2008-07-28T00:22:00 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4102:e7b402f52d0d 1.8-dom-bblank/MoinMoin/macro2/ (PageSize.py _base.py): macro2._base: new base class for lists with numbers+pagelinks
2008-07-28T00:22:01 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4103:a4d9890bfebb 1.8-dom-bblank/MoinMoin/ (macro2/PageHits.py macro/PageHits.py): Migrate PageHits macro to new interface
2008-07-28T00:23:50 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4104:767f999a7147 1.8-dom-bblank/MoinMoin/ (macro/__init__.py macro2/PageHits.py macro2/PageSize.py): macro/macro2: cleanup, remove unused code
2008-07-28T08:15:40 <ThomasWaldmann> moin
2008-07-28T09:23:37 * johill stops over in FRA and will go to HAJ soon
2008-07-28T09:25:31 <johill> hopefully
2008-07-28T09:27:43 <ThomasWaldmann> wb johill :)
2008-07-28T09:45:16 <dreimark> moin
2008-07-28T09:45:32 <dreimark> hi johill
2008-07-28T09:48:46 <xorAxAx> johill: how are you online?
2008-07-28T09:48:55 <xorAxAx> dns tunneling doesnt work for me anymore in german locations
2008-07-28T10:02:36 <TheSheep> moin
2008-07-28T10:36:04 <dreimark> byeongweon: it would be nice if I could later edit the UnderConstruction text using the gui editor
2008-07-28T10:36:49 <dreimark> I tried it and added therefore the two new bug reports
2008-07-28T10:37:29 <byeongweon> dreimark: ok.
2008-07-28T10:38:51 <dreimark> if you want a copy I've added a package to my homepage at MM:ReimarBauer
2008-07-28T10:39:05 <dreimark> attachment:UnderConstruction.zip
2008-07-28T10:40:38 <dreimark> ThomasWaldmann: test {{attachment:abc.png||height="484",width="484",align=top}}
2008-07-28T10:41:02 * dreimark was wondering what the meaning of top is because of the result
2008-07-28T10:41:30 <dreimark> the result is that the word test is on top of the image
2008-07-28T10:43:47 <byeongweon> dreimark: http://moinmo.in/ReimarBauer/UnderConstruction is tihs page what you want to tell me?
2008-07-28T10:56:29 <PawelPacana> moin
2008-07-28T11:16:52 <TheSheep> hi PawelPacana
2008-07-28T11:27:37 <zenhase> moin
2008-07-28T12:09:44 <CIA-53> zenhase <zh@punyco.de> default * 4024:c2ee4633b9e8 1.8-wsgi-fkrupicka/ (67 files in 32 dirs): Merged with 1.8
2008-07-28T12:10:33 <zenhase> \o/
2008-07-28T12:11:44 <zenhase> uhm ... wrong userstring :o
2008-07-28T12:11:59 <zenhase> why does cloning not pull in the other .hgrc? :(
2008-07-28T12:12:57 <ThomasWaldmann> put it into your homedir
2008-07-28T12:13:39 <zenhase> this is from the one in my homedir
2008-07-28T12:13:47 <zenhase> which is not the same i want for moin :o
2008-07-28T12:19:41 <ThomasWaldmann> zenhase: btw, if you clone one of my repos, that doesn't mean you want to commit using my identity :)
2008-07-28T12:20:29 <zenhase> hmm, that is right too :)
2008-07-28T12:21:48 <zenhase> software has to be 'do what i mean' not 'do what i say' ;)
2008-07-28T12:30:26 <TheSheep> zenhase: I prefer 'do what I say' as long as I also have enough info to make the decissions
2008-07-28T12:32:15 <ThomasWaldmann> xorAxAx: dreimark: maybe it is a good idea that moon first concentrates on the javascript code related bugs.
2008-07-28T12:32:53 <moon> ThomasWaldmann: should I ?
2008-07-28T12:33:03 <moon> ThomasWaldmann: ok. I will do that.
2008-07-28T12:33:14 <ThomasWaldmann> (we can fix the one or other python code related bug ourselves. easier than when javascript problems are left over at the end.)
2008-07-28T12:33:27 <ThomasWaldmann> moon: ah, sorry, I thought you left.
2008-07-28T12:34:12 <moon> ThomasWaldmann: don't mention it. anyway I will do that.
2008-07-28T12:35:02 <moon> ThomasWaldmann: to tell the truth, some problem so deeply related with other moinmoin code not just converter
2008-07-28T12:35:08 <ThomasWaldmann> if there are no js code related bugs left, you can of course also fix python code bugs. :)
2008-07-28T12:35:23 <moon> xorAxAx: ok. thanks :)
2008-07-28T12:37:14 <ThomasWaldmann> moon: did you see what waldi is currenlty doing?
2008-07-28T12:37:19 <moon> ThomasWaldmann: upper msg for you.
2008-07-28T12:37:37 <moon> ThomasWaldmann: no I didn't.
2008-07-28T12:37:44 <moon> ThomasWaldmann: what's that?
2008-07-28T12:38:06 <ThomasWaldmann> he's changing how wiki parser and formatter work together
2008-07-28T12:38:31 <moon> ThomasWaldmann: oh, that's cool!
2008-07-28T12:38:42 <ThomasWaldmann> old way is that the parser parses line by line and immediately produces some output using the formatter (e.g. html)
2008-07-28T12:39:51 <ThomasWaldmann> new way is parser creates a complete DOM tree in memory (converter wiki markup > domtree) and then gives the tree to a domtree > html converter.
2008-07-28T12:40:24 <moon> ThomasWaldmann: oh. it seems like more general idea.
2008-07-28T12:40:59 <ThomasWaldmann> the current code processing what the GUI editor POSTs is also a converter (but it tries to convert (x)html > wiki)
2008-07-28T12:41:50 <ThomasWaldmann> some new dom style converter (doesn't exist yet), could go xhtml > domtree, domtree > wiki
2008-07-28T12:42:22 <ThomasWaldmann> or domtree > whatever
2008-07-28T12:43:25 <moon> ThomasWaldmann: sounds good :)
2008-07-28T12:43:42 <ThomasWaldmann> (and also the stuff for feeding the gui editor could be changed to be more flexible and work based on dom trees)
2008-07-28T12:43:43 <moon> ThomasWaldmann: but implementation seem like little bit hard.
