-18T07:44:10  <dreimark> hi ei-grad
  15 2010-05-18T07:44:23  <dreimark> do you have looked at py.test and the test?
  16 2010-05-18T07:46:25  <ei-grad> hi dreimark, yes, i'm working on it now
  17 2010-05-18T07:47:48  <ei-grad> its more complex than i recognized at first look)
  18 2010-05-18T07:49:26  <dreimark> btw. can you stay connected on irc?
  19 2010-05-18T07:50:48  <ei-grad> yep
  20 2010-05-18T07:54:00  <dreimark> ei-grad: ThomasWaldmann did most of that code and decided to skip the test after the Config refactoring
  21 2010-05-18T07:54:22  <dreimark> you can look at it how it works before that
  22 2010-05-18T07:54:30  <dreimark> also you can ask him too
  23 2010-05-18T10:34:45  <ThomasWaldmann> moin
  24 2010-05-18T10:45:44  <ThomasWaldmann> ei-grad: did you install the windows server already?
  25 2010-05-18T10:45:59  <ei-grad> yes
  26 2010-05-18T10:46:36  <ThomasWaldmann> the 2008r2 eval?
  27 2010-05-18T10:46:49  <ei-grad> 2003...
  28 2010-05-18T10:47:30  <ThomasWaldmann> ok, good enough
  29 2010-05-18T10:52:05  <ThomasWaldmann> maybe also extend your wiki pages with your ideas about what / how to implement after coding phase starts
  30 2010-05-18T11:39:09  <ThomasWaldmann> valeuf: in the test changeset you posted, indenting in line 29 and 30 looks different
  31 2010-05-18T11:39:43  <ThomasWaldmann> http://pastie.org/965181
  32 2010-05-18T11:40:16  <ThomasWaldmann> can you check in the src whether it is correct and spaces-only (no tabs)?
  33 2010-05-18T11:41:20  <ThomasWaldmann> s/generate/baseline_shift/g
  34 2010-05-18T11:42:03  <ThomasWaldmann> 34/36 pep8
  35 2010-05-18T11:43:34  <ThomasWaldmann> 37 check grammar
  36 2010-05-18T11:44:29  <ThomasWaldmann> valeuf: other than that (and that the test still uses strings), looks good for me
  37 2010-05-18T12:21:43  <valeuf> ThomasWaldmann: thanks
  38 2010-05-18T12:22:16  <valeuf> ThomasWaldmann: actually I did it with my eeePC during a lecture this morning, I suddenly realize that I do not have a correct configuration file for vim
  39 2010-05-18T12:22:29  <valeuf> ThomasWaldmann: so tabulation are completely messy
  40 2010-05-18T12:22:57  <valeuf> ThomasWaldmann: I wil fix it, (and copy my .vimconfig for my main computer to my eeePC)
  41 2010-05-18T12:23:12  <ThomasWaldmann> btw, the stackoverflow stuff doesn't look usable
  42 2010-05-18T12:23:28  <ThomasWaldmann> first thing is still comparing strings
  43 2010-05-18T12:23:56  <ThomasWaldmann> the tree based thing there is just usable to compare full trees, so maybe not that useful either
  44 2010-05-18T12:24:17  <valeuf> waldi: I finally found how to add support for the missing stuffs in the html_out, actually, I was completely wrong before, because I thought there was a <super></super> tag, but it is using <span> with base-line attribute
  45 2010-05-18T12:25:08  <valeuf> waldi: I will rewrite some more complete specs for the DOM tree, because I saw quite many different stuffs compare to http://moinmo.in/BastianBlank/TreeOutputFormatter/Spec
  46 2010-05-18T12:25:38  <valeuf> ThomasWaldmann: yep, stackoverflow did not help a lot, but it was interesting to read
  47 2010-05-18T12:26:13  <valeuf> ThomasWaldmann: the problem is that I find really complicate stuffs to compare the whole XML tree, otherwise it is usually using string comparaison
  48 2010-05-18T12:27:25  <ThomasWaldmann> well, it should only check what the test is intended to check
  49 2010-05-18T12:28:23  <ThomasWaldmann> otherwise you'll likely have to change lots of tests after doing some change in the converters, even if unrelated to most tests
  50 2010-05-18T12:32:19  <ThomasWaldmann> valeuf: http://stackoverflow.com/questions/222375/elementtree-xpath-select-element-based-on-attribute
  51 2010-05-18T12:33:00  <ThomasWaldmann> check the corresponding stuff in EmeraldTree
  52 2010-05-18T12:33:17  <xorAxAx> "Using ElementTree 1.2.6, the attribute xpath features are only available in 1.3 and beyond"
  53 2010-05-18T12:33:21  <xorAxAx> which version was forked?
