2008-06-22T00:00:08 <johill> check it out, what do you think?
2008-06-22T00:01:56 <dreimark> xorAxAx: seems I have a ghostwriter
2008-06-22T00:02:15 <xorAxAx> do we have a 1.8 test wiki?
2008-06-22T00:02:55 <johill> don't think so
2008-06-22T00:03:03 <dreimark> it is too new
2008-06-22T00:05:21 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3745:08dd5b1a1f1c 1.8/MoinMoin/macro/WikiConfigHelp.py: wiki config help macro: add id to headings so TOC can pick them up
2008-06-22T00:10:43 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3746:510db268574b 1.8/MoinMoin/config/multiconfig.py: make mail settings group
2008-06-22T00:14:02 <dreimark> good night
2008-06-22T00:14:48 <ThomasWaldmann> gn dreimark
2008-06-22T00:15:24 <johill> night
2008-06-22T00:15:29 <johill> the one ugly thing is the default %r stuff
2008-06-22T00:15:34 <johill> [<MoinMoin.auth.MoinAuth instance at 0x10425490>]
2008-06-22T00:17:38 <xorAxAx> johill: hmm, arent there any computed settings?
2008-06-22T00:17:42 <xorAxAx> only literals and names?
2008-06-22T00:18:38 <johill> well yeah, there are actually
2008-06-22T00:18:49 <johill> like attachmentspace = 20*1000*1000 or something
2008-06-22T00:18:50 <xorAxAx> i dont see any block where you set them :)
2008-06-22T00:18:58 <johill> ?
2008-06-22T00:19:01 <xorAxAx> ah, i meant settings that depend on others
2008-06-22T00:19:16 <johill> oh
2008-06-22T00:19:20 <johill> I cheated
2008-06-22T00:19:23 <johill> there was one, and I removed that
2008-06-22T00:19:53 <johill> that was show_nonexist_qm
2008-06-22T00:20:12 <johill> I suspect nonexist_qm was a very old setting
2008-06-22T00:20:35 <johill> I should remove it, it's not even used any more
2008-06-22T00:20:54 <xorAxAx> no???
2008-06-22T00:21:04 <xorAxAx> but its still a user setting?
2008-06-22T00:21:11 <xorAxAx> its the most important one of moin :-)
2008-06-22T00:21:32 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3747:c280acec2aaa 1.8/MoinMoin/config/multiconfig.py:
2008-06-22T00:21:32 <CIA-52> remove nonexist_qm setting
2008-06-22T00:21:32 <CIA-52> it was only used to fill checkbox_defaults['show_nonexist_qm']
2008-06-22T00:21:39 <johill> umm
2008-06-22T00:21:45 <johill> checkbox_defaults['show_nonexist_qm'] is teh default for that setting
2008-06-22T00:21:52 <xorAxAx> nice
2008-06-22T00:22:12 <johill> and I had copied over the value there
2008-06-22T00:22:35 <johill> xorAxAx: also, it doesn't make sense to compute them depending on others at that point
2008-06-22T00:22:42 <johill> because they'd be computed based on the default value
2008-06-22T00:22:46 <xorAxAx> indeed
2008-06-22T00:22:54 <johill> (so it only makes some sense if we want to change the default)
2008-06-22T00:22:57 <johill> in any case
2008-06-22T00:22:59 <johill> this was the only one
2008-06-22T00:23:05 <xorAxAx> wow, this structure forces people to keep this thing clean :)
2008-06-22T00:23:09 <johill> heh
2008-06-22T00:23:19 <johill> well the indentation could use some work ;)
2008-06-22T00:24:02 <johill> now we only need to make wiki formatting in the help texts work
2008-06-22T00:24:09 <johill> and copy over the descriptions
2008-06-22T00:24:45 <johill> is there any way to print out the original expression that was parsed instead of using %r?
2008-06-22T00:24:56 <xorAxAx> no
2008-06-22T00:25:01 <johill> ok
2008-06-22T00:25:23 <xorAxAx> except if you wrap them
2008-06-22T00:25:51 <xorAxAx> with a simple Expr class that is evaluated while setting the attr on the class but shown literally on the web
2008-06-22T00:25:59 <xorAxAx> (in cases where you would need it
2008-06-22T00:26:33 <johill> you mean Expr('2*100')
2008-06-22T00:26:39 <johill> instead of 2*100?
2008-06-22T00:29:51 <xorAxAx> where necessary
2008-06-22T00:29:57 <johill> right
2008-06-22T00:30:04 <johill> yeah that's doable
2008-06-22T00:30:13 <johill> it'd only be really useful in two cases right now
2008-06-22T00:36:38 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3748:dc917a577d38 1.8/MoinMoin/ (config/multiconfig.py macro/WikiConfigHelp.py):
2008-06-22T00:36:38 <CIA-52> add multiconfig.DefaultExpression for documentation
2008-06-22T00:36:38 <CIA-52> this allows showing the actual expression used as the default rather than
2008-06-22T00:36:38 <CIA-52> using '%r', use sparingly
2008-06-22T00:40:26 <xorAxAx> ok, there are 7 translated versions of helponconfiguration
2008-06-22T00:40:39 <johill> actually translated?