2008-07-28T12:44:21 <ThomasWaldmann> it's lots of work, but better testable and more powerful than the old stuff
2008-07-28T12:45:09 <moon> ThomasWaldmann: I agree. if that stuff work well. it will be wonderful.
2008-07-28T12:49:45 <ThomasWaldmann> (but waldi's stuff is rather for moin 2.0, while moon's stuff should be in moin 1.8)
2008-07-28T12:50:04 <moon> ThomasWaldmann: hehe. ok.
2008-07-28T12:52:19 <ThomasWaldmann> moin 1.8 could go alpha/beta after SOC and just include the pieces that are rather finished and easy mergable.
2008-07-28T13:01:07 <xorAxAx> yeah
2008-07-28T13:07:22 <dreimark> ThomasWaldmann: ok
2008-07-28T13:09:54 <dennda> Is there an overview of the usual release cycles of moin?
2008-07-28T13:11:33 <ThomasWaldmann> http://moinmo.in/MoinMoinTodo/Release_1.8
2008-07-28T13:11:54 <ThomasWaldmann> dennda: "when it is ready" :P
2008-07-28T13:12:40 <dennda> debianish, ic
2008-07-28T13:12:54 <dennda> Makes most sense imho
2008-07-28T13:12:56 <ThomasWaldmann> no, we are faster :)
2008-07-28T13:13:02 <dennda> heh
2008-07-28T13:13:41 <xorAxAx> well, time based releases with good release engineering have their advantages as well
2008-07-28T13:13:45 <xorAxAx> and its certainly more agile
2008-07-28T13:14:11 <dennda> that's true, but if your product still has many faults you shouldn't push it too much, imho
2008-07-28T13:14:14 <xorAxAx> on the other hand, llvm does it like that and pypy found a failing assert in every release of llvm
2008-07-28T13:14:21 <dennda> as seen lately with hardy heron, if you ask me
2008-07-28T13:14:45 <xorAxAx> python does time based releases and gets i right
2008-07-28T13:14:50 <dennda> yes
2008-07-28T13:14:54 <xorAxAx> s/i r/it r/
2008-07-28T13:15:08 <ThomasWaldmann> dennda: 2 hardy upgrades hit the wall, 1 still to fix
2008-07-28T13:15:41 <dennda> Well, I think they rushed it a bit
2008-07-28T13:16:12 <dennda> At least that's what I thought when I started it back then. Can't remember the exact bugs I noticed
2008-07-28T13:16:34 <ThomasWaldmann> the strange thing is that the upgrade shortly after the release worked without problems, but currently there is some locale (gzip?) related nasty problem.
2008-07-28T13:20:06 <dennda> hm
2008-07-28T13:20:22 <dennda> I usually don't upgrade my systems
2008-07-28T13:24:08 <xorAxAx> real men use linux 0.9
2008-07-28T13:25:08 * xorAxAx did an upgrade to hardly on saturday. unfortunately, the machine is using a damn ralink rt73usb wifi stick whose driver had been broken for a lot of years. so the kernel is still .15 or .20
2008-07-28T13:26:09 <xorAxAx> (in the hardly kernel, the dev is being recognized but with ehci it stalls the system on io and without it limits the bandwidth at 50 kB/s)
2008-07-28T13:26:45 <dennda> it has hardy, not hardly :)
2008-07-28T13:27:10 <dennda> s/has/is
2008-07-28T13:27:32 <ThomasWaldmann> hardly upgradeable :P
2008-07-28T13:27:37 <dennda> yes
2008-07-28T13:27:59 <dennda> I knew there'd be a good pun on that, but I didn't know which ;)
2008-07-28T13:28:11 <xorAxAx> sure it is hardly
2008-07-28T13:28:28 * xorAxAx uses agrep to find another nick name for intrepid
2008-07-28T13:29:10 <dennda> petrined?
2008-07-28T13:29:19 <dennda> ah doesn't work
2008-07-28T13:29:26 <xorAxAx> intrigue
2008-07-28T13:29:55 <xorAxAx> hmm, it has too many positive meanings
2008-07-28T13:30:13 <xorAxAx> inbreed
2008-07-28T13:30:21 <xorAxAx> ubuntu inbreed, yeah, not bad :)
2008-07-28T13:30:54 <xorAxAx> insipid, better
2008-07-28T13:33:55 <ThomasWaldmann> moon: for the dialogue boxes displayed (e.g. when adding/editing a link), it would be nice if the cursor was positioned in the first entry field.
2008-07-28T13:40:35 <ThomasWaldmann> zenhase: if i title search for foobar, it tells me:
2008-07-28T13:40:44 <ThomasWaldmann> Please use a more selective search term instead of ""
2008-07-28T13:59:23 <johill> xorAxAx: dns tunnelling sort of worked, but I gave up
2008-07-28T13:59:35 <johill> xorAxAx: just used my cellphone
2008-07-28T13:59:55 <johill> anyway, need lunch now, didn't have time anywhere due to flight delays etc.
2008-07-28T14:04:11 <ThomasWaldmann> dennda: how's debugging going?
2008-07-28T14:05:01 <dennda> ThomasWaldmann: Yesterday night I rewrote the info action. It *should* work, but I can only test after I rewrote the piece of code that actually puts the infos into the revisions in the first place. (Which I am doing right now)
2008-07-28T14:05:12 <dennda> For now you'll get a key error
2008-07-28T14:05:31 <ThomasWaldmann> yes, seen that
2008-07-28T14:05:35 <dennda> which is expected
2008-07-28T14:07:06 <johill> you should catch it though and display nothing just in case somebody creates revisions like that accidentally or something
2008-07-28T14:07:09 <dennda> As far as I can tell, that happens in PageEditor.py, saveText and the former _write_file. (the latter is being discarded)
2008-07-28T14:07:33 <dennda> johill: Hm, that's clearly a misuse of the code then
2008-07-28T14:08:39 <dennda> I wonder, how would one accidentally create revisions like that?
2008-07-28T14:08:53 <dennda> You must use the pageeditor, mustn't you?