  54 2010-05-18T12:35:13  <valeuf> xorAxAx: ElementTree 1.3
  55 2010-05-18T12:35:30  <ThomasWaldmann> all that ElTr stuff is quite old
  56 2010-05-18T12:35:39  <valeuf> xorAxAx: but check ElementPath.py
  57 2010-05-18T12:35:50  <valeuf> you will see that it is a really limited support
  58 2010-05-18T12:39:15  <ThomasWaldmann> valeuf: as waldi has forked it and we'll bundle it anyway, we can easily extend it, if needed :)
  59 2010-05-18T12:39:49  <valeuf> ThomasWaldmann: yep, it can be useful, we need to check with waldi
  60 2010-05-18T12:39:53  <ThomasWaldmann> valeuf: documenting what exactly is needed (and still missing) could help
  61 2010-05-18T12:40:22  <valeuf> ThomasWaldmann: ok, I will try to see what is missing
  62 2010-05-18T12:42:21  <ThomasWaldmann> def test_Element_findtext(): pass
  63 2010-05-18T12:42:35  <ThomasWaldmann> great test :) never makes troubles :))
  64 2010-05-18T12:44:45  <ThomasWaldmann> looks like the tests are only for the most basic stuff
  65 2010-05-18T12:45:11  <ThomasWaldmann> if EmTr can already do more, it would be nice to have better coverage
  66 2010-05-18T13:12:23  <valeuf> ThomasWaldmann: cool, can I also write my test like that ? :p
  67 2010-05-18T13:16:02  <ThomasWaldmann> no :P
  68 2010-05-18T13:16:19  * ThomasWaldmann tries to write some findtext tests
  69 2010-05-18T13:21:00  <CIA-55> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 57:c1feee84d5d8 EmeraldTree/emeraldtree/tests/test_tree.py: add tests for findtext()
  70 2010-05-18T13:21:44  <ThomasWaldmann> waldi: i also tried with multiple text children, but it seems this is not supported. not sure whether that is a bug or intended.
  71 2010-05-18T13:26:10  <CIA-55> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 58:9d0db78957e0 EmeraldTree/emeraldtree/tests/test_tree.py: cosmetic changes for findtext() tests
  72 2010-05-18T14:20:40  <CIA-55> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 59:33f5e57f1b97 EmeraldTree/emeraldtree/tests/test_tree.py: more xpath tests, 3 of them are skipped because they fail
  73 2010-05-18T14:22:21  *** ei-grad has quit IRC
  74 2010-05-18T14:22:29  <ThomasWaldmann> valeuf: waldi: ../foo, //foo, .//foo are failing
  75 2010-05-18T14:22:53  <waldi> ThomasWaldmann: i don't know about this path stuff
  76 2010-05-18T14:23:34  <ThomasWaldmann> some fails are related to text nodes getting into result list, although they were not selected
  77 2010-05-18T14:24:49  <ThomasWaldmann> waldi: maybe review the tests. i hope they are correct, but I have no experience with xpath stuff.
  78 2010-05-18T14:25:08  <waldi> yep
  79 2010-05-18T14:51:25  <CIA-55> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 60:0c6020f1a2be EmeraldTree/emeraldtree/tests/test_tree.py: more xpath tests, 1 is skipped because it is unsupported
  80 2010-05-18T14:52:23  *** ei-grad has joined #moin-dev
  81 2010-05-18T14:52:28  <ThomasWaldmann> http://effbot.org/zone/element-xpath.htm talks about [position], but there seems to be no code for that
  82 2010-05-18T14:53:42  <ThomasWaldmann> valeuf: for your xpath based tests: first have a look at the tests I added (see above). then you'll know what works and what not.
