MoinMoin: MoinMoinChat/Logs/moin-dev/2006-06-20

2006-06-20T00:47:42  <ThomasWaldmann> Kepplar: as you might have found meanwhile, the existing MetaDict implementation doesnt write to disk until you call sync()
2006-06-20T00:52:17  <nwp_> ThomasWaldmann: just come across a possible issue with cookies in general...
2006-06-20T00:52:40  <nwp_> python's cookie library throws an exception when you ask it to parse a cookie with "illegal" characters in it
2006-06-20T00:53:16  <nwp_> it's more than possible that there will be cookies out there from apps running on the same servers as moin servers that create such cookies
2006-06-20T00:53:23  <nwp_> specifically, colons in a key
2006-06-20T00:53:42  <nwp_> Apparently WebSVN might be one example of such an app.
2006-06-20T00:54:15  <nwp_> and some perl auth libraries for apache do it by default
2006-06-20T00:54:27  <nwp_> (and in old versions, can only do it that way)
2006-06-20T00:55:30  <mvirkkil> nwp_: Just trap the exception, and hope it's not a vital cookie?
2006-06-20T00:55:34  <nwp_> So when I get a chance, I'll have a look to see whether there's an easy way round it, or whether it's likely to be fixed by the python libs relaxing their checking a little (has been known before in a similar case it seems), or whether we just document it and forget about it
2006-06-20T00:55:42  <nwp_> mvirkkil: it always is a vital cookie
2006-06-20T00:55:51  <nwp_> we only have vital cookies
2006-06-20T00:56:25  <nwp_> what you/I think of as a cookie, python thinks about as a key/value pair within one big cookie
2006-06-20T00:56:50  <nwp_> and if an unrelated key has illegal chars, we can't read our (legal) one
2006-06-20T00:57:31  <nwp_> which is teh suck.
2006-06-20T00:58:14  <nwp_> sorry, I mean teh 5ux0r
2006-06-20T00:58:16  <nwp_> :-P
2006-06-20T01:02:02  <ThomasWaldmann> i never had this case yet, but sounds like trouble
2006-06-20T01:03:39  <ThomasWaldmann> maybe do a fixed python Cookie lib, we can include them under support/ as we do with other broken python libs, too
2006-06-20T02:11:56  <nwp_> it won't be common, and the other apps should be fixed, but if there's an easy way round it, it would make sense to do it. I think there's a bug filed on trac about this, and they were looking at a workaround. I didn't have time to look into how they worked-around though
2006-06-20T11:34:12  <Kepplar> damn, uni machine fell over. lost irc connect. may have missed your answers
2006-06-20T15:30:31  <mvirkkil> please read the "Passthrough blocks" section on http://fedoraproject.org/wiki/MoinDocBookProject/SupportedElements
2006-06-20T15:30:42  <mvirkkil> I'd really like to hear your input regarding this
2006-06-20T15:33:31  <mvirkkil> xorAxAx: --^
2006-06-20T15:33:37  <mvirkkil> ThomasWaldmann: --^
2006-06-20T15:35:00  <mvirkkil> in this case the "db" namespace would be for docbook. Something like "math" could be suitable for mathml.
2006-06-20T15:36:09  <mvirkkil> I want to keeep the hardcoded list of the known namespaces in the parser, even though they could be resolved when the formatter gets them. But this way there is a canonical namespace-mapping.
2006-06-20T15:37:18  <mvirkkil> This would allow for mixing different xml stuff, like embedding an svg image inside a docbook element. The html formatter could just ignore the docbook namespaced stuff, and render the svg instead.
2006-06-20T15:38:42  <mvirkkil> Some fallback mechanism could even be created within the parser for some of the formats, if the formatter doesn't understand them. Like rendering mathml or svg to an image or something, maybe...
2006-06-20T15:39:13  <mvirkkil> That might be a bit complex for the time being, but it would in theory be possible.
2006-06-20T15:48:46  * mvirkkil is cleaning up the idea
2006-06-20T17:58:09  <Kepplar> !seen ThomasWaldmann
2006-06-20T17:59:05  <starshine> Kepplar: no bots in here, but I can scroll up...
2006-06-20T17:59:36  <starshine> ~4pm mytime yesterday in this channel.. so um..
2006-06-20T17:59:48  <starshine> 17h about
2006-06-20T18:03:37  <starshine> gtg
2006-06-20T18:11:44  <Kepplar> thanks
2006-06-20T18:11:50  <Kepplar> i got d/c so i cant scroll up this time :)
2006-06-20T18:58:31  <ThomasWaldmann> moin
2006-06-20T18:59:22  <ThomasWaldmann> mvirkkil: i would prefer to have moin related content on the moin wiki
2006-06-20T19:02:21  <ThomasWaldmann> fpletz: the xapian install worked (but we didnt test too much because of nonrelated stuff)
2006-06-20T19:07:34  <ThomasWaldmann> mvirkkil: do you think it is nice/necessary to put xml fragments into wikied docbook pages?