2008-07-28T14:09:23 <ThomasWaldmann> or xmlrpc
2008-07-28T14:09:59 <dennda> I don't know that, let me look it up
2008-07-28T14:10:24 <ThomasWaldmann> putPage or so
2008-07-28T14:10:38 <johill> well yeah, you shouldn't really do that, but it's possible quite easily and I'm not sure it should blow up
2008-07-28T14:10:48 <johill> anyway, I don't think I'm thinking straight after travelling for over 18 hours
2008-07-28T14:11:08 <dennda> hehe
2008-07-28T14:11:18 <dennda> poor johill
2008-07-28T14:11:36 <dennda> Ok, I can catch it. Not much of a problem
2008-07-28T14:11:58 <johill> fix the other place first ;)
2008-07-28T14:12:08 <dennda> yes
2008-07-28T14:12:13 <dennda> that's the place with that html form thing
2008-07-28T14:12:19 <johill> ok
2008-07-28T14:12:27 <johill> lunch now, and then we'll see if I can think again
2008-07-28T14:12:32 <dennda> ok
2008-07-28T14:12:34 <dennda> get coffee
2008-07-28T14:12:36 <dennda> :)
2008-07-28T14:21:37 <PawelPacana> someone said that drinking coffee is borrowing energy from the next day ;)
2008-07-28T14:22:15 <dennda> well, drink coffee the next day then, too
2008-07-28T14:49:57 <zenhase> hmm, coding one handed is hard
2008-07-28T14:50:24 <xorAxAx> pair programming :)
2008-07-28T14:50:56 <zenhase> over the last few days pain in my right hand came up
2008-07-28T14:51:20 <zenhase> i might have some light CTS going on :(
2008-07-28T14:56:15 <grzywacz> lol @ pair programming ;)
2008-07-28T14:57:09 <zenhase> i'll get a wrist brace/splint later from a friend
2008-07-28T14:57:26 <zenhase> hope this will speed uphealing
2008-07-28T14:57:31 <grzywacz> Get a proper keyboard, perhaps?
2008-07-28T14:57:41 <zenhase> meanwhile i type lefthanded
2008-07-28T14:58:12 <zenhase> emacs is such a pain without a right hand :O
2008-07-28T14:58:35 <dennda> Can't emacs simulate a right hand for you?
2008-07-28T14:58:37 <zenhase> grzywacz: i don't think it's the keyboard
2008-07-28T14:58:43 <dennda> There's likely a plugin
2008-07-28T14:58:52 <grzywacz> zenhase, I've been having less problems since I bought a proper one
2008-07-28T14:58:57 <xorAxAx> ah, emacs, the bad rsi guy
2008-07-28T14:59:02 <xorAxAx> zenhase: are you using workrave?
2008-07-28T14:59:12 <zenhase> xorAxAx: hmm?
2008-07-28T14:59:13 <grzywacz> oh yes, workrave helps as well
2008-07-28T14:59:17 <grzywacz> zenhase, go install it
2008-07-28T14:59:24 <xorAxAx> zenhase: install it before you die
2008-07-28T14:59:38 * xorAxAx has been using it mainly to relax the eyes a bit
2008-07-28T14:59:46 <zenhase> what is it?
2008-07-28T14:59:48 <xorAxAx> but for general rsi issues it works as well
2008-07-28T14:59:54 <xorAxAx> zenhase: a programm that controls your breaks
2008-07-28T14:59:54 <grzywacz> The eye exercises it proposes are quite good indeed.
2008-07-28T15:00:08 <xorAxAx> zenhase: and gives you exercises
2008-07-28T15:00:40 <zenhase> hmm ok
2008-07-28T15:00:51 <zenhase> hmm cannot use it
2008-07-28T15:00:54 <xorAxAx> one important step to do is to actually configure the break intervas
2008-07-28T15:01:00 <zenhase> linux or windows
2008-07-28T15:01:03 <dennda> both
2008-07-28T15:01:04 <xorAxAx> the defaults dont fit to everybody
2008-07-28T15:01:07 <grzywacz> both
2008-07-28T15:01:09 <zenhase> i work on osx
2008-07-28T15:01:09 <xorAxAx> dennda: neither for him
2008-07-28T15:01:21 <dennda> ah
2008-07-28T15:01:29 <dennda> Then that weird keyboard layout likely is the cause :)
2008-07-28T15:01:37 <zenhase> no weird layout here
2008-07-28T15:01:38 <xorAxAx> zenhase: i guess then you will find a complementary osx version somewhere for only 200 USD (shareware)
2008-07-28T15:01:43 <zenhase> i have good ol US
2008-07-28T15:01:44 <grzywacz> xorAxAx, do you do microbreaks? Or just the every-hours ones?
2008-07-28T15:01:57 <xorAxAx> grzywacz: micro every 10 min., large break every 60 minutes
2008-07-28T15:02:04 <grzywacz> xorAxAx, ok.
2008-07-28T15:02:12 <xorAxAx> (4 min. is the large break)
2008-07-28T15:02:14 <grzywacz> xorAxAx, I found micro ones a little bit too disturbing...
2008-07-28T15:02:40 <xorAxAx> well, focusing your eyes at the screen for 10 min. is already too much
2008-07-28T15:02:45 <grzywacz> zenhase, http://www.virtual-hideout.net/reviews/microsoft_4000_kb/image26.jpg
2008-07-28T15:02:46 * dennda tests workrave, just for fun
2008-07-28T15:02:59 <xorAxAx> grzywacz: and if they are short enough (15 secs or something like that in my case) its nice
2008-07-28T15:03:03 <grzywacz> xorAxAx, maybe I'll give it a try again, since I've already got used to the long ones ;)
2008-07-28T15:03:16 <dennda> You really use that?
2008-07-28T15:03:49 <xorAxAx> yes, its adding up its ~ 5 digits to the 25 digits in my gnome toolbar
2008-07-28T15:03:51 <zenhase> grzywacz: i hate those bent keyboards
2008-07-28T15:03:54 <grzywacz> dennda, sure
2008-07-28T15:04:01 <grzywacz> zenhase, well, bad for you. They do wonders.