2008-06-22T00:40:47 <johill> or just some of the docs
2008-06-22T00:40:51 <johill> and not the settings texts
2008-06-22T00:40:57 <xorAxAx> actually translated but mostly outdated
2008-06-22T00:41:17 <xorAxAx> so at least translators think that it is useful :)
2008-06-22T00:42:51 <johill> :)
2008-06-22T01:01:18 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3749:b0d5f7f41d4b 1.8/MoinMoin/config/multiconfig.py: split out session settings
2008-06-22T01:01:19 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3750:d5b786be06d7 1.8/MoinMoin/ (config/multiconfig.py session.py): make anonymous_session_lifetime a proper setting
2008-06-22T01:11:07 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3751:507867abcc19 1.8/ (MoinMoin/config/multiconfig.py docs/CHANGES): remove url_prefix setting
2008-06-22T01:13:16 <ThomasWaldmann> johill: merge 1.7 before there are too many conflicts :)
2008-06-22T01:13:45 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3752:5cca96861760 1.8/MoinMoin/macro/WikiConfigHelp.py: WikiConfigHelp: remove "No description provided" text, it's obvious anyway
2008-06-22T01:14:20 <johill> ThomasWaldmann: I did a while ago
2008-06-22T01:14:22 <johill> shouldn't be much
2008-06-22T01:15:30 <dennda> good night
2008-06-22T01:15:51 <johill> nigiht
2008-06-22T01:15:53 <ThomasWaldmann> gn everybody :)
2008-06-22T01:16:24 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3758:4a18d48f8a85 1.8/ (47 files in 5 dirs): merge 1.7
2008-06-22T01:21:52 <johill> wtf
2008-06-22T01:22:20 <johill> why an option edit_ticketing??
2008-06-22T02:09:30 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3759:2b78179acbea 1.8/MoinMoin/config/multiconfig.py: normalise mail_enabled to a boolean
2008-06-22T02:15:23 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3760:f004cd750c86 1.8/MoinMoin/config/multiconfig.py: remove subscribable_events stuff from config
2008-06-22T02:51:28 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3761:24e9fecdd81c 1.8/MoinMoin/config/multiconfig.py: config: move auth post-processing to proper place
2008-06-22T02:51:29 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3762:95266d0c8a6f 1.8/wikiconfig.py: fix logo string
2008-06-22T08:22:13 <dreimark> current CopyPage action is not compatible to hierarchic_acls
2008-06-22T08:22:28 <dreimark> and should be excluded in 1.6.4 too
2008-06-22T08:22:38 <dreimark> moin
2008-06-22T08:31:29 <dreimark> and from that point of view the nr part needs to be removed
2008-06-22T08:41:27 <ThomasWaldmann> moin
2008-06-22T09:30:18 <xorAxAx> moin byeongweon
2008-06-22T10:00:10 <johill> moin
2008-06-22T10:11:25 <johill> xorAxAx: I was working on a "show wiki configuration" macro too, but it always shows the password_checker because the function gets bound to the object as a method
2008-06-22T10:11:36 <johill> xorAxAx: any idea how to figure out whether it's the same?
2008-06-22T10:12:05 <xorAxAx> "the same"?
2008-06-22T10:12:22 <johill> well once it is
2008-06-22T10:12:23 <johill> <function _default_password_checker at 0x10426630>
2008-06-22T10:12:29 <johill> the other time it is
2008-06-22T10:12:30 <johill> <bound method Config._default_password_checker of <wikiconfig_local.Config object at 0x10433870>>
2008-06-22T10:12:49 <johill> but the only difference is that the function was bound
2008-06-22T10:13:12 <xorAxAx> how do you check currently?
2008-06-22T10:13:22 <xorAxAx> or how do you handle the default_pwd_checker?
2008-06-22T10:14:15 <johill> it's just a module-level function that gets reassigned to the class
2008-06-22T10:14:33 <xorAxAx> no, i mean, what do you want to show for that function?
2008-06-22T10:14:40 <xorAxAx> or attribute
2008-06-22T10:14:44 <xorAxAx> the repr of the function object?
2008-06-22T10:15:11 <johill> no no showing is fine
2008-06-22T10:15:27 <johill> I use the DefaultExpression('...') class so it's just shown as _default_password_checker
2008-06-22T10:15:38 <johill> but I'm writing a macro that shows all settings that differ from the defaults
2008-06-22T10:15:43 <xorAxAx> ah
2008-06-22T10:15:46 <johill> and this one differs although it doesn't
2008-06-22T10:15:58 <xorAxAx> how do you traverse the instance?
2008-06-22T10:16:15 <xorAxAx> with dir() and getattr?