2006-06-20T21:00:46  <ThomasWaldmann> fpletz: reading?
2006-06-20T21:15:29  <ThomasWaldmann> fpletz: just merge main :)
2006-06-20T21:21:29  <Kepplar> \o/
2006-06-20T21:21:34  <Kepplar> right
2006-06-20T21:21:37  <Kepplar> im partially
2006-06-20T21:21:39  <Kepplar> back
2006-06-20T21:21:49  <Kepplar> have i missed much?
2006-06-20T21:22:17  <Kepplar> kepplar@jaxteran:~$ su
2006-06-20T21:22:17  <Kepplar> Password:
2006-06-20T21:22:59  <ThomasWaldmann> just enter your password :)
2006-06-20T21:23:21  <Kepplar> i cancelled the rest of the paste :P
2006-06-20T21:23:31  <Kepplar> how you been?
2006-06-20T21:23:50  <ThomasWaldmann> working on moin details the whole day
2006-06-20T21:23:57  <Kepplar> details?
2006-06-20T21:24:39  <ThomasWaldmann> ldap, filesystem search stuff, filters, ...
2006-06-20T21:24:51  <ThomasWaldmann> see hg.thinkmo.de
2006-06-20T21:25:01  <Kepplar> cool
2006-06-20T21:29:01  <Kepplar> er
2006-06-20T21:29:07  <Kepplar> ive been getting bogged down in design
2006-06-20T21:29:14  <Kepplar> especially over how to select the engine type
2006-06-20T21:29:23  <Kepplar> debating having a builder class
2006-06-20T21:29:46  <Kepplar> that supplied to the engine manager to generate and register the specific engine by a name
2006-06-20T21:30:04  <Kepplar> that involves another class and you might as well just create the damn object yourself
2006-06-20T21:30:07  <Kepplar> :/
2006-06-20T21:30:17  <Kepplar> any advice?
2006-06-20T21:32:24  <ThomasWaldmann> why dont you just create the storage engine in the cfg until you get something working?
2006-06-20T21:32:40  <Kepplar> hmm ok
2006-06-20T21:32:54  <Kepplar> i just hate working with hacked architecture :P
2006-06-20T21:38:34  <ThomasWaldmann> maybe get something working at all soon
2006-06-20T21:40:11  <Kepplar> thats what im doing
2006-06-20T21:40:24  <Kepplar> becvaus
2006-06-20T21:40:33  <Kepplar> because we decided against the flat interfaces
2006-06-20T21:40:35  <mvirkkil> ThomasWaldmann: Are you referring to the docbook plans, when you say you'd like docbook related stuff to be on the wiki?
2006-06-20T21:40:38  <Kepplar> its thrown my gameplan completly
2006-06-20T21:40:55  <mvirkkil> ThomasWaldmann: The things i've got on the fedora wiki?
2006-06-20T21:42:57  <mvirkkil> ThomasWaldmann: And regarding embedding xml fragments in to the wiki: this is a request I've got a lot. My mentor requested I'd research the subject. My original plan was a horrible kludge compared to this :O)
2006-06-20T21:43:07  <mvirkkil> ThomasWaldmann: This would allow embedding mathml in the wiki syntax.
2006-06-20T21:44:22  <mvirkkil> ThomasWaldmann: but the reason I'm getting requests for it, is because people want to put stuff in the docbook, that there is no wikisyntax for, and that can't be handled nicely with macros.
2006-06-20T21:44:42  <ThomasWaldmann> xml is not quite the wiki way
2006-06-20T21:45:00  <mvirkkil> ThomasWaldmann: I know. But that's the cleanest solution I've found so far :/
2006-06-20T21:45:21  <mvirkkil> ThomasWaldmann: And people seem to be dead set on wanting to embed xml.
2006-06-20T21:45:57  <mvirkkil> ThomasWaldmann: Do you have a better suggestion?
2006-06-20T21:46:42  <mvirkkil> ThomasWaldmann: This would be a cleanish and extendable way, while still allowing stuff to be done "the wiki way" around it.
2006-06-20T21:49:19  <mvirkkil> ascii2doc is _very_ interesting by the way.
2006-06-20T21:50:02  <mvirkkil> written in python. It's an alternative standalone wikisyntax (ascii) -> html/docbook/whatever format converter.
2006-06-20T21:50:19  <mvirkkil> And the syntax seems well documented.
2006-06-20T21:52:48  <mvirkkil> ThomasWaldmann: I could move the pages to the moin wiki if you really want...
2006-06-20T22:09:02  <Kepplar> ...