2008-07-28T15:04:15 * xorAxAx dislikes them as well
2008-07-28T15:04:36 <grzywacz> Bad for you as well. ;-)
2008-07-28T15:04:58 * xorAxAx never had problems but indeed i am dreaming every night about the good ol PC AT keyboards
2008-07-28T15:05:02 <zenhase> but they are not like a typical laptop layout
2008-07-28T15:05:12 <zenhase> so i can't use those bent ones
2008-07-28T15:05:19 <zenhase> i am also mostly a laptop worker
2008-07-28T15:05:39 <zenhase> so i have no gain from buying a new keyboard
2008-07-28T15:05:43 <xorAxAx> thats actually a large issue for some people
2008-07-28T15:05:52 <xorAxAx> the rsi resulting from too much laptop work
2008-07-28T15:06:10 <grzywacz> Laptop keyboard have always seemed the worst offenders to me...
2008-07-28T15:06:13 <zenhase> what is the difference between rsi and cts btw?
2008-07-28T15:06:15 <grzywacz> *keyboards
2008-07-28T15:06:30 <xorAxAx> grzywacz: its also the general posture
2008-07-28T15:06:35 <dennda> Funny software. I'll give it a try
2008-07-28T15:06:42 <zenhase> i think more pauses and alsp a better posture will already help a lot
2008-07-28T15:07:30 <grzywacz> Anyway, for the bent ones - it's a matter of getting your hands work *straight* instead of the weird stuff people do when using typical keyboards. That + typing without looking = major win. I have no problems with my wrists since I've switched. ;)
2008-07-28T15:08:24 <grzywacz> This helps as well: http://en.wikipedia.org/wiki/Table_football
2008-07-28T15:08:25 <grzywacz> ;-)
2008-07-28T15:08:33 <zenhase> funny thing is ... now that my right hand is kind of not fully functional i have to look at the keyboard again to find keys
2008-07-28T15:08:54 <zenhase> left hand doesn't know about right keyboard side at all
2008-07-28T15:09:27 <zenhase> bbl, getting some splint for my wrsist
2008-07-28T15:15:56 <ThomasWaldmann> grzywacz: http://moinmo.in/MoinMoinBugs/UserpreferencesKeyerrorJid can you have a look?
2008-07-28T15:17:39 <grzywacz> ThomasWaldmann, will do, but not immediately. I'm officially at work now. ;)
2008-07-28T15:21:41 <ThomasWaldmann> ah ;)
2008-07-28T15:47:43 <PawelPacana> working hard or hardly working ;)
2008-07-28T15:48:13 <dennda> yay
2008-07-28T16:01:42 <dennda> puh...
2008-07-28T16:02:38 <xorAxAx> and, do you like the pauses? :)
2008-07-28T16:03:12 <dennda> I think I didn't configure it correctly
2008-07-28T16:03:24 <xorAxAx> probably
2008-07-28T16:03:33 <dennda> I now got a panel applet and it counts the time, but it didn't do anything special yet
2008-07-28T16:03:48 <dennda> Maybe I need to read the Documentation or something
2008-07-28T16:03:49 <xorAxAx> well, it starts doing something special when the counters are at 0:00
2008-07-28T16:04:06 <xorAxAx> you are not active enough to make it trigger
2008-07-28T16:08:30 <CIA-53> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com default * 4374:fd192e91e361 1.8-storage-cdenter/MoinMoin/ (Page.py PageEditor.py action/info.py):
2008-07-28T16:08:30 <CIA-53> storage: Page-saving and info action.
2008-07-28T16:08:30 <CIA-53> * Rewrote Page.size() to work correctly
2008-07-28T16:08:30 <CIA-53> * Re-activating PageEditor._write_file() for now
2008-07-28T16:08:30 <CIA-53> * Adjusting PageEditor.saveText() accordingly (Still need to get revno from html-form
2008-07-28T16:08:34 <CIA-53> * Fix bugs / errors
2008-07-28T16:08:36 <CIA-53> * Displaying the Version-Log of a Page works now. (Diff-Action still needs to be rewritten.)
2008-07-28T16:09:41 <dennda> It is dependant on how much you type? What if I am reading a book or something?
2008-07-28T16:09:52 <dennda> Oh, that just triggered a micro pause ;)
2008-07-28T16:10:32 <xorAxAx> i think its not dependant by default
2008-07-28T16:10:55 <xorAxAx> but its counting your strokes
2008-07-28T16:11:13 <xorAxAx> and the journey length of your mouse pointer
2008-07-28T16:11:35 <xorAxAx> and how many breaks you do by yourself without your advisor "workrave"
2008-07-28T16:11:41 <dennda> ic, could need a webcam-eyetracking plugin
2008-07-28T16:12:31 <xorAxAx> refa 2.0
2008-07-28T16:13:10 <dennda> refa?
2008-07-28T16:14:09 <ThomasWaldmann> fuer interessierte hier: http://wiki.python.de/User_Group_Stuttgart
2008-07-28T16:15:09 <dennda> StuPID? :D
2008-07-28T16:16:06 <xorAxAx> dennda: http://de.wikipedia.org/wiki/REFA-Methodenlehre
2008-07-28T16:16:23 <xorAxAx> dennda: http://de.wikipedia.org/wiki/Zeitstudie
2008-07-28T16:17:11 <dennda> ah
2008-07-28T16:22:11 <ThomasWaldmann> dennda: nein, StuPIG :P
2008-07-28T16:22:26 <dennda> :)
2008-07-28T16:25:50 * dreimark had read also stupid logo :)
2008-07-28T16:28:50 <ThomasWaldmann> hehe
2008-07-28T16:29:01 <waldi> ThomasWaldmann: something is broken, the current code allows preview results to be cached ...
2008-07-28T16:29:18 * ThomasWaldmann tried to find something funny for the name, but I am open for better/funnier names
2008-07-28T16:30:04 <TheSheep> ThomasWaldmann: you can use my old oink logo http://oink.sheep.art.pl/raw/oink.png
2008-07-28T16:30:21 <dennda> TheSheep: !!! great !!!
2008-07-28T16:30:32 <ThomasWaldmann> TheSheep: cool :)
2008-07-28T16:31:09 <waldi> ah, unchanged preview ...
