1 2012-05-14T00:00:29  *** dwcramer has quit IRC
   2 2012-05-14T00:13:29  *** dwcramer has joined #moin-dev
   3 2012-05-14T01:44:47  *** spy has quit IRC
   4 2012-05-14T04:57:44  *** dwcramer has quit IRC
   5 2012-05-14T06:18:03  *** Trip_ has joined #moin-dev
   6 2012-05-14T06:21:15  *** MattMaker has quit IRC
   7 2012-05-14T07:38:52  *** Trip_ has quit IRC
   8 2012-05-14T07:45:57  *** MattMaker has joined #moin-dev
   9 2012-05-14T08:21:31  <dreimark> moin
  10 2012-05-14T09:29:56  *** MattMaker has quit IRC
  11 2012-05-14T09:38:33  *** MattMaker has joined #moin-dev
  12 2012-05-14T09:52:53  *** MattMaker has quit IRC
  13 2012-05-14T10:08:33  <breton> ThomasWaldmann: I wonder, is it possible to have something like a global meta for item? One file or entry somewhere for each itemid. Is there something like this in moin2 now?
  14 2012-05-14T10:09:43  <breton> we could keep there th whole branching stuff
  15 2012-05-14T10:23:55  *** xjjk has quit IRC
  16 2012-05-14T10:24:07  *** xjjk has joined #moin-dev
  17 2012-05-14T10:48:31  <breton> hmm, where is ACL is saved now?
  18 2012-05-14T10:48:49  <breton> s/ACL is/ACL/
  19 2012-05-14T10:49:02  <dreimark> breton: in meta
  20 2012-05-14T10:50:33  <dreimark> http://readthedocs.org/docs/moin-20/en/latest/devel/api/MoinMoin.storage.html?highlight=acl
  21 2012-05-14T11:29:33  *** greg_f has joined #moin-dev
  22 2012-05-14T11:42:44  <ThomasWaldmann> breton: first define precisely what you need, then we see how to store
  23 2012-05-14T12:30:39  * xiaq is reading jinja templates
  24 2012-05-14T12:31:02  * xiaq is wondering where to put summary on the page :)
  25 2012-05-14T12:35:18  <xiaq> question: how can i embed the werkzeug debugger into the page?
  26 2012-05-14T12:37:20  <xiaq> the way i used to debug templates is inserting [i]pdb.set_trace into the template and dubug in the terminal, but since werkzeug provides a nice web debugger ;) let's try it
  27 2012-05-14T12:53:16  <xiaq> another question: does Jinja has filter corresponding to str.format? the format filters corresponds to % operator
  28 2012-05-14T12:53:31  <xiaq> s/filter/builtin filter/g
  29 2012-05-14T12:56:49  <ronny> xiaq: i dont think so, but it should be easy to make one, about 3-4 lines or so
  30 2012-05-14T12:58:40  * xiaq sticks to Jinja builtin format
  31 2012-05-14T12:59:57  * xiaq gets TypeError with         {{ _('Summary: %s')|format(item.meta['summary']) }}
  32 2012-05-14T13:00:00  <xiaq>         {{ _('Summary: %s')|format(item.meta['summary']) }}
  33 2012-05-14T13:00:12  <xiaq> what's wrong?
  34 2012-05-14T13:00:33  <xiaq> {{ 'Summary: %s'|format(item.meta['summary']) }} works fine
  35 2012-05-14T13:02:04  <ronny> no idea? whats the type _ returns?
  36 2012-05-14T13:02:39  <xiaq> don't know yet, since the Werkzeug debugger interprets _ as last evaluated value...
  37 2012-05-14T13:04:03  <xiaq> seems to have something to do with unicode/str stuff...
  38 2012-05-14T13:06:21  * xiaq is finally going to learn the terrible unicode/str lessons every Python programmer has to learn
  39 2012-05-14T13:10:25  <dreimark> xiaq: watch http://nedbatchelder.com/text/unipain.html
  40 2012-05-14T13:10:52  <dreimark> _ is used normally for i18n
  41 2012-05-14T13:11:05  <xiaq> dreimark: yeah known this
  42 2012-05-14T13:11:30  <xiaq> dreimark: thx for the link, but gtg soon
  43 2012-05-14T13:12:10  <dreimark> which type is item.meta['summary'] ?
