2006-07-07T00:59:28 <Kepplar> prehaps a seperate header box?
2006-07-07T09:04:01 <ThomasWaldmann> moin
2006-07-07T09:05:08 <ThomasWaldmann> Kepplar: this is not yet decided
2006-07-07T09:05:41 <ThomasWaldmann> maybe we begin with merging meta and data before feeding all into editor box (and splitting them again afterwards)
2006-07-07T09:06:03 <ThomasWaldmann> later we could have a separate box for metadata or a form or ...
2006-07-07T10:18:42 <Kepplar> moin
2006-07-07T10:19:03 <Kepplar> yea
2006-07-07T10:19:04 <Kepplar> ok
2006-07-07T10:19:12 <Kepplar> ok
2006-07-07T10:19:15 <Kepplar> assuming that
2006-07-07T10:19:23 <Kepplar> we can make get/set data only data
2006-07-07T10:19:27 <Kepplar> and ignore metadata
2006-07-07T10:19:33 <Kepplar> as its combining out the output anyway
2006-07-07T10:19:41 <Kepplar> and will be saved in the same place for flatfile
2006-07-07T10:19:54 <Kepplar> sure technically that means more operations for the legacy system to work
2006-07-07T10:20:13 <Kepplar> but fits perfetly for the new systems
2006-07-07T10:21:18 <Kepplar> what also is interesting
2006-07-07T10:21:21 <Kepplar> is technically
2006-07-07T10:21:26 <Kepplar> one could load the entire wiki to memory
2006-07-07T10:21:46 <Kepplar> (might have to write a pesudo paging method to make it realistic
2006-07-07T10:21:48 <Kepplar> )
2006-07-07T10:21:55 <Kepplar> and then save it under a different engine
2006-07-07T10:22:02 <Kepplar> ala instant migration
2006-07-07T10:22:02 <Kepplar> hmm
2006-07-07T10:22:10 <Kepplar> especially if the db/svn is remote
2006-07-07T10:24:04 <xorAxAx> whats the advantage compared to doing it item-wise?
2006-07-07T10:24:55 <Kepplar> paging?
2006-07-07T10:25:18 <Kepplar> because id like to see you get 10000000 items in memory on a 64MB 486 DX33
2006-07-07T10:25:21 <Kepplar> :)
2006-07-07T10:27:41 <Kepplar> so
2006-07-07T10:27:58 <Kepplar> metadata
2006-07-07T10:28:18 <Kepplar> form -------| |------ storage
2006-07-07T10:28:29 <Kepplar> data
2006-07-07T10:28:35 <Kepplar> for legacy
2006-07-07T10:28:59 <Kepplar> metadata----------metadata
2006-07-07T10:29:13 <Kepplar> form --------|
2006-07-07T10:29:22 <Kepplar> data-------------data
2006-07-07T10:29:30 <Kepplar> for all other storage implementations
2006-07-07T10:29:33 <Kepplar> and eventually
2006-07-07T10:29:35 <Kepplar> in the future
2006-07-07T10:29:45 <Kepplar> metadata ----------- metadata ---------- metadata
2006-07-07T10:29:47 <Kepplar> well
2006-07-07T10:29:53 <Kepplar> [meta]data is more appreopiete
2006-07-07T11:25:07 <xorAxAx> ThomasWaldmann: has already somebody else arrived?
2006-07-07T12:09:47 <mvirkkil> xorAxAx: Starting the reimplementation of the rest parser?
2006-07-07T12:09:59 <Kepplar> yo
2006-07-07T12:10:00 <mvirkkil> s/rest/rst/
2006-07-07T12:10:57 <xorAxAx> mvirkkil: not yet
2006-07-07T12:13:18 <xorAxAx> mvirkkil: felixw and i scheduled a meeting in the late afternoon
2006-07-07T12:18:20 <mvirkkil> xorAxAx: As long as you call formatter methods only (no rawHTML, please) I'll merge you work in to my branch and make sure my formatter handles it.
2006-07-07T12:20:14 <xorAxAx> mvirkkil: i am not sure if we can guarantee that
2006-07-07T12:20:27 <xorAxAx> i mean, some rawhtml call will always remain
2006-07-07T12:20:37 <mvirkkil> xorAxAx: For what stuff?
2006-07-07T12:20:51 <xorAxAx> for all stuff that is not supported y the sformatter api
2006-07-07T12:21:16 <mvirkkil> xorAxAx: Do you have an example of something that might be common?
2006-07-07T12:22:46 <mvirkkil> xorAxAx: One way you could do, is write some hasattr-checks, and call the formatter methods if the formatter has them, and if it doesn't use the rawhtml as a fallback.
2006-07-07T12:23:06 <mvirkkil> xorAxAx: Similar to what I did with admonitions.
2006-07-07T12:23:34 <mvirkkil> xorAxAx: Because that way it would be easy to implement support for it, if the target (docbook) supports it.
2006-07-07T12:23:59 <xorAxAx> we ant to introduce the admonitions ...
2006-07-07T12:24:15 <mvirkkil> xorAxAx: yeah, I know.
2006-07-07T12:25:06 <mvirkkil> xorAxAx: I was referring to the way I was calling hasattr(formatter,"admonition") to check if the formatter supports it. Now that it will go in to formatterbase those checks can of course be removed.