2008-07-28T16:33:06 <ThomasWaldmann> TheSheep: looks better now :)
2008-07-28T16:33:42 <xorAxAx> TheSheep: its not stu specific, any pig could use it :)
2008-07-28T16:34:45 <TheSheep> or this one: http://sheep.art.pl/misc/piggy04.png
2008-07-28T16:35:03 <xorAxAx> :)
2008-07-28T16:35:42 <ThomasWaldmann> heh
2008-07-28T16:37:09 <TheSheep> (I have it as svg too)
2008-07-28T16:38:11 <CIA-53> Pawel Pacana <pawel.pacana@gmail.com> default * 4374:08881cc14c0b 1.8-mercurialbackend-ppacana/MoinMoin/storage/backends/hg.py: (hg) Simplified dict keys mangling using generators.
2008-07-28T16:38:12 <CIA-53> Pawel Pacana <pawel.pacana@gmail.com> default * 4375:7bfab535fa97 1.8-mercurialbackend-ppacana/MoinMoin/storage/backends/hg.py: (hg) Changed misleading name.
2008-07-28T16:38:13 <CIA-53> Pawel Pacana <pawel.pacana@gmail.com> default * 4376:2679409ecfec 1.8-mercurialbackend-ppacana/MoinMoin/storage/_tests/test_backends.py: (test_backends) Added test for listing revisions after rename.
2008-07-28T16:44:19 <CIA-53> Bastian Blank <bblank@thinkmo.de> default * 4105:17d26555d23b 1.8-dom-bblank/MoinMoin/converter2/link.py: Link converter - Split recursor logic in base class, add pagelinks converter
2008-07-28T16:44:19 <CIA-53> Bastian Blank <bblank@thinkmo.de> default * 4106:376278c72402 1.8-dom-bblank/MoinMoin/Page.py: Page - Save pagelinks into cache
2008-07-28T16:45:07 <dennda> PawelPacana: I merged yesterday
2008-07-28T16:45:54 <PawelPacana> dennda: cool
2008-07-28T16:46:18 <PawelPacana> i've seen you have ported info action to work with new backend
2008-07-28T16:47:17 <dennda> Yes
2008-07-28T16:47:26 <dennda> Doing diff-action right now
2008-07-28T16:52:05 <waldi> ThomasWaldmann: if the pagelinks are not cached, should it create the cache or pretend that there exists none?
2008-07-28T16:54:51 <PawelPacana> question about renames: is the upper layer code responsible for committing 'rename' revision with comment and meta stuff after firing item.rename() or should it be done completely on backend layer?
2008-07-28T16:55:25 <xorAxAx> PawelPacana: is there a rename operation?
2008-07-28T16:55:30 <xorAxAx> if so, the upper layers should use it
2008-07-28T16:56:49 <PawelPacana> xorAxAx: there is
2008-07-28T16:58:25 <ThomasWaldmann> waldi: it should create the cache, of course
2008-07-28T16:58:56 <PawelPacana> xorAxAx: this question shoud specifically go to johill, since revision is not increased in new fs backend after rename
2008-07-28T16:58:56 <ThomasWaldmann> (and if there are no pagelinks, there will be a cache with an empty list, not no cache)
2008-07-28T16:59:50 <xorAxAx> PawelPacana: hmm, good question. is there a need for a rev increase from moins pov?
2008-07-28T17:00:01 <xorAxAx> PawelPacana: the name is some kind of virtual item metadata property
2008-07-28T17:00:10 <xorAxAx> so changing metadata of the item shouldnt need a new rev
2008-07-28T17:03:48 <PawelPacana> actually changing metadata should need a new revision -> ACL
2008-07-28T17:04:59 <xorAxAx> PawelPacana: not if you hcange item metadata
2008-07-28T17:05:46 <TheSheep> xorAxAx: so renames are impossible to rollback and don't appear on recent changes?
2008-07-28T17:05:50 <TheSheep> xorAxAx: cool!
2008-07-28T17:06:10 <xorAxAx> hmm
2008-07-28T17:06:17 <PawelPacana> xorAxAx: items with metadata dont have revisions
2008-07-28T17:06:26 <xorAxAx> PawelPacana: huh?
2008-07-28T17:06:32 <xorAxAx> is that written down somewhere?
2008-07-28T17:06:32 <PawelPacana> items with revisions have revision metadata
2008-07-28T17:06:55 <xorAxAx> i mean, it doesnt make sense to change the spec every month via irc and probably end up with 3 different pictures of the whole in 3 or 4 minds :)
2008-07-28T17:07:08 <xorAxAx> i admit that you wouldnt be the one writing but rather reading it
2008-07-28T17:07:26 <xorAxAx> PawelPacana: so, do you know of such a wiki page?
2008-07-28T17:07:46 <johill> PawelPacana: of course items with metadata also have revisions
2008-07-28T17:08:11 <johill> ACL isn't item metadata tho but revision metadat
2008-07-28T17:08:24 <dennda> johill: welcome back :)
2008-07-28T17:08:57 <PawelPacana> johill: i agree ACL is revision metadata
2008-07-28T17:08:57 <TheSheep> johill: what about name? is it item metadata or revision metadata?
2008-07-28T17:09:49 <xorAxAx> TheSheep: indeed, that needs to be resolved. we need denotional semantics for the api
2008-07-28T17:10:16 <xorAxAx> who writes a few haskell monads and ports them into a coq program? :)
2008-07-28T17:10:35 * johill tunes out
2008-07-28T17:11:05 <xorAxAx> johill: :-)
2008-07-28T17:11:06 <PawelPacana> johill: so basicaly i can do item.changes_metadata(); item.publish... and then item.create_revision() ?
2008-07-28T17:11:23 <johill> what for?
2008-07-28T17:11:32 <xorAxAx> johill: well, the queestion is what a rename() changes
2008-07-28T17:11:32 <johill> can you start with what you're trying to do maybe?
2008-07-28T17:11:44 <johill> what does rename have to do with that??
2008-07-28T17:12:07 <xorAxAx> johill: the following choices are available: a) item metadata b) rev metadata c) the weather d) a new revision is created
2008-07-28T17:12:28 <johill> eh?
2008-07-28T17:12:30 <xorAxAx> johill: i think he is thinking about how the rename op works
2008-07-28T17:12:33 <johill> neither?
2008-07-28T17:12:36 <johill> it just renames the thing?
2008-07-28T17:12:41 <johill> the name isn't part of metadata
2008-07-28T17:12:54 <xorAxAx> ok
2008-07-28T17:13:00 <xorAxAx> an item consists of a name, item metadata and revisions
2008-07-28T17:13:01 <johill> and creating a new rev is done by the upper layer
2008-07-28T17:13:11 <xorAxAx> why a new rev?