  44 2012-05-14T13:12:48  <xiaq> unicode
  45 2012-05-14T13:14:06  <xiaq> Jinja doc has:
  46 2012-05-14T13:14:06  <xiaq> {{ _('Hello %(user)s!')|format(user=user.username) }}
  47 2012-05-14T13:14:18  <dreimark> well that both is usually ascii
  48 2012-05-14T13:14:19  <xiaq> so i just changed to:
  49 2012-05-14T13:14:22  <xiaq> {{ _('Summary: %(summary)s')|format(summary=item.meta['summary']) }}
  50 2012-05-14T13:14:31  <dreimark> and ascii is 1byte code in unicode too
  51 2012-05-14T13:14:35  <xiaq> which gives me "TypeError: not all arguments converted during string formatting
  52 2012-05-14T13:15:04  <xiaq> dreimark: Ha, as CJK person i knew Unicode and utf8 well already
  53 2012-05-14T13:15:24  <xiaq> dreimark: i just avoided processing unicode in Python every time :)
  54 2012-05-14T13:15:55  * xiaq gtg
  55 2012-05-14T13:15:59  <dreimark> true, we have some simlar problems in our lang
  56 2012-05-14T13:19:55  <dreimark> format(value, *args, **kwargs)
  57 2012-05-14T13:19:57  <dreimark> Apply python string formatting on an object:
  58 2012-05-14T13:36:02  *** spy has joined #moin-dev
  59 2012-05-14T13:57:14  *** dwcramer has joined #moin-dev
  60 2012-05-14T14:23:01  *** MattMaker has joined #moin-dev
  61 2012-05-14T14:59:34  *** dwcramer has quit IRC
  62 2012-05-14T15:14:11  *** kanha_ has joined #moin-dev
  63 2012-05-14T15:15:09  *** kanha has quit IRC
  64 2012-05-14T15:42:22  *** dave_largo has joined #moin-dev
  65 2012-05-14T15:48:14  <breton> ThomasWaldmann: http://etherpad.osuosl.org/moin2-bms 28-30, 49, 51 and 24 regarging multiple parentid. Will read the comments in ~8h from now.
  66 2012-05-14T15:49:59  <breton> gn
  67 2012-05-14T16:23:14  *** dwcramer has joined #moin-dev
  68 2012-05-14T16:24:10  * ThomasWaldmann answered on breton's etherpad
  69 2012-05-14T17:27:59  <MaikuMori> moin
  70 2012-05-14T17:40:54  *** DasIch_ has joined #moin-dev
  71 2012-05-14T17:41:01  *** DasIch has quit IRC
  72 2012-05-14T17:41:01  *** DasIch_ is now known as DasIch
  73 2012-05-14T18:03:06  *** dwcramer has quit IRC
  74 2012-05-14T18:44:24  <ThomasWaldmann> hi MaikuMori
  75 2012-05-14T18:44:30  <MaikuMori> Heya
  76 2012-05-14T18:48:22  <ThomasWaldmann> spy and me have just found we have the same "how to present this thing" issue as for the ticket items
  77 2012-05-14T18:48:35  <ThomasWaldmann> (for the blog items in that case)
  78 2012-05-14T18:49:38  <ThomasWaldmann> and I think it might repeat for DasIch and the pastebin, so it would be good if you all discuss how to solve best
  79 2012-05-14T18:50:39  <ThomasWaldmann> DasIch: there are some ideas in the ticket system etherpad, but they need more thinking / details
  80 2012-05-14T18:51:05  *** dwcramer has joined #moin-dev
  81 2012-05-14T18:55:50  <MaikuMori> ThomasWaldmann: have you read http://moinmo.in/EugeneSyromyatnikov/PluginConcept ?