2006-07-07T12:26:28 <mvirkkil> xorAxAx: So you could write if hasattr(formatter, "foo"): getattr(formatter, "foo"); else: formatter.rawHTML("<foo>")
2006-07-07T12:26:33 <mvirkkil> xorAxAx: or something.
2006-07-07T12:28:20 <xorAxAx> yeah, thats dirty, as i said
2006-07-07T12:28:38 <mvirkkil> xorAxAx: But cleaner than just calling rawHTML
2006-07-07T12:29:36 <mvirkkil> xorAxAx: Because you'd at least have some generic interface, even though not all formatters would support it. rawHTML is the ugliest thing in moinmoin.
2006-07-07T12:30:07 <mvirkkil> xorAxAx: So it really should be used as an absolute last resort, IMNHO.
2006-07-07T12:33:13 <xorAxAx> no, rawHTML is necessary because the formatter api is limited
2006-07-07T12:34:16 <mvirkkil> xorAxAx: But some formatters could be extended. As long as you just check if the formatter supports something, before calling rawHTML it could all "just work".
2006-07-07T12:36:48 <mvirkkil> xorAxAx: And I'm not even sure what wouldn't be supported by the formatter api. Do you have an example?
2006-07-07T12:37:36 <mvirkkil> xorAxAx: Something like <font size=6>HELLO</font> isn't currently supported.
2006-07-07T12:38:31 <mvirkkil> xorAxAx: So you would check if the formatter has a set_font method (which none of them has), and call it, but if it doesn't, then use rawHTML.
2006-07-07T12:38:48 <mvirkkil> or someting.
2006-07-07T12:39:34 <mvirkkil> So when later the latex-formatter gets perfect, it can easily add support for the set_font method.
2006-07-07T12:40:47 <xorAxAx> hmm, this doesnt scale
2006-07-07T12:40:53 <xorAxAx> somehow :)
2006-07-07T12:41:37 <mvirkkil> xorAxAx: Why not?
2006-07-07T12:42:00 <xorAxAx> because every formatter would invent its own orthogonal methods
2006-07-07T12:42:30 <mvirkkil> xorAxAx: You mean every macro and parser would invent methods?
2006-07-07T12:42:36 <xorAxAx> all
2006-07-07T12:42:45 <xorAxAx> macros, parser and of course formatters
2006-07-07T12:42:50 <mvirkkil> xorAxAx: Because unless no one is calling a formatter method, there is no point in a formatters calling them.
2006-07-07T12:42:53 <xorAxAx> and nobody would understand anyone
2006-07-07T12:43:06 <mvirkkil> xorAxAx: Well, nobody understands rawhtml
2006-07-07T12:43:23 <mvirkkil> xorAxAx: And people just crunch out rawhtml and that is a lot worse.
2006-07-07T12:43:47 <xorAxAx> crunch out?
2006-07-07T12:43:52 <mvirkkil> xorAxAx: Spew out.
2006-07-07T12:43:57 <mvirkkil> xorAxAx: Write out.
2006-07-07T12:44:08 <mvirkkil> xorAxAx: call formatter.rawHTML()
2006-07-07T12:51:10 <mvirkkil> xorAxAx: Well, I just hope you don't come across anything that would force you to call it.
2006-07-07T12:51:34 <mvirkkil> xorAxAx: If you notice something that forces you to call rawHTML I'd be interested in hearing about it.
2006-07-07T12:52:07 <mvirkkil> xorAxAx: If you could maybe drop me a line here on irc, if you come across such a case?
2006-07-07T12:52:26 <xorAxAx> mvirkkil: yes
2006-07-07T12:52:59 <mvirkkil> xorAxAx: thanks :)
2006-07-07T14:08:36 <felixw> clear
2006-07-07T14:55:14 <xorAxAx> felixw: it looks like i still need ~ an hour
2006-07-07T15:24:16 <Kepplar> moin
2006-07-07T15:27:51 <felixw> xorAxAx: OK.
2006-07-07T15:52:25 <xorAxAx> felixw: or maybe another hour :) we have some problems here
2006-07-07T15:54:32 <felixw> xorAxAx: OK. :-)
2006-07-07T15:55:13 <felixw> join #europython
2006-07-07T15:55:19 <felixw> sorry ;-)
2006-07-07T16:20:59 <newz2000> I've heard 1.6 is going to be a cool upgrade, any guess when it'll be ready for release?
2006-07-07T16:25:13 <Kepplar> september id imagine
2006-07-07T16:34:08 <xorAxAx> newz2000: thats a #moin question
2006-07-07T16:34:40 <newz2000> xorAxAx: ok
2006-07-07T16:34:54 <newz2000> Kepplar: thanks
2006-07-07T16:49:17 <xorAxAx> felixw: problem still not fixed :-/
2006-07-07T17:09:15 <xorAxAx> felixw: can we delay it for tomorrow?
2006-07-07T17:09:22 <xorAxAx> felixw: or for how long will you be here?
2006-07-07T17:17:25 <felixw> xorAxAx: Not sure yet. Maybe I wanna go back tomorrow, but I'll check for trains, OK?
2006-07-07T17:17:36 <felixw> xorAxAx: I'll tell you soon.
2006-07-07T17:18:33 <xorAxAx> hmm, looks like i might be free at 18:00, finally
2006-07-07T17:22:47 <felixw> xorAxAx: Ok, cool. I'd like to talk a bit about how to do the text processing. MoinMoin: MoinMoinChat/Logs/moin-dev/2006-07-07 (last edited 2007-10-29 19:07:54 by localhost)