2008-07-28T17:13:13 <johill> where the name is the handle to use it from
2008-07-28T17:13:15 <xorAxAx> is that necessary?
2008-07-28T17:13:20 <johill> well because it adds '## renamedf from foo'
2008-07-28T17:13:29 <johill> no
2008-07-28T17:13:31 <xorAxAx> yeah, but thats just a specific detail of the wiki
2008-07-28T17:13:35 <johill> exactly
2008-07-28T17:13:49 <xorAxAx> thats completly irrelevant when your name is pawel or dennda
2008-07-28T17:14:00 <xorAxAx> (ok, less in the latter case probably :))
2008-07-28T17:14:15 <johill> yup
2008-07-28T17:14:18 <xorAxAx> PawelPacana: is that doable for the hg backend?
2008-07-28T17:14:19 <TheSheep> johill: so it changes the data of the page every time too?
2008-07-28T17:14:26 <TheSheep> johill: what about images and such?
2008-07-28T17:14:29 <johill> no, rename doesn't chanke fukc
2008-07-28T17:14:35 <johill> god I can't type
2008-07-28T17:14:43 <johill> what's so hard about rename?
2008-07-28T17:14:47 <johill> just rename the damn thing
2008-07-28T17:14:58 <xorAxAx> a wiki state consists of a function name -> item
2008-07-28T17:15:12 <TheSheep> johill: how does it affect the page history -- that's the hard part
2008-07-28T17:15:12 <johill> yeah
2008-07-28T17:15:21 <xorAxAx> an item consists of item metadata \times (revno -> revision)
2008-07-28T17:15:25 <PawelPacana> xorAxAx: that would involve some internal revisions ommitting
2008-07-28T17:15:30 <johill> and to support consisten state snapshots, you need to have an id internally
2008-07-28T17:15:38 <xorAxAx> a revision consists of rev metadata \times data
2008-07-28T17:15:53 <dennda> Look at the MemoryBackend. There is a name <-> id mapping
2008-07-28T17:15:53 <johill> TheSheep: eh, history is part of the item, not the name of it
2008-07-28T17:16:10 <johill> it's unimportant tho
2008-07-28T17:16:18 <johill> it's a detail to implement consistent wiki state snapshots
2008-07-28T17:16:25 <TheSheep> johill: so anything happens to the history is correct?
2008-07-28T17:16:35 <johill> ???
2008-07-28T17:16:41 <TheSheep> johill: in particular, the rename may be logged or not?
2008-07-28T17:17:04 <xorAxAx> TheSheep: the state of a wiki is not revisioned
2008-07-28T17:17:11 <xorAxAx> TheSheep: only every item is
2008-07-28T17:17:13 <johill> exactly
2008-07-28T17:17:24 <johill> it's not logged until the upper layer adds the '## rnaemed from foo' revision
2008-07-28T17:17:26 <xorAxAx> which is a conceptual issue for wikisync for example
2008-07-28T17:17:49 <TheSheep> johill: and it doesn't add it for all pages that are not of type wiki
2008-07-28T17:18:10 <TheSheep> johill: adding it to images would be pretty silly
2008-07-28T17:18:16 <johill> I gues that would be, indeed
2008-07-28T17:18:33 <xorAxAx> how about an exif comment? :)
2008-07-28T17:18:53 <xorAxAx> .oO(just joking
2008-07-28T17:18:54 <TheSheep> so, basically, you can renmae pages in the wiki amd it doesn't appear on recent changes
2008-07-28T17:19:06 <TheSheep> you can vadalize the whole wiki without any form of peer review
2008-07-28T17:19:14 <xorAxAx> well, how does the spec define the RC?
2008-07-28T17:19:32 <TheSheep> xorAxAx: fuck the spec, you want it to be usable, not correct
2008-07-28T17:20:08 * TheSheep goes to drink a glass of water
2008-07-28T17:20:15 <xorAxAx> TheSheep: well, but we need to agree on a set of things :)
2008-07-28T17:20:23 <xorAxAx> spec in the sense of a written down wiki page
2008-07-28T17:20:32 <johill> I'm not going to think about it now, I don't really care
2008-07-28T17:20:50 <xorAxAx> otherwise backend implementors like pawel will still wonder what a rename is
2008-07-28T17:20:54 <johill> you can go ahead and fix it, or maybe require creating a new rev with appropriate log metadata
2008-07-28T17:21:05 <xorAxAx> the iwki page says "a rename changes the name of an item"
2008-07-28T17:21:18 <johill> I'm in a bad mood right now anyway so it's probably not good to discuss that with me
2008-07-28T17:21:27 <xorAxAx> johill: jetlag?
2008-07-28T17:21:45 <xorAxAx> TheSheep: also the issue that you brought up is more visible if you have a page that describes the interface that people agreed upon
2008-07-28T17:21:59 <johill> I only arrived at 2, and we had a spanish kid gropu in front of us on the plane, little sleep
2008-07-28T17:22:09 <xorAxAx> johill: hmm, when did you depart?
2008-07-28T17:22:22 <johill> boarding was at 4:45 or so
2008-07-28T17:22:28 <johill> pm
2008-07-28T17:22:29 <xorAxAx> johill: the best rule is to use the target tz sleep patterns
2008-07-28T17:22:34 <johill> I tried
2008-07-28T17:22:36 <xorAxAx> hmm, then you should have slepts
2008-07-28T17:22:39 <johill> but that meant sleep on plane
2008-07-28T17:22:40 <xorAxAx> s/.$//
2008-07-28T17:22:42 <xorAxAx> :-)
2008-07-28T17:22:50 <xorAxAx> indeed
2008-07-28T17:22:58 <johill> so now I'm shitty and waiting for it to be late enough to go to bed
2008-07-28T17:23:27 <ThomasWaldmann> better dont sleep on plane, you never know whether something explodes or some drunken maids open the door inflight
2008-07-28T17:24:07 <dennda> In either case it's probably better to be asleep, don't you think? :)
2008-07-28T17:25:07 <ThomasWaldmann> |-)
2008-07-28T17:26:10 <johill> anyway, you guys go ahead and design how to do rename logging cleanly, I just don't see a way without putting the design upside down totally
2008-07-28T17:26:34 <johill> [by using an item name that never changes, and using name in metadata, and indexing, like in user storage]
2008-07-28T17:27:42 <xorAxAx> how did lanius implement it?