2008-06-22T10:16:38 <johill> yeah I think, let me check :)
2008-06-22T10:16:56 <johill> yes
2008-06-22T10:20:14 <xorAxAx> then traverse instance.__dict__
2008-06-22T10:20:33 <xorAxAx> that will protect you from the evil descriptors :)
2008-06-22T10:20:42 <johill> oh
2008-06-22T10:21:30 <johill> no that doesn'tw ork
2008-06-22T10:21:41 <johill> or I have to traverse .__classs__.__dict__ too
2008-06-22T10:22:15 <johill> yeah that works
2008-06-22T10:22:38 <johill> but not really
2008-06-22T10:23:15 <xorAxAx> hmm?
2008-06-22T10:23:17 <johill> now I have to look at __parent__ recursively
2008-06-22T10:23:26 <johill> or whatever it is, super(..)
2008-06-22T10:23:27 <xorAxAx> ah, yes
2008-06-22T10:23:58 <xorAxAx> well, there are some descriptors on multiconfig
2008-06-22T10:24:03 <xorAxAx> you will need to handle them somehow
2008-06-22T10:24:24 <xorAxAx> you can unpack bound methods by getattr'ing im_func
2008-06-22T10:25:19 <johill> well what I did for all the properties etc. was that I put them into a new class
2008-06-22T10:25:27 <johill> class ConfigFunctionality(object):
2008-06-22T10:25:30 <johill> # all methods
2008-06-22T10:25:36 <johill> class DefaultConfig(ConfigFunctionality):
2008-06-22T10:25:38 <johill> pass
2008-06-22T10:26:10 <johill> so if I only traverse down to DefaultConfig I'm fine
2008-06-22T10:26:47 <xorAxAx> which properties are there?
2008-06-22T10:27:27 <johill> iwid for example
2008-06-22T10:27:34 <johill> a whole bunch
2008-06-22T10:28:06 <johill> it avoids finding methods like __init__ too
2008-06-22T10:29:10 <xorAxAx> the iwid cannot be overriden
2008-06-22T10:29:28 <xorAxAx> are there any properties that deal with data that can be set by the config writer?
2008-06-22T10:30:40 <johill> not sure, but what is your point?
2008-06-22T10:32:29 <xorAxAx> because they might need to be handled differently. but i didnt find any, ok
2008-06-22T10:32:45 <xorAxAx> johill: well, if you want to stop traversal, then you have to implement it manually anyway
2008-06-22T10:33:02 <xorAxAx> you can write a method on your base class
2008-06-22T10:33:05 <xorAxAx> that does it
2008-06-22T10:33:19 <xorAxAx> (checks __dict__ and if there is no entry, asks superclass)
2008-06-22T10:34:40 <johill> I was going to just use .__class__.mro()
2008-06-22T10:35:08 <xorAxAx> yes, probably better :)
2008-06-22T10:35:09 <johill> but I won nothing over dir()
2008-06-22T10:35:13 <johill> <bound method Config._default_password_checker of <wikiconfig_local.Config object at 0x103818b0>>
2008-06-22T10:35:26 <xorAxAx> huh, you got that out of __dict__?
2008-06-22T10:35:44 <johill> or __class__.__dict__, yes
2008-06-22T10:36:48 <xorAxAx> then your code must be wrong :)
2008-06-22T10:36:52 <xorAxAx> works for me here
2008-06-22T10:37:48 <johill> oh
2008-06-22T10:37:58 <johill> due to getattr, yeah
2008-06-22T10:38:07 <johill> I wasn't looking at the dict
2008-06-22T10:40:09 <xorAxAx> ?
2008-06-22T10:40:25 <johill> I was only using the __dict__ to find the keys and still using getattr
2008-06-22T10:40:27 <xorAxAx> getattr(i.__dict__, foo) should work fine
2008-06-22T10:40:35 <xorAxAx> hmm, no
2008-06-22T10:40:40 <johill> well I was using getattr(i, key) anyway
2008-06-22T10:40:45 <xorAxAx> yeah
2008-06-22T10:40:48 <johill> never mind I fixed it now
2008-06-22T10:40:57 <johill> by just returning the value from the iterator too
2008-06-22T10:41:15 <johill> it's working now
2008-06-22T10:42:01 <johill> shows 7 settings for DE
2008-06-22T10:42:41 <xorAxAx> dont forget to restrict the rendering for the macro to superusers only
2008-06-22T10:42:54 <johill> yeah I was wondering about that
2008-06-22T10:43:01 <johill> is superuser the right level?
2008-06-22T10:43:21 <johill> is there anything stronger?
2008-06-22T10:44:53 <xorAxAx> no
2008-06-22T10:46:32 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3763:0284379adbc8 1.8/MoinMoin/ (config/multiconfig.py macro/WikiConfig.py):
2008-06-22T10:46:32 <CIA-52> some config refactorings, add WikiConfig macro
2008-06-22T10:46:32 <CIA-52> The WikiConfig macro shows all wiki settings that have non-default values;
2008-06-22T10:46:32 <CIA-52> it is restricted to superusers.