  82 2012-05-14T18:59:53  *** dwcramer has quit IRC
  83 2012-05-14T19:01:32  *** greg_f has quit IRC
  84 2012-05-14T19:02:27  * ThomasWaldmann just reading it
  85 2012-05-14T19:04:34  <ronny> scary
  86 2012-05-14T19:10:45  *** dwcramer has joined #moin-dev
  87 2012-05-14T19:14:36  *** esyr has quit IRC
  88 2012-05-14T19:14:53  *** esyr has joined #moin-dev
  89 2012-05-14T19:22:00  *** esyr_ has joined #moin-dev
  90 2012-05-14T19:22:00  *** esyr has quit IRC
  91 2012-05-14T19:23:53  * ThomasWaldmann hopes we won't reinvent zope2 :D
  92 2012-05-14T19:28:04  *** esyr_ has quit IRC
  93 2012-05-14T19:46:26  *** esyr has joined #moin-dev
  94 2012-05-14T20:06:49  *** esyr has quit IRC
  95 2012-05-14T20:08:28  *** esyr has joined #moin-dev
  96 2012-05-14T20:14:28  *** esyr has quit IRC
  97 2012-05-14T20:23:17  *** esyr has joined #moin-dev
  98 2012-05-14T20:29:17  <esyr> hm.
  99 2012-05-14T20:31:09  <esyr> ThomasWaldmann: MaikuMori: is it only me or you get 504 when trying to get history/ol page revs on ep too?
 100 2012-05-14T20:40:35  *** esyr has quit IRC
 101 2012-05-14T20:43:45  <ThomasWaldmann> yup, 504 here too.
 102 2012-05-14T20:50:08  *** esyr has joined #moin-dev
 103 2012-05-14T20:57:41  *** esyr has quit IRC
 104 2012-05-14T20:58:25  *** esyr has joined #moin-dev
 105 2012-05-14T21:05:22  <ThomasWaldmann> esyr: yup, 504 here too.
 106 2012-05-14T21:05:40  <esyr> that's bad.
 107 2012-05-14T21:06:05  <esyr> eventually during reconnect i loss revision where paste problem haven't been introduced yet.
 108 2012-05-14T21:06:11  <esyr> *i lost
 109 2012-05-14T21:06:31  <esyr> ok, try to remember what was there.
 110 2012-05-14T21:10:11  <MaikuMori> ThomasWaldmann: regarding interfaces being 'javaish'
 111 2012-05-14T21:10:15  <MaikuMori> Well it's a design pattern, most of them are used in Java.
 112 2012-05-14T21:11:28  <MaikuMori> ABC's are almost the same. Anyhow I agree that zope is more or less this legacy package which everyones tries to forget.
 113 2012-05-14T21:11:43  <ThomasWaldmann> no ' for plurals, please :)
 114 2012-05-14T21:12:43  <MaikuMori> Alright :)
 115 2012-05-14T21:14:17  <MaikuMori> The only reason I brought it up is because many plugin systems use zope.interface. I personally would choose ABCs.
 116 2012-05-14T21:22:34  * ThomasWaldmann looks at zope.interface
 117 2012-05-14T21:31:06  <MaikuMori> Hmm, ep died.
 118 2012-05-14T21:34:18  <esyr> MaikuMori: looks like there's only one thread for all :\
 119 2012-05-14T21:34:28  * esyr tried to reload those 504 pages
 120 2012-05-14T21:34:35  <MaikuMori> It's back now.
 121 2012-05-14T21:36:48  <esyr> yeah.
 122 2012-05-14T21:36:48  <esyr> so, now you know how to ddos ep :)
 123 2012-05-14T21:41:21  *** esyr has quit IRC
 124 2012-05-14T21:44:07  <ThomasWaldmann> well, i looked at z.i, and i somehow can't help: i don't like it
 125 2012-05-14T21:45:43  <ThomasWaldmann> it's not PEP8, it's somehow looks javaesque and it's not really needed (convince me if you think otherwise :)
 126 2012-05-14T21:47:25  <MaikuMori> I don't think I even want to try.
 127 2012-05-14T21:47:55  <MaikuMori> Have you seen the recent versions of the plugin EP?
 128 2012-05-14T21:48:56  <MaikuMori> Some of the proposed stuff seems way out of scope at the moment. Correct me if I'm wrong
 129 2012-05-14T21:49:36  *** esyr has joined #moin-dev
 130 2012-05-14T21:50:23  <ThomasWaldmann> MaikuMori: the ep needs better separation, in fundamental stuff and in applications of that (e.g. for moin)
 131 2012-05-14T21:51:01  <MaikuMori> Agree. I meant the 'plugins provide tests' and similar stuff
 132 2012-05-14T21:52:29  <ThomasWaldmann> line?