2008-07-28T17:28:03 <johill> the same way we do now, I'd guess, or some hack
2008-07-28T17:28:15 <xorAxAx> well, one could just require the backend to add a "rename revision"
2008-07-28T17:28:17 <johill> but he hadn't gotten rid of logfiles completely
2008-07-28T17:28:19 <xorAxAx> that doesnt change the data
2008-07-28T17:28:31 <johill> you mean frontend
2008-07-28T17:28:35 <xorAxAx> no, backend
2008-07-28T17:28:49 <johill> that would be kinda sad
2008-07-28T17:28:52 <TheSheep> just handle it the same as metadata changes
2008-07-28T17:28:57 <johill> since you couldn't change other metadata at the same time
2008-07-28T17:29:04 <xorAxAx> the storage api would require every backend to change the name of an item and add a specific revision with rev metadata "rename from x t oy"
2008-07-28T17:29:06 <johill> TheSheep: that's turning it upside down as I said above, it's _not_ metadata
2008-07-28T17:29:31 <johill> anyway, I'm not going to bother thinking about it now
2008-07-28T17:29:38 <TheSheep> johill: let me see, one revision of an item can have one name, and other rvision -- other name, sounds like data attached to revision, e.g. revision metadata
2008-07-28T17:29:43 <johill> you guys write a wiki page about it or something once you understand what you're doing
2008-07-28T17:29:49 <johill> clearly, at this point you do not
2008-07-28T17:29:49 <xorAxAx> its a bit of abstraction inversion (an item doesnt need to know its name) but better than nothing probably
2008-07-28T17:30:18 <xorAxAx> johill: well, you are the mentor of the guy who is doing the spec things and you changed the spec as well. is there a spec page? :)
2008-07-28T17:30:44 <xorAxAx> TheSheep: yeah, but then its not normalised
2008-07-28T17:30:55 <xorAxAx> TheSheep: you would have the name in the global mapping and the rev metadata
2008-07-28T17:31:27 <xorAxAx> TheSheep: in my model, you still have knowledge about your name but no strict denormalization at the whole time
2008-07-28T17:31:44 <xorAxAx> in fact, you only need to know your name when you rename :))
2008-07-28T17:32:05 <xorAxAx> hmm, regarding renames: assuming somebody wants to implement a kind of namespace dispatching backend
2008-07-28T17:32:19 <xorAxAx> that mounts different backends in namespaces. what if somebody does a cross-namespace rename?
2008-07-28T17:32:28 <TheSheep> xorAxAx: I'm not saying to put the name into metadata, I'm just saying that a rename might incerement the revision number without changing the data
2008-07-28T17:32:35 <xorAxAx> hmm, i guess that merging backend would handle that
2008-07-28T17:33:02 <xorAxAx> TheSheep: i cant get that from your sentence
2008-07-28T17:33:29 <TheSheep> xorAxAx: if you want cross-backend renames, then don't do a rename, just save a new page, but then history stays with the old item
2008-07-28T17:34:52 <xorAxAx> TheSheep: why would it stay? you could move history like you do in a conversion use case
2008-07-28T17:34:58 <xorAxAx> i think you would actually like to do so
2008-07-28T17:35:35 <TheSheep> xorAxAx: what if you are renamng to a name that existed before?
2008-07-28T17:35:48 <TheSheep> xorAxAx: should the histories be merged then?
2008-07-28T17:35:49 <waldi> ThomasWaldmann: okay
2008-07-28T17:35:51 <xorAxAx> TheSheep: in my scenario or generally?
2008-07-28T17:36:11 <xorAxAx> currently its not possible to rename to a page name where there is a (deleted or not) item with that name
2008-07-28T17:36:28 <xorAxAx> and you would actually have to rename it first
2008-07-28T17:36:38 <xorAxAx> if merging took place, the backends wouldnt do that nyway
2008-07-28T17:37:07 <TheSheep> xorAxAx: with hg backend, not only the histories of the new page get merged, but the old page also retains its history
2008-07-28T17:37:34 <TheSheep> xorAxAx: I think it's advantegous, but it makes the tests fail
2008-07-28T17:37:36 <xorAxAx> TheSheep: well, thats not the storage api semantics :-P
2008-07-28T17:37:48 <TheSheep> xorAxAx: then what is it?
2008-07-28T17:38:00 <TheSheep> xorAxAx: there is no api to manipulate the storage history
2008-07-28T17:38:00 <xorAxAx> that names dont have revisions, items do
2008-07-28T17:38:07 <xorAxAx> renames change the name of an item
2008-07-28T17:38:16 <xorAxAx> there can only be exactly one item with a distinct name
2008-07-28T17:38:19 <xorAxAx> at the same time
2008-07-28T17:38:47 <TheSheep> xorAxAx: that's all nice if you only live in today
2008-07-28T17:38:53 <TheSheep> xorAxAx: but items have history
2008-07-28T17:38:58 <xorAxAx> TheSheep: is that necessary to manipulate the history?
2008-07-28T17:39:05 <xorAxAx> TheSheep: items dont have history, they have revisions :)
2008-07-28T17:39:14 <xorAxAx> s/is that/is it/
2008-07-28T17:39:50 <TheSheep> xorAxAx: in this naive approach you propose, the only sane way of 'ranming' an item would be deleting it and creating with a new name and same content
2008-07-28T17:40:10 <xorAxAx> TheSheep: approach? i just explain the current state of the spec as far as i know
2008-07-28T17:40:23 <xorAxAx> TheSheep: no, why?
2008-07-28T17:40:32 <TheSheep> xorAxAx: because as soon as you try to take the history with rename, you admit that items have not only revisions, but also histories -- e.g. the revisions are linked with each other
2008-07-28T17:40:38 <xorAxAx> an item is renamed by changing its name and mutating the global mapping of the tate
2008-07-28T17:41:02 <xorAxAx> TheSheep: i dont understand
2008-07-28T17:41:28 <xorAxAx> revisions are indeed rather a list than a revno -> rev mapping (they are continuguous)
2008-07-28T17:41:36 <xorAxAx> but where does the history come into place?