2008-06-22T10:47:45 <johill> oh and settings that moin doesn't know about are shown in italic
2008-06-22T10:48:00 <johill> so if you write some_foo = 7 in your config it shows some_foo in italic
2008-06-22T10:51:13 <xorAxAx> probably a legend would make sense :)
2008-06-22T10:54:03 <johill> :)
2008-06-22T10:55:19 <johill> This table shows all settings in this wiki that do not have default values. Settings that the configuration system doesn't know about are shown in italic, those may be due to third-party extensions needing configuration or settings that were removed from Moin.
2008-06-22T10:57:29 <johill> so what do we do wrt. translation?
2008-06-22T10:57:44 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3764:0747f38efe23 1.8/MoinMoin/macro/WikiConfig.py: add explanatory text to WikiConfig macro
2008-06-22T10:59:29 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3765:a9c67b33a5e6 1.8/MoinMoin/macro/ (WikiConfig.py WikiConfigHelp.py): fix dependencies of WikiConfig and WikiConfigHelp macros
2008-06-22T11:00:07 <xorAxAx> johill: hmm
2008-06-22T11:00:41 <xorAxAx> johill: how about a separate _() function called differently that works with .po files that have config keys as msgid?
2008-06-22T11:01:01 <johill> aside from the techincal problem that we have no "render this text as wiki to html" function aside from the one built into _()
2008-06-22T11:01:29 <johill> xorAxAx: config keys for msgid isn't nice
2008-06-22T11:01:36 <johill> for example the help text I pasted above should really be in that file too
2008-06-22T11:01:55 <xorAxAx> yeah, indeed
2008-06-22T11:02:02 <johill> or that is a corner case, but group help texts etc. for sure
2008-06-22T11:02:43 <johill> plus if we ever change a description there's surely a reason so we want translators to notice
2008-06-22T11:02:45 <xorAxAx> well, then its easy and you modify the _() collector code to sort the msgs into the correct bucket (normal, config) and add _() everywhere
2008-06-22T11:03:05 <xorAxAx> i wonder which part of moin does the collection currently
2008-06-22T11:03:12 <johill> and add a config=True keyword to _()?
2008-06-22T11:03:34 <xorAxAx> isnt necessary
2008-06-22T11:03:42 <xorAxAx> the _() code itself can just load both files
2008-06-22T11:03:51 <xorAxAx> and lookup in the single dict as usual
2008-06-22T11:04:08 <xorAxAx> the collection of strings is the harder probleme
2008-06-22T11:05:21 <johill> tbh, I have no idea how i18n really works
2008-06-22T11:05:40 <xorAxAx> and i dont know which software component is doing the string collection
2008-06-22T11:06:05 <johill> there's a TextFinder in MoinMoin/i18n/tools/check_i18n.py
2008-06-22T11:06:56 <johill> but that file doesn't seem to write po files
2008-06-22T11:07:44 <johill> ahrg ahrg
2008-06-22T11:07:52 <johill> _("LDAP server %(server)s failed." % {'server': server})
2008-06-22T11:09:06 <xorAxAx> :-)
2008-06-22T11:10:00 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3731:6d89eb828dc9 1.7/MoinMoin/auth/ldap_login.py: ldap_login: fix wrong _() use
2008-06-22T11:10:07 <johill> but that doesn't actually collect the strings
2008-06-22T11:10:51 <johill> hmm
2008-06-22T11:10:52 <johill> I don't know
2008-06-22T11:10:57 <johill> have to go now anyway
2008-06-22T11:11:34 <dreimark> Walk through AST tree and collect text from gettext calls
2008-06-22T11:20:00 <dreimark> from check_i18n
2008-06-22T11:20:37 <dreimark> Usage: check_i18n.py [lang ...]
2008-06-22T11:54:44 <johill> dreimark: yeah but it doesn't write po files
2008-06-22T11:54:46 <johill> so what gives?
2008-06-22T12:24:13 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3766:6a618481be28 1.8/MoinMoin/config/multiconfig.py:
2008-06-22T12:24:13 <CIA-52> add help texts to configuration items
2008-06-22T12:24:13 <CIA-52> The display macro needs to format them as wiki still, right now the
2008-06-22T12:24:13 <CIA-52> help text is shown 'raw'.
2008-06-22T12:33:04 <dreimark> johill: http://wiki.wxpython.org/Internationalization#head-a5665108d7852c4069838327388fc11fbc524dfb
2008-06-22T12:46:53 <dreimark> found it it is in the i18n Makefile wiki2po.py
2008-06-22T12:47:37 <dreimark> which is in i18n/tools/
2008-06-22T12:47:50 <dreimark> johill:
2008-06-22T13:56:34 <johill> grzywacz: I killed cfg.subscribable_events/_subscribable_events, what was that for? I didn't see it used anywhere
2008-06-22T13:57:14 <johill> dreimark: wiki2po is to pull the translated files down
2008-06-22T14:02:18 <dreimark> and to write the po file
2008-06-22T14:03:56 <johill> ok but how do we collect strings from the source code?