 133 2012-05-14T21:53:44  <MaikuMori> Ah it's a bit changed now 158, 162
 134 2012-05-14T21:53:57  <MaikuMori> also 123
 135 2012-05-14T21:54:35  <MaikuMori> Alright I'm off to bed. Will do a lot more work tomorrow.
 136 2012-05-14T21:58:20  *** esyr has quit IRC
 137 2012-05-14T22:02:22  *** esyr has joined #moin-dev
 138 2012-05-14T22:03:11  <ThomasWaldmann> esyr: your internet connection seems quite unstable right now
 139 2012-05-14T22:13:23  <ronny> MaikuMori: you wrote that plugin proposal?
 140 2012-05-14T22:13:39  <MaikuMori> Which one?
 141 2012-05-14T22:13:53  <ronny> http://moinmo.in/EugeneSyromyatnikov/PluginConcept
 142 2012-05-14T22:14:09  <MaikuMori> Nope, it's by EugeneSyromyatnikov
 143 2012-05-14T22:14:15  <ronny> ok
 144 2012-05-14T22:14:25  <ronny> its totally overengineered
 145 2012-05-14T22:15:39  <ThomasWaldmann> ronny: your suggestion?
 146 2012-05-14T22:16:18  <ronny> start from current reality instead and engineer well selected extension points
 147 2012-05-14T22:16:33  *** dwcramer has quit IRC
 148 2012-05-14T22:16:39  <ronny> i.e. backends, authentication, actions
 149 2012-05-14T22:16:52  <ronny> well, and themes
 150 2012-05-14T22:17:25  <ronny> all ofthose things have well defined and/or well known interfaces
 151 2012-05-14T22:18:32  <ronny> some need to handle contect creation, others just request/response or data
 152 2012-05-14T22:19:10  <ronny> gien how they work its even interesting to see if there is a way to handle some of it in wikiconfig instead of discovery
 153 2012-05-14T22:19:24  <ronny> i.e. actions and backends and auth
 154 2012-05-14T22:19:39  <ronny> macros may have some merit
 155 2012-05-14T22:19:55  <ronny> but on the other hand its trobulesome if system level installs break wikipages due to removing one
 156 2012-05-14T22:20:14  <ronny> so well
 157 2012-05-14T22:20:25  <ronny> im not sure if autodiscovery alone makes sense
 158 2012-05-14T22:20:33  <ronny> except for naming them for easy use in a config
 159 2012-05-14T22:21:05  <ronny> which leaves one withh well defined intefaces already there, entry-points for discovery and config for enabling
 160 2012-05-14T22:21:11  <ronny> do you follow?
 161 2012-05-14T22:21:21  <ThomasWaldmann> MaikuMori: ^^ (hope you read it tomorrow in the logs)
 162 2012-05-14T22:22:03  <MaikuMori> Reading it now, but yeah consider me already away. Those are good points.
 163 2012-05-14T22:23:19  <ThomasWaldmann> i agree about keeping good interfaces we already have
 164 2012-05-14T22:25:08  <ThomasWaldmann> and if one wants to be able to have some extensions NOT enabled, the easiest way is to just list all stuff one wants to have
 165 2012-05-14T22:26:18  <ronny> who is EugeneSyromyatnikov? esyr?
 166 2012-05-14T22:26:27  <ThomasWaldmann> yes
 167 2012-05-14T22:26:38  <ThomasWaldmann> much easier :D
 168 2012-05-14T22:26:39  <ronny> k
 169 2012-05-14T22:27:19  <ronny> esyr: you commited the basic crime of premature generalization + use of java patterns
 170 2012-05-14T22:28:37  <ThomasWaldmann> :D
 171 2012-05-14T22:29:34  <ronny> thats how zope started :P
 172 2012-05-14T22:29:39  <esyr> bwahahahah.
 173 2012-05-14T22:29:59  <esyr> ThomasWaldmann: connectivity from collocation to freenode, yes.