2006-06-20T22:09:07  <Kepplar> whats wrong with the line
2006-06-20T22:09:26  <Kepplar> str_location = self.generate_rev_location(itrv_revision)?
2006-06-20T22:09:33  <Kepplar> no ? obviously :)
2006-06-20T22:10:59  <ThomasWaldmann> mvirkkil: we have wiki markup because it is easier to enter than html. thus, introducing xml for stuff is nonwiki. of course, if some stuff can only be done with xml, you have a problem. :)
2006-06-20T22:11:37  <Kepplar> i have to agree with ThomasWaldmann, if xml was the way forward, we wouldnt need wiki markup at all
2006-06-20T22:11:57  <Kepplar> tbh i wouldn't say easier, i would just say more idiot proof =)
2006-06-20T22:12:42  <ThomasWaldmann> google made really a mess with that rel nofollow stuff
2006-06-20T22:13:34  <Kepplar> ThomasWaldmann: i hope to have flatfile working and intergrated by the end of the week
2006-06-20T22:13:48  <Kepplar> wireless bridge should be here by thursday or friday so i'll commit and push then
2006-06-20T22:15:16  <mvirkkil> ThomasWaldmann: We are not disagreeing :)
2006-06-20T22:16:17  <Kepplar> this is unbelieable...two functions have the same line, one passes, the other fails
2006-06-20T22:18:17  <ThomasWaldmann> Kepplar: i want daily commits and if you are here, I guess you also have network to commit
2006-06-20T22:18:40  <Kepplar> this machine isnt on the lan
2006-06-20T22:18:49  <Kepplar> er
2006-06-20T22:18:53  <Kepplar> this machine isnt on MY lan
2006-06-20T22:18:58  <Kepplar> this machine is online
2006-06-20T22:19:58  <ThomasWaldmann> and you can commit to your local repo even if the machine isnt online
2006-06-20T22:20:15  <Kepplar> ok?
2006-06-20T22:20:26  <Kepplar> out of interest, whats the point? :)
2006-06-20T22:21:11  <ThomasWaldmann> the point is showing you are working on your task, committing what you have done. code, comments, whatever.
2006-06-20T22:21:58  <Kepplar> that would be shown if i did a push though?
2006-06-20T22:22:44  <ThomasWaldmann> it is published by that
2006-06-20T22:23:23  <ThomasWaldmann> if there is no commit for a week, you could also have been idle, who knows?
2006-06-20T22:24:26  <Kepplar> still dont really get it but ok
2006-06-20T22:24:37  <Kepplar> anyway what shall i do with this mysterious bug?
2006-06-20T22:25:33  <ThomasWaldmann> there is no bug
2006-06-20T22:26:58  <Kepplar> its saying syntax prob
2006-06-20T22:27:27  <ThomasWaldmann> you either have to give verbatim stuff or find yourself
2006-06-20T22:33:25  <mvirkkil> ThomasWaldmann: Is there a risk that some of my stuff wouldn't get accepted in to moin-trunk?
2006-06-20T22:40:27  <ThomasWaldmann> it's hard to tell that before :)
2006-06-20T22:48:14  <Kepplar> hmm
2006-06-20T22:48:27  <Kepplar> eric gave the clue awasy with the collapsable text bit in the wrong place
2006-06-20T22:48:32  <Kepplar> something to do with the docstrings
2006-06-20T22:48:34  <Kepplar> although god knows what
2006-06-20T22:48:39  <Kepplar> fixed though
2006-06-20T22:48:58  <mvirkkil> ThomasWaldmann: I mainly mean the xml-parser stuff? For embedding xml?
2006-06-20T22:50:35  <ThomasWaldmann> i personally don't use docbook, so I maybe won't care much if there can be xml in it or not
2006-06-20T22:52:21  <ThomasWaldmann> but maybe check if key: value isn't good enough for some of that stuff
2006-06-20T22:56:01  <xorAxAx> mvirkkil: as long as you can parse it?
2006-06-20T22:56:07  <xorAxAx> mvirkkil: which integration problem do you see?
2006-06-20T22:56:37  <xorAxAx> mvirkkil: you want to use a parser?
2006-06-20T22:56:37  <Kepplar> hey xorAxAx
2006-06-20T22:56:42  <xorAxAx> hi Kepplar
2006-06-20T23:06:12  <ThomasWaldmann> moin voyager :)
2006-06-20T23:28:18  <mvirkkil> xorAxAx: I'
2006-06-20T23:28:28  <mvirkkil> xorAxAx: I'm not following your questions
2006-06-20T23:29:10  <mvirkkil> xorAxAx: I want to write what is called a "parser" in moin, to handle preformatted blocks of xml.
2006-06-20T23:29:47  <mvirkkil> xorAxAx: In reality the xml parsing inside the block would be done with a real xml parser.