2008-06-22T14:04:11 <johill> to write the original po file that translators use
2008-06-22T14:04:14 * dreimark looks at mk_POTFILES
2008-06-22T14:13:57 <dreimark> LOL
2008-06-22T14:14:18 <dreimark> johill: we should read README in i18n
2008-06-22T14:14:29 <dreimark> For using the i18n build system, one needs to have "gettext" installed.
2008-06-22T14:14:37 <dreimark> xgettext can extract these strings automatically from
2008-06-22T14:14:39 <dreimark> the files listed in POTFILES.in.
2008-06-22T14:15:35 <grzywacz> johill, events that appear in notification preferences that the user can subscribe to?
2008-06-22T14:16:27 <grzywacz> grep -r _subscribable_events MoinMoin | wc -l
2008-06-22T14:16:27 <grzywacz> 16
2008-06-22T14:30:09 <xorAxAx> hey my student
2008-06-22T14:30:22 <xorAxAx> grzywacz: can you decipher amadeus ticket dumps? :)
2008-06-22T14:30:44 <johill> grzywacz: well gread the grep result more carefully
2008-06-22T14:31:12 <johill> grzywacz: 7 in config
2008-06-22T14:31:17 <johill> grzywacz: and the others are get_subscribable_events
2008-06-22T14:31:25 <xorAxAx> johill: ?
2008-06-22T14:31:31 <xorAxAx> johill: get_subscribable_events is used only of couse
2008-06-22T14:31:58 <johill> xorAxAx: there was cfg._subscribable_events
2008-06-22T14:32:08 <johill> which I just killed because it isn't used anywhere
2008-06-22T14:32:12 <xorAxAx> ?
2008-06-22T14:32:14 <xorAxAx> its a cache
2008-06-22T14:32:15 <johill> hence, I asked grzywacz
2008-06-22T14:32:23 <johill> no
2008-06-22T14:32:29 <johill> cfg.subscribable_events isn't used either
2008-06-22T14:32:36 <xorAxAx> /config/multiconfig.py: self._subscribable_events = events.get_subscribable_events()
2008-06-22T14:32:39 <xorAxAx> loads the cache
2008-06-22T14:32:44 <xorAxAx> ./config/multiconfig.py: return getattr(self, "_subscribable_events")
2008-06-22T14:32:47 <xorAxAx> reads th cache
2008-06-22T14:33:21 <johill> look
2008-06-22T14:33:24 <johill> I understand the code
2008-06-22T14:33:29 <johill> tell me where it's used
2008-06-22T14:33:50 <grzywacz> xorAxAx, nope, amadeus is not us :p
2008-06-22T14:34:04 <xorAxAx> grzywacz: dont they use the same 3 char codes? :)
2008-06-22T14:34:05 <johill> besides, aren't events completely static? so they could be cached in the events module if needed?
2008-06-22T14:34:17 <grzywacz> xorAxAx, what do you want to know?
2008-06-22T14:34:20 <xorAxAx> johill: yes, indeed, its broken
2008-06-22T14:34:30 <xorAxAx> grzywacz: i want to reopen a case that happened a few years ago
2008-06-22T14:34:39 <xorAxAx> grzywacz: i want to understand why my tickets were cancelled
2008-06-22T14:34:48 <xorAxAx> you probably still remember the stroy
2008-06-22T14:34:56 <grzywacz> xorAxAx, uh. I'm just a little developer. ;)
2008-06-22T14:34:58 <xorAxAx> so i need somebody who can decifer these 3 char ticketing codes
2008-06-22T14:35:01 <johill> xorAxAx: well imho it shouldn't be cached in cfg either way
2008-06-22T14:35:10 <johill> so since it wasn't used, I removed it
2008-06-22T14:35:15 <grzywacz> xorAxAx, like, give me the list, I may be able to find someone who knows ;)
2008-06-22T14:35:32 <xorAxAx> johill: yeah, it was meant to be used as a lazy cache to avoid importing all modules if there are no events
2008-06-22T14:35:33 <johill> xorAxAx: events.get_subscribable_events() is entirely static. if there's need for a cache, it can do it
2008-06-22T14:35:47 <xorAxAx> but brokenly implemented
2008-06-22T14:36:34 <johill> xorAxAx: I don't understand, how can you avoid importing the events module?
2008-06-22T14:37:57 <johill> and what modules are you talking about?
2008-06-22T14:38:04 <johill> I think you're confusing it with notification modules? rather than events?
2008-06-22T14:38:09 <xorAxAx> hmm, indeed
2008-06-22T14:38:31 <johill> that's entirely unrelated though
2008-06-22T14:38:51 <xorAxAx> yep
2008-06-22T14:38:54 <johill> all events are just classes defined in events/__init__.py
2008-06-22T14:39:13 <johill> so now do you still see any reason for caching something like that or putting it into cfg?