 174 2012-05-14T22:30:32  <ronny> esyr: how do you think about it after my notes?
 175 2012-05-14T22:30:41  <esyr> i'm thinking.
 176 2012-05-14T22:31:13  <esyr> ronny: well, i can't see how my proposal conflicts with your notes currently.
 177 2012-05-14T22:31:26  <ronny> esyr: its an antithesis
 178 2012-05-14T22:31:40  *** dwcramer has joined #moin-dev
 179 2012-05-14T22:31:42  <ronny> also note that in code registration is useless
 180 2012-05-14T22:31:47  <esyr> i mean, i can't see why predefined extension points can't live together with generic modular core.
 181 2012-05-14T22:32:04  <ronny> esyr: why is there need for a generic modular core to begin with?
 182 2012-05-14T22:32:21  <ronny> moin is not an endles flexible application frameworks
 183 2012-05-14T22:32:37  <esyr> ronny: why not? :)
 184 2012-05-14T22:32:51  <ronny> esyr: overengineering sucks
 185 2012-05-14T22:32:51  <esyr> i mean, i'd like to use moin in any imaginable way.
 186 2012-05-14T22:32:58  <esyr> and it fits!
 187 2012-05-14T22:33:01  <esyr> 1.x mostly.
 188 2012-05-14T22:33:25  <esyr> it has some drawbacks though, but it suits for nearly any application.
 189 2012-05-14T22:33:29  <ronny> but well, we have storages, we have actions, we have auth, we have themes, we have macros
 190 2012-05-14T22:33:38  <ronny> all of those have interfaces already
 191 2012-05-14T22:33:57  <ronny> so we just need ways to register new ones with a name -> entrypoints
 192 2012-05-14T22:34:13  <esyr> ronny: yes, for me action is rather  tight restriction.
 193 2012-05-14T22:34:37  <ronny> hmm, what would you like to see?
 194 2012-05-14T22:34:50  <ronny> note tht there are contenttypes and their neogation
 195 2012-05-14T22:35:12  <ronny> so crud actions fit very many things
 196 2012-05-14T22:35:13  <esyr> well, i'd like to have possibility to redefine and extend any part of moin.
 197 2012-05-14T22:35:22  <ronny> speak in usecases
 198 2012-05-14T22:35:29  <ronny> not abstract gibberish
 199 2012-05-14T22:35:35  <esyr> ronny: for example, if i want to integrate moin with ldap.
 200 2012-05-14T22:35:46  <ronny> for auth?
 201 2012-05-14T22:35:49  <ronny> thats auth
 202 2012-05-14T22:35:51  <esyr> for auth and acl.
 203 2012-05-14T22:36:12  <esyr> what should i do if i want to get acl from ldap? or from authz?
 204 2012-05-14T22:36:29  <esyr> for different paths.
 205 2012-05-14T22:36:38  <ronny> ThomasWaldmann: any intents to have external acl?
 206 2012-05-14T22:36:50  <esyr> i need to have appropriate entry points for redefining permission checks, fine.
 207 2012-05-14T22:36:58  <ronny> esyr: basically you need a router that cobines metadata with ldap data
 208 2012-05-14T22:37:07  <ronny> since acl is in meta
 209 2012-05-14T22:37:09  <esyr> i don't say this is big problem in particular.
 210 2012-05-14T22:37:25  <ronny> the thing is, dont inent absolute overly general interfaces
 211 2012-05-14T22:37:37  <esyr> ronny: well, in authz approach all permissions for different paths are stored in external file, for example.
 212 2012-05-14T22:37:42  <esyr> (that's other usexase)
 213 2012-05-14T22:37:42  <ronny> they make programming things harder and more confusing than necessary
 214 2012-05-14T22:37:52  <esyr> well, anyway.
 215 2012-05-14T22:38:02  <esyr> i don't insist on my approach.
 216 2012-05-14T22:38:13  <esyr> i don't argue that it's not overengineered.
 217 2012-05-14T22:38:23  <esyr> (mmm, double negative, sorry)
 218 2012-05-14T22:38:33  <ThomasWaldmann> well, I'ld like a rather general approach, but it should be easy to use
 219 2012-05-14T22:38:53  <esyr> but i think approach should be more general than in 1.x.