2008-07-28T17:41:46 <xorAxAx> whats wrong with the abovely explained model?
2008-07-28T17:41:59 <xorAxAx> (this time wrong wrt to usefulness :))
2008-07-28T17:42:42 <TheSheep> xorAxAx: forget it, you are right, I'm sorry I even brought it up
2008-07-28T17:42:58 <xorAxAx> no problem. a wiki page would have helped here :)
2008-07-28T17:43:07 <TheSheep> I'm inventing problems and assaluting jetlagged devs :)
2008-07-28T17:43:20 <xorAxAx> well, back to our rc issue
2008-07-28T17:43:34 <xorAxAx> what do you think about my meta-revisions that are used for renames?
2008-07-28T17:44:02 <TheSheep> xorAxAx: I'm sorry, I really can't continue it right now, be back later
2008-07-28T17:44:07 <xorAxAx> hmm, you would need to duplicate the data file for those ...
2008-07-28T17:44:13 <xorAxAx> TheSheep: is it hot in .pl as well? :)
2008-07-28T17:44:28 <xorAxAx> (ok, duplicating data sounds indeed bad from a usefulness point of view :))
2008-07-28T17:47:37 <xorAxAx> PawelPacana: is it hot?
2008-07-28T17:49:56 <PawelPacana> xorAxAx: quite, around 28C now
2008-07-28T17:50:27 <xorAxAx> wow, here its 31° outside and probably 35° in my flat. but this uni building has air conditioning :)
2008-07-28T17:50:50 <waldi> ua, WantedPages on a fresh underlay needs ages
2008-07-28T17:51:48 <xorAxAx> waldi: about 10 -20 min. on the main branch probably
2008-07-28T17:51:57 <waldi> yeah
2008-07-28T17:52:16 <ThomasWaldmann> hmpf, lenny frozen with moin 1.7.0
2008-07-28T17:53:48 * waldi usualy say that frozen means: we want to take an eye on it but if _you_, the maintainer, can explain why this is needed to fix problems we have no problem
2008-07-28T17:54:11 <waldi> ui, my code lists more wanted pages
2008-07-28T17:54:25 <waldi> but below 1% difference
2008-07-28T17:54:31 <xorAxAx> umm
2008-07-28T17:54:38 <ThomasWaldmann> hopefully the maintainer wakes up early enough and packages 1.7.1 at least
2008-07-28T17:54:48 <xorAxAx> that means it either recognizes the links incorrectly or the old parser crashed parsing some pages --> no pagelinks
2008-07-28T17:55:08 <xorAxAx> ThomasWaldmann: is it still jonas?
2008-07-28T17:55:15 <xorAxAx> the weird guy from dk :)
2008-07-28T17:55:28 * xorAxAx met him once
2008-07-28T17:55:42 <ThomasWaldmann> yes
2008-07-28T17:56:07 <waldi> xorAxAx: well, only a complete dump could tell this
2008-07-28T17:56:10 <xorAxAx> PawelPacana: hmm, do you have any idea how the storage interface can be changed to make renames show up on RC?
2008-07-28T17:56:31 <xorAxAx> waldi: well, 3 lines modifications of the wantedpages macro(s)
2008-07-28T17:57:40 <waldi> xorAxAx: yeah
2008-07-28T18:03:21 <PawelPacana> xorAxAx: not yet
2008-07-28T18:05:58 <xorAxAx> hmm, how about having a special item that tracks wiki wide state?
2008-07-28T18:06:13 <xorAxAx> with some kind of constant name
2008-07-28T18:06:38 <xorAxAx> in this item, one could store all metadata nd changes that affect the whole wiki
2008-07-28T18:06:57 <johill> sounds like a bottleneck
2008-07-28T18:07:15 <xorAxAx> why bottleneck?
2008-07-28T18:07:44 <xorAxAx> you need some linear global history store anyway ... the question is just whether it should be implicit or explicit
2008-07-28T18:07:46 <johill> updated all the time when editing, no?
2008-07-28T18:07:57 <xorAxAx> with lanius i invented a list of (page, revno, date) tripels
2008-07-28T18:08:09 <xorAxAx> which is queried by news()
2008-07-28T18:08:44 <xorAxAx> johill: well, it will be in either case for the aforementioned reason
2008-07-28T18:09:50 <johill> news is actually calculated on the fly now
2008-07-28T18:10:07 <johill> oh no there's the news optimisation
2008-07-28T18:10:10 <johill> but it could be
2008-07-28T18:10:21 <johill> of course we can force to put this into news
2008-07-28T18:10:27 <johill> seems a bit odd tho
2008-07-28T18:10:51 <xorAxAx> how should news be computed on the fly? do you want to read all n items to show RC?
2008-07-28T18:10:59 <waldi> ThomasWaldmann: MoinMoin.widged.browser uses request.write ...
2008-07-28T18:11:14 <johill> yeah no we just read the binlog, ignore me, I'm only 30% functioning
2008-07-28T18:11:27 <xorAxAx> which binlog?
2008-07-28T18:11:31 <xorAxAx> waldi: :-)
2008-07-28T18:11:38 <johill> news
2008-07-28T18:12:02 <xorAxAx> johill: on which level is it implemented?
2008-07-28T18:12:09 <johill> backend
2008-07-28T18:12:23 <johill> sql doesn't really need that I guessed
2008-07-28T18:12:30 <johill> since it has all revisions in a single table anyway
2008-07-28T18:13:10 <xorAxAx> well, news items could point to the state item
2008-07-28T18:13:29 <johill> state item?
2008-07-28T18:13:33 <xorAxAx> and the state item only captures those revisions that dont fit anywhere lese
2008-07-28T18:13:43 <xorAxAx> that leads to non-redudant normalised storage
2008-07-28T18:13:51 <xorAxAx> 18:05:56 < xorAxAx> hmm, how about having a special item that tracks wiki wide state?
2008-07-28T18:13:54 <xorAxAx> 18:06:11 < xorAxAx> with some kind of constant name
2008-07-28T18:13:57 <xorAxAx> 18:06:36 < xorAxAx> in this item, one could store all metadata nd changes that affect the whole wiki
2008-07-28T18:14:07 <ThomasWaldmann> waldi: maybe i have time to refactor it later