2008-06-22T14:39:31 <grzywacz> :)
2008-06-22T14:39:34 <johill> now, get_subscribable_events is fairly expensive
2008-06-22T14:39:35 <johill> it could cache
2008-06-22T14:39:43 <johill> but that's trivial to do at the module level just in that function
2008-06-22T14:40:24 <xorAxAx> grzywacz: ok, i will scan it someday and send it to you
2008-06-22T14:41:13 <johill> so anyhow, I'm convinced it's bogus, just wanted to ask in case there was supposed to be a reason for it and it was just done wrong
2008-06-22T14:42:01 <grzywacz> johill, I guess not.
2008-06-22T14:42:11 <johill> ok
2008-06-22T14:42:13 <johill> thanks
2008-06-22T14:44:30 <johill> xorAxAx: cfg.event_handlers behaves exactly like you were saying :)
2008-06-22T14:44:40 <xorAxAx> hehe
2008-06-22T14:46:22 <johill> some ugly code though
2008-06-22T14:46:41 <johill> with getattr(self, '_event_handlers', None) rather than setting _event_handlers = None by default
2008-06-22T14:50:31 <johill> dreimark: any idea how to fix the embed tests? make it allow both audio/ and application/?
2008-06-22T14:52:16 <CIA-52> Johannes Berg <johannes AT sipsolutions DOT net> default * 3767:0b7f7b9226df 1.8/MoinMoin/config/multiconfig.py: config: improve some lazy-loading code
2008-06-22T14:54:03 <dreimark> johill: please show the assertion failure of the test
2008-06-22T14:56:02 <dreimark> check if is fixed if you add your mimetype to mimetypes_embed
2008-06-22T14:56:56 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 89:83abc483373a 1.7-extensions/data/plugin/parser/text_x_arnica.py: text_x_arnica: internal _get_files routine added to select only valid image files
2008-06-22T14:56:56 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 90:7e0445f8120e 1.7-extensions/data/plugin/parser/text_x_arnica.py: renamed get_files to define_thumb_webnails
2008-06-22T14:56:57 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 91:16d762c9c3e7 1.7-extensions/data/plugin/action/arnica_slides.py: arnica_slides: fixed missing html closing
2008-06-22T14:56:58 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 92:bb0f35d0e9c5 1.7-extensions/data/plugin/parser/text_x_arnica.py: text_x_arnica: only item file is tested for isPicture
2008-06-22T14:57:00 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 93:64797a5fc3d9 1.7-extensions/data/plugin/parser/text_x_arnica.py: text_x_arnica: check alias for [[ and ]]
2008-06-22T14:57:03 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 94:68102c2191f9 1.7-extensions/data/plugin/parser/text_x_arnica.py: text_x_arnica: small refactoring of html_* methods
2008-06-22T14:57:06 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 95:564e8ffe3d4b 1.7-extensions/data/plugin/macro/arnica.py: macro.arnica: fixed typo in docstring
2008-06-22T14:57:09 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 96:63da0711188e 1.7-extensions/data/plugin/action/arnica_slides.py: arnica_slides: fixed bug in comparison for image_for_webnail
2008-06-22T16:01:19 <johill> dreimark: it would not be fixed
2008-06-22T16:01:33 <johill> > assert 'application/ogg' in u'Current configuration does not allow embedding of the file test.ogg because ...eForEmbedObject?action=AttachFile&do=get&target=test.ogg">test.ogg</a>'
2008-06-22T16:04:28 <johill> dreimark: so it'd still fail because application/ogg won't be in the embedding code
2008-06-22T16:10:00 <dreimark> johill: it needs only to be added to the config var mimetypes_embed
2008-06-22T16:11:40 <johill> yes I know but it'll still fail
2008-06-22T16:11:53 <johill> because the html will not contain applicaiont/ogg, it'll contain audio/ogg
2008-06-22T16:12:28 <dreimark> aha, yes i see
2008-06-22T16:13:25 <dreimark> but you should now have less failing tests or ?
2008-06-22T16:14:23 <johill> yes, one less
2008-06-22T16:14:29 <johill> E assert mimetype in result
2008-06-22T16:14:29 <johill> > assert 'application/ogg' in u'\n<object data="./AutoCreatedMoinMoinTemporaryTestPageForEmbedObject?action=...e="hidden" value="False" valuetype="data"><p>Embedded audio/ogg</p>\n</object>'
2008-06-22T16:14:39 <johill> E assert 'value="transparent"' in result
2008-06-22T16:14:39 <johill> > assert 'value="transparent"' in u'\n<object data="./AutoCreatedMoinMoinTemporaryTestPageForEmbedObject?action=...e="hidden" value="False" valuetype="data"><p>Embedded audio/ogg</p>\n</object>'
2008-06-22T16:16:23 <dreimark> I change the test to the mpeg file
2008-06-22T16:18:41 <dreimark> and kill ogg, because i need one other for application
2008-06-22T16:18:46 <dreimark> johill:
2008-06-22T16:21:53 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 97:477255e09c8c 1.7-extensions/data/plugin/action/arnica_slides.py: arnica_slides: simple css elements added based on a mockup from Radomir Dopieralski
2008-06-22T16:22:26 <johill> dreimark: hmm
2008-06-22T16:22:39 <johill> video/mpeg mpeg mpg mpe
2008-06-22T16:26:11 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3732:f413e3c28df6 1.7/MoinMoin/macro/_tests/test_EmbedObject.py: test_EmbedObject: removed testing for application/ogg because the mime type of ogg files was changed to audio/ogg. Added a testcase for application/pdf.
2008-06-22T16:29:39 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3768:b03223ab1818 1.8/MoinMoin/macro/_tests/test_EmbedObject.py: test_EmbedObject: removed testing for application/ogg because the mime type of ogg files was changed to audio/ogg. Added a testcase for application/pdf.
2008-06-22T16:29:53 <johill> that'll even work here :)
2008-06-22T16:30:18 <dreimark> good
2008-06-22T16:30:22 * dreimark fixes Makefile
2008-06-22T16:30:38 <johill> no failures now
2008-06-22T16:32:34 * dreimark has test_packages.py[6] ....F.
2008-06-22T16:33:03 <dreimark> def testAllCreate(self):
2008-06-22T16:33:18 <dreimark> zf.write(filename.encode("cp437"), zipname)
2008-06-22T16:33:27 <dreimark> E [failure to get at sourcelines from <TracebackEntry /home/ree/workspace/moin-1.8/encodings/cp437.py:12>]
2008-06-22T16:33:46 <dreimark> UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 78: ordinal not in range(128)
2008-06-22T16:34:22 <johill> well I tested only embed the other stuff was fine before
2008-06-22T16:35:06 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3733:6879f85a6faa 1.7/Makefile: fixed path for check_i18n
2008-06-22T16:36:27 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3769:8a6960133ec2 1.8/Makefile: fixed path for check_i18n
2008-06-22T16:38:42 <dreimark> huh, in 1.7 this test is skipped
2008-06-22T16:41:09 <johill> weird
2008-06-22T16:42:36 * dreimark copies the traceback
2008-06-22T16:43:19 <ThomasWaldmann> re
2008-06-22T16:43:24 <johill> can't really help, I have to go in a minute
2008-06-22T16:46:34 <dreimark> cu
2008-06-22T16:46:47 <dreimark> http://moinmo.in/ReimarBauer?action=AttachFile&do=view&target=18dev_test_package_failing_test.txt
2008-06-22T16:47:18 <dreimark> it should skip in 1.8 too but doesn't
2008-06-22T16:57:05 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3770:749d7148b080 1.8/MoinMoin/script/account/ (create.py disable.py resetpw.py):
2008-06-22T16:57:05 <CIA-52> script.account: create user checks for already known email adress or username,
2008-06-22T16:57:05 <CIA-52> disable and resetpw checks if the user exists
2008-06-22T17:01:41 <ThomasWaldmann> dreimark: filename = AttachFile.getFilename(self.request, page.page_name, attname) < SENSE?
2008-06-22T17:01:53 <ThomasWaldmann> (from the traceback you posted above)
2008-06-22T17:03:23 <ThomasWaldmann> > zf.write(filename.encode("cp437"), zipname) < SENSE?
2008-06-22T17:07:13 <ThomasWaldmann> obviously that .encode("cp437") is wrong and needs to get removed
2008-06-22T17:07:39 <dreimark> sure but why does this test fail now and not in 1.7
2008-06-22T17:08:08 <ThomasWaldmann> it depends on the attachment names you have in the test
2008-06-22T17:09:48 <dreimark> yeah
2008-06-22T17:12:35 <ThomasWaldmann> dreimark: please fix in 1.6, 1.7 (and then pull from 1.7 into 1.8)
2008-06-22T17:13:46 * ThomasWaldmann looks at traceback_log_dir
2008-06-22T17:14:17 <dreimark> a view days ago I asked xorAxAx about this .encode("cp437") and he explained that this is the standard definition for zip files
2008-06-22T17:14:22 <dreimark> http://moinmo.in/MoinMoinChat/Logs/moin-dev/2008-06-12
2008-06-22T17:14:38 <dreimark> ThomasWaldmann:
2008-06-22T17:20:02 <ThomasWaldmann> obviously we did not look at what was required at that place
2008-06-22T17:20:44 <ThomasWaldmann> the first argument to write has to be a filename in the file system, not a filename for use withing the zip file (that's the 2nd arg)
2008-06-22T17:27:57 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3734:e458af23d64d 1.7/MoinMoin/action/PackagePages.py: PackagePages.collectpackage: removed encoding from file name of zipfile
2008-06-22T17:28:21 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 2666:349056d118f2 1.6/MoinMoin/action/PackagePages.py: PackagePages.collectpackage: removed encoding from file name of zipfile
2008-06-22T17:30:54 <dreimark> added 4 changesets with 2 changes to 4 files (+1 heads)
2008-06-22T17:31:02 <dreimark> if I do hg pull ../moin-1.7
2008-06-22T17:32:14 <dreimark> ThomasWaldmann: seems normal or ?
2008-06-22T17:37:14 <dreimark> bbl
2008-06-22T17:40:45 <dreimark> http://moinmo.in/MacroMarket/FormCreate
2008-06-22T17:46:25 <ThomasWaldmann> dreimark: you need to merge after pulling
2008-06-22T18:16:03 * ThomasWaldmann changes failure.py
2008-06-22T18:46:53 <CIA-52> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3771:c57d79281471 1.8/ (3 files in 3 dirs): error handling: remove cfg.traceback_log_dir (we just use logging.exception), only output text tracebacks if allowed by cfg.traceback_show
2008-06-22T18:49:33 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3776:4a2e168399ca 1.8/MoinMoin/ (action/PackagePages.py auth/ldap_login.py): branch merge
2008-06-22T18:58:12 <ThomasWaldmann> zenhase: for improving the global error handler it would be nice to know the current's request output mimetype
2008-06-22T18:58:49 <ThomasWaldmann> s/current's request/current request's/ :)
2008-06-22T18:59:16 <ThomasWaldmann> currently the error handler emits htmls without knowing if that is appropriate
2008-06-22T19:01:27 <ThomasWaldmann> also it does not care about whether a content-type header is already emitted or not
2008-06-22T19:02:44 <ThomasWaldmann> johill: you should add your name to header of multiconfig
2008-06-22T19:11:10 * ThomasWaldmann edits multiconfig
2008-06-22T19:20:03 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 2667:44c6c172639c 1.6/MoinMoin/_tests/test_packages.py: test_package: fixed broken testAllCreate
2008-06-22T19:20:16 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3735:e6de0a13519c 1.7/MoinMoin/_tests/test_packages.py: test_package: fixed broken testAllCreate
2008-06-22T19:21:24 <CIA-52> Reimar Bauer <rb.proj AT googlemail DOT com> default * 3778:7ce6ee352251 1.8/MoinMoin/_tests/test_packages.py: branch merge
2008-06-22T20:09:36 <CIA-52> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3779:9aad6c344482 1.8/MoinMoin/config/multiconfig.py: multiconfig: PEP8 fixes, typos, no code change
2008-06-22T20:09:37 <CIA-52> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3780:1190e9351eea 1.8/MoinMoin/ (3 files in 3 dirs): merged main
2008-06-22T20:22:57 * ThomasWaldmann adds help texts to multiconfig
2008-06-22T21:40:49 <CIA-52> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3781:a637edbeb5fa 1.8/MoinMoin/config/multiconfig.py: multiconfig: added more setting descriptions
2008-06-22T21:41:56 <ThomasWaldmann> johill: nice WikiConfig* macros :)
2008-06-22T22:02:10 <dreimark> bbl
2008-06-22T22:15:14 <gizmach> gn
2008-06-22T22:33:25 <dreimark> good night melita
2008-06-22T22:33:28 <dreimark> gizmach:
2008-06-22T22:35:41 <dreimark> johill: some additional thoughts about CopyPage, with hacl it needs to be disabled on default independent from actions_excluded.
2008-06-22T22:38:22 <dreimark> good night
2008-06-22T22:39:09 <ThomasWaldmann> gn dreimark
2008-06-22T22:52:29 <CIA-52> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3782:791011b5c3fd 1.8/MoinMoin/config/multiconfig.py: multiconfig: better grouping of settings
2008-06-22T22:52:49 <ThomasWaldmann> hi LotekThirteen
2008-06-22T22:55:02 <ThomasWaldmann> johill: i guess it would be better manageable without sorting and without 'options' (just using options_no_groupname and sorting manually)
2008-06-22T22:55:40 <xorAxAx> ThomasWaldmann: why without sorting?
2008-06-22T22:55:46 <ThomasWaldmann> usually one wants the most important settings at top of the table and maybe group settings together (that are not necessarily sorting together)
2008-06-22T23:05:44 <dreimark> another one problem I don't want to forget is about RenamePage and hacl. The savest solution is to exclude it too. Otherwise one can rename just only the page which defines the acls or a subpage to an other hierarchy page or just a new pagename with no acls afterwards.
2008-06-22T23:05:51 <xorAxAx> the latter issue sounds like subgroups :-) but importance is an issue, yep
2008-06-22T23:06:20 <dreimark> bye for today MoinMoin: MoinMoinChat/Logs/moin-dev/2008-06-22 (last edited 2008-06-21 22:15:02 by IrcLogImporter)