 220 2012-05-14T22:38:59  <ronny> esyr: exotic uecses that make sense will help reshaping the interfaces we have
 221 2012-05-14T22:39:08  <esyr> at least, i want to have ability to redefine each part of moin.
 222 2012-05-14T22:39:11  <ronny> storage is more flexible now
 223 2012-05-14T22:39:20  <esyr> but not at cost of geeral use, of course.
 224 2012-05-14T22:39:50  <ronny> esyr: a meta goal like "redefine each part" is bound to mess with general use
 225 2012-05-14T22:40:05  <ronny> break it down to use cases and reshape the apis in small iterations
 226 2012-05-14T22:40:20  <ronny> that tends to create less pain than a big overgeneral start
 227 2012-05-14T22:40:37  <ronny> i have a few projects that still hurt from the overgeneral start
 228 2012-05-14T22:41:04  <esyr> ronny: well, for me, i think, that proper definition of interfaces and policies is enough for this.
 229 2012-05-14T22:41:25  <esyr> moin1.x is build adhoc in many cases, so redefining it is pai.
 230 2012-05-14T22:41:30  <esyr> *is built
 231 2012-05-14T22:41:35  <esyr> *in many places
 232 2012-05-14T22:41:41  <esyr> *is pain
 233 2012-05-14T22:42:10  <esyr> moin2 is better in these terms, we just should settle some policy related to this.
 234 2012-05-14T22:42:13  <ThomasWaldmann> we are not reimplementing 1.x plugins
 235 2012-05-14T22:42:29  <esyr> of course, this in meaningless in 2.x context.
 236 2012-05-14T22:42:46  <esyr> (but we can provide appropriate pluginprovider, hehe)
 237 2012-05-14T22:44:04  <ThomasWaldmann> no need
 238 2012-05-14T22:46:17  <ThomasWaldmann> esyr: would a shell account here help you for irc backlog/connectivity?
 239 2012-05-14T22:46:24  <ronny> esyr: so work on the interfaces, and then we can just use entryoints to make additional implementations known, and enable them by lisinh in a wikiconfig
 240 2012-05-14T22:47:50  <esyr> hmmm.
 241 2012-05-14T22:48:14  <esyr> ThomasWaldmann: well, if you want to give me shell, i would not argue against it :)
 242 2012-05-14T22:48:44  <esyr> ronny: well, the main question is how different parts should connect with each other.
 243 2012-05-14T22:49:12  <ronny> esyr: thats already defined in moin
 244 2012-05-14T22:49:17  <ronny> it works fine
 245 2012-05-14T22:49:27  <ronny> you stuff things together in wikiconfig
 246 2012-05-14T22:49:58  <esyr> ronny: do you mean moin1.x?
 247 2012-05-14T22:50:49  <ronny> hmm, i mean moin2, there wikiconfig is basically a flask app config
 248 2012-05-14T22:50:57  <ronny> i should use bettr wording
 249 2012-05-14T22:50:59  <ronny> and some sleep
 250 2012-05-14T22:51:02  <ronny> good night
 251 2012-05-14T22:51:48  <esyr> ah.
 252 2012-05-14T22:52:19  <esyr> well, this leads to explicit manual work and that makes me sad panda.
 253 2012-05-14T22:54:02  <ronny> esyr: better than random plugins installed by the server admin breaking your wiki
 254 2012-05-14T22:54:07  <ThomasWaldmann> esyr: query
 255 2012-05-14T22:55:30  *** in3xes has joined #moin-dev
 256 2012-05-14T23:01:12  <esyr> ronny: good point.
 257 2012-05-14T23:01:23  <esyr> hmmm.
 258 2012-05-14T23:18:30  *** dave_largo has quit IRC
 259 2012-05-14T23:20:17  <esyr> hmm, looks like i know in what directions i should move related to concept.
 260 2012-05-14T23:20:26  <esyr> try to averhaul it today or tomorrow.

MoinMoin: MoinMoinChat/Logs/moin-dev/2012-05-14 (last edited 2012-05-13 22:45:04 by IrcLogImporter)