  83 2010-05-18T14:58:10  <CIA-55> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 61:6c0a8bd9aa9a EmeraldTree/emeraldtree/tests/test_tree.py: more xpath attribute tests (with other and without value)
  84 2010-05-18T15:12:15  *** ei-grad has quit IRC
  85 2010-05-18T15:50:39  *** kursor has joined #moin-dev
  86 2010-05-18T15:51:39  *** franklin has quit IRC
  87 2010-05-18T16:12:36  <esyr_> moin
  88 2010-05-18T16:26:10  <ThomasWaldmann> hi esyr_
  89 2010-05-18T16:30:41  *** esyr_ is now known as esyr
  90 2010-05-18T16:31:00  *** esyr is now known as eSyr
  91 2010-05-18T16:48:22  <valeuf> ThomasWaldmann: I will check
  92 2010-05-18T16:50:56  *** diofeher has joined #moin-dev
  93 2010-05-18T17:30:02  *** diofeher has quit IRC
  94 2010-05-18T17:32:50  *** kursor has quit IRC
  95 2010-05-18T17:38:29  *** kursor has joined #moin-dev
  96 2010-05-18T19:43:53  *** kursor has quit IRC
  97 2010-05-18T20:13:54  <ThomasWaldmann> re
  98 2010-05-18T20:18:57  <dreimark> re
  99 2010-05-18T21:17:05  <ThomasWaldmann> how about using restful urls in moin2?
 100 2010-05-18T21:17:33  <ThomasWaldmann> i just discussed with TheSheep and it looks like he has found a nice way to do it
 101 2010-05-18T21:17:54  <ThomasWaldmann> http://hg.hatta-wiki.org/hatta-dev/file/700208414722/hatta.py#l2582 < similar to there
 102 2010-05-18T21:19:26  <TheSheep> moin probably needs some more complictaed patterns though
 103 2010-05-18T21:20:21  <ThomasWaldmann> using something like that we can still have /ItemName, just actions need some prefix not allowed for items, e.g. /+edit/ItemName
 104 2010-05-18T21:20:37  <TheSheep> I stole that from launchpad, btw
 105 2010-05-18T21:21:15  <ThomasWaldmann> hehe
 106 2010-05-18T21:21:34  <TheSheep> you could also have different prefixes for different purposes, user pages could use ~
 107 2010-05-18T21:22:51  <TheSheep> then you would have /~UserName for the general user preferences, /~UserName/ for their home page, and ~UserName/PageName for their subpages
 108 2010-05-18T21:23:20  <TheSheep> that's one possibility at least
 109 2010-05-18T21:24:04  <TheSheep> /~UserName/+preferences works too
 110 2010-05-18T21:24:30  <TheSheep> superusers then don't need to 'become' an user to edit their prefs
 111 2010-05-18T21:24:36  <ThomasWaldmann> or /+preferences/UserName :)
 112 2010-05-18T21:24:49  <TheSheep> true
 113 2010-05-18T21:25:16  <TheSheep> you can also have /PageName/+edit instead of /+edit/PageName
 114 2010-05-18T21:25:31  <TheSheep> but in case of actions I think it makes more sense to have them at the beginning
 115 2010-05-18T21:26:11  <TheSheep> at least for the most important ones like edit or raw
 116 2010-05-18T21:33:45  <ThomasWaldmann> if we have bigger amounts of named/optional params (like now with qs), how is it handled?
 117 2010-05-18T21:34:29  <ThomasWaldmann> and will qs be completely unused then?
 118 2010-05-18T21:39:31  <TheSheep> no, the params get passed to the function as keyword arguments
 119 2010-05-18T21:39:45  <TheSheep> so you can can use that mapping with current moin's urls too
 120 2010-05-18T21:40:00  <TheSheep> using the mapping system is not what makes the urls RESTful
 121 2010-05-18T21:40:06  <TheSheep> it's their design
 122 2010-05-18T21:40:32  <TheSheep> so it's two separate issues -- using a mapping library to handle url parsing, which I think is a good idea anyways
 123 2010-05-18T21:40:49  <TheSheep> and redesigning urls to be more restful, which may be a bad idea
 124 2010-05-18T22:34:32  *** franklin has joined #moin-dev
 125