1 2011-12-20T00:00:48 <ThomasWaldmann> no, it is not
2 2011-12-20T00:04:02 <dreimark> ok
3 2011-12-20T00:04:41 *** raignarok
4 2011-12-20T00:05:53 <CIA-59> Reimar Bauer <rb.proj AT googlemail DOT com> default * 1173:358f0945a24e 2.0/docs/user/ (6 files): docs: proofread moinwiki, rest, creolewiki, mediawiki, markups, search. Thanks to Stefan Bobev (GCI2011)
5 2011-12-20T00:07:45 <dreimark> gn
6 2011-12-20T00:13:20 <ThomasWaldmann> gn dreimark
7 2011-12-20T01:24:23 *** qxcv
8 2011-12-20T01:47:17 <cdbs> ThomasWaldmann: there?
9 2011-12-20T01:47:37 <cdbs> ThomasWaldmann: I've given up on that task now, since I wasn't able to find out how to actually access the unprotected backend
10 2011-12-20T02:22:27 <ThomasWaldmann> well, did you look at the protecting middleware?
11 2011-12-20T02:24:19 <ThomasWaldmann> cdbs:
12 2011-12-20T02:24:27 <cdbs> I did
13 2011-12-20T02:24:35 <cdbs> there was a class on that
14 2011-12-20T02:24:41 <cdbs> but I arrived at nowhere
15 2011-12-20T02:24:42 <cdbs> anyway
16 2011-12-20T02:25:11 <cdbs> I might look into it after I'm done with my current issue (I just fixed a new issue) and then will fix a test as part of another task, then will come back
17 2011-12-20T02:36:04 <ThomasWaldmann> cdbs: you see the prot mw gets "indexer" - that is basically the unprotected storage.
18 2011-12-20T02:36:16 <cdbs> oh right
19 2011-12-20T02:36:16 <ThomasWaldmann> the prot mw mostly wraps that
20 2011-12-20T02:36:32 <cdbs> I nearly arrived at that, but gave up instead of trying
21 2011-12-20T02:37:03 <ThomasWaldmann> you see all methods just delegate to same indexer methods + check acls if needed.
22 2011-12-20T02:37:05 <cdbs> ThomasWaldmann: On a side note, can you please close this pull request? https://bitbucket.org/thomaswaldmann/moin-2.0/pull-request/17/fix-issue-130-account_create-fails
23 2011-12-20T02:38:33 <ThomasWaldmann> you mean reject?
24 2011-12-20T02:38:36 <cdbs> yes
25 2011-12-20T02:38:51 <cdbs> so that the other pull request for issue #89 (which I just registered) can get through
26 2011-12-20T02:39:10 <cdbs> which is part of the same GCI task
27 2011-12-20T02:39:45 <ThomasWaldmann> i get json back if i click on reject :D
28 2011-12-20T02:40:47 <ThomasWaldmann> doesn't work, bb broken
29 2011-12-20T02:40:55 <cdbs> BB is terribly broken
30 2011-12-20T02:41:13 <cdbs> see, for instance, https://bitbucket.org/thomaswaldmann/moin-2.0/pull-request/18/fix-issue-89-settings-personal-settings
31 2011-12-20T02:41:31 <cdbs> the pull request which I just registered
32 2011-12-20T02:41:35 <cdbs> it doesn't even have a diff!
33 2011-12-20T02:41:42 <cdbs> and neither does it accept comments
34 2011-12-20T02:45:59 <ThomasWaldmann> one sees the diff after following the link to the cs
35 2011-12-20T02:46:45 *** eSyr
36 2011-12-20T02:47:55 <ThomasWaldmann> what AttributeError exactly are you catching there?
37 2011-12-20T02:48:43 <cdbs> ThomasWaldmann: When there's no openid specified (i.e. openid = None in the User constructor) we should just ignore that check and move on
38 2011-12-20T02:49:15 <cdbs> ThomasWaldmann: In the User constructor, if no OpenID is specified, it defaults to None, which doesn't register that variable
39 2011-12-20T02:49:24 <cdbs> and hence leading up to an AttributeError
40 2011-12-20T02:49:41 <cdbs> when we try to check the openid for duplicates
41 2011-12-20T02:49:43 <ThomasWaldmann> there is lots of code being executed in that try/except
42 2011-12-20T02:50:01 <cdbs> so?
43 2011-12-20T02:50:15 <cdbs> ah wait, is it a tab fail?
44 2011-12-20T02:50:25 <ThomasWaldmann> it's not a good idea to make that wide "pass" exception handlers
45 2011-12-20T02:50:25 * cdbs looks at the diff again
46 2011-12-20T02:51:13 <ThomasWaldmann> just imagine there is some problem in that method that is called. you will never see it that way.
47 2011-12-20T02:51:27 <cdbs> ThomasWaldmann: The advantage of this approach is that, if there's an AttributeError, it would catch it and just move on. If there's some other problem, an exception will be thrown
48 2011-12-20T02:51:53 <cdbs> I can add another except block, for generic Exception, which just throws that exception again
49 2011-12-20T02:52:11 <ThomasWaldmann> user.search_users(openid=form['openid'].value) what is if in that code there is an AE happening?
50 2011-12-20T02:52:34 <cdbs> hmm, riiiight
51 2011-12-20T02:53:02 <cdbs> I can add a check before the if condition, that checks if openid is set. If not, then it just skips over to the next if statement
52 2011-12-20T02:53:05 <cdbs> would that be okay
53 2011-12-20T02:53:06 <ThomasWaldmann> also, maybe we should look if we can't avoid that AE happening
54 2011-12-20T02:53:17 <ThomasWaldmann> maybe there is just something missing in User
55 2011-12-20T02:54:13 <cdbs> probably set the openid to "empty" or some other string? so that it can be checked when needed?
56 2011-12-20T02:54:20 <ThomasWaldmann> did you check how other User attributes are handled?
57 2011-12-20T02:54:44 <cdbs> hmm
58 2011-12-20T02:56:07 <ThomasWaldmann> maybe email or alias or ...
59 2011-12-20T02:56:13 <cdbs> the optional arguments to user, with the exception of openid, just aren't used outside the user class
60 2011-12-20T02:56:49 <cdbs> the required ones like email sure are used, though. But since we can depend on email existing always, we don't need a check for that
61 2011-12-20T02:57:31 <ThomasWaldmann> yeah, better look for sth optional
62 2011-12-20T02:57:31 <cdbs> ThomasWaldmann: as for aliasname, hmm, there's an "if not aliasname" check everywhere
63 2011-12-20T02:57:51 <cdbs> I can do the same in this as well
64 2011-12-20T02:59:45 <ThomasWaldmann> self.__dict__.update(copy.deepcopy(self._cfg.user_defaults))
65 2011-12-20T02:59:53 <ThomasWaldmann> that might be the interesting line
66 2011-12-20T03:01:17 <ThomasWaldmann> and if you look into that user_defaults, you see openid=None - so how can it cause an AE?
67 2011-12-20T03:02:35 <cdbs> that's weird
68 2011-12-20T03:03:48 <ThomasWaldmann> (i admit the user.py code in general is rather weird. it already got better since storage refactoring, but might need more refactoring.)
69 2011-12-20T03:03:49 <cdbs> maybe the attribute is cleared by python just because its None?
70 2011-12-20T03:04:07 <ThomasWaldmann> python has nothing against None :)
71 2011-12-20T03:05:11 <cdbs> as a matter of fact, I'm also not able to reproduce the bug myself, can you try it with a newly-created user?
72 2011-12-20T03:05:38 <cdbs> could be a problem on your side
73 2011-12-20T03:05:58 <ThomasWaldmann> reproducing should always come first :)
74 2011-12-20T03:06:50 * ThomasWaldmann tries
75 2011-12-20T03:08:29 <ThomasWaldmann> uhoh
76 2011-12-20T03:09:16 <cdbs> means?
77 2011-12-20T03:09:51 <ThomasWaldmann> my fault
78 2011-12-20T03:09:59 <ThomasWaldmann> a bug in a local configuration file
79 2011-12-20T03:10:39 <cdbs> so, it doesn't occur again?
80 2011-12-20T03:10:39 <ThomasWaldmann> sorry for waste of time
81 2011-12-20T03:10:45 <cdbs> ah, no problem
82 2011-12-20T03:10:58 <ThomasWaldmann> i found that i have overwritten the config default of user_defaults there
83 2011-12-20T03:11:02 <cdbs> I'll get back on the previous issue
84 2011-12-20T03:11:18 <ThomasWaldmann> and it was a older copy of that default, without openid stuff
85 2011-12-20T03:11:43 <cdbs> ThomasWaldmann: so, please reject this pull request before I run rollback again, and screw BB up again :)
86 2011-12-20T03:11:52 <ThomasWaldmann> so it was correct at some time, but wasn't updated when openid was added to the builtin default
87 2011-12-20T03:12:09 <ThomasWaldmann> i can't reject, it doesn't work
88 2011-12-20T03:12:30 <cdbs> hmm
89 2011-12-20T03:12:38 <ThomasWaldmann> i get some json on the screen which i expect should go somewhere else :)
90 2011-12-20T03:12:59 <cdbs> If I were you guys, I would've chosen Github over BB in a jiffy
91 2011-12-20T03:13:16 <ThomasWaldmann> well, we don't use git :)
92 2011-12-20T03:13:25 <cdbs> but the migration is quite painless
93 2011-12-20T03:13:49 <ThomasWaldmann> we use mercurial for some good reasons, we won't move to git
94 2011-12-20T03:13:51 <cdbs> okay, so I'll get back on the unprotected storage issue starting tomorrow
95 2011-12-20T03:14:06 <cdbs> thanks a lot
96 2011-12-20T03:14:13 <ThomasWaldmann> but yes, github seems better than bb
97 2011-12-20T03:14:53 <ThomasWaldmann> if you have questions about storage, just ask here
98 2011-12-20T03:15:23 <cdbs> thanks
99 2011-12-20T03:15:33 <ThomasWaldmann> i might be unavailable tomorrow UTC evening, though, but there are some other mentors here
100 2011-12-20T03:15:44 <cdbs> time for dinner, bye
101 2011-12-20T03:15:50 <ThomasWaldmann> cu :)
102 2011-12-20T05:19:49 *** qxcv
103 2011-12-20T07:30:28 *** qxcv
104 2011-12-20T08:34:03 <dreimark> qxcv: hi
105 2011-12-20T08:34:20 <dreimark> it would be better you work on the wiki page logged in
106 2011-12-20T08:35:11 <dreimark> moin
107 2011-12-20T09:09:59 *** eSyr
108 2011-12-20T09:10:04 *** esyrd
109 2011-12-20T10:19:59 *** greg_f
110 2011-12-20T10:50:35 <ThomasWaldmann> moin
111 2011-12-20T14:21:04 *** qxcv
112 2011-12-20T15:52:33 *** marchael_
113 2011-12-20T15:53:46 *** marchael
114 2011-12-20T15:53:46 *** eSyr
115 2011-12-20T15:53:52 *** marchael_
116 2011-12-20T15:54:24 *** eSyr
117 2011-12-20T16:24:01 *** marchael
118 2011-12-20T18:14:52 * ThomasWaldmann is away
119 2011-12-20T18:20:53 *** greg_f
120 2011-12-20T18:26:01 *** pkumar
121 2011-12-20T18:36:27 *** brunomartin
122 2011-12-20T19:34:12 *** raignarok
123 2011-12-20T20:13:01 * dreimark EP2012 July 2nd-8th 2012.
124 2011-12-20T20:22:26 *** marchael
125 2011-12-20T20:59:43 <dreimark> bbl
126 2011-12-20T21:09:47 *** pkumar
127 2011-12-20T22:40:49 <ThomasWaldmann> re
128 2011-12-20T23:42:38 *** raignarok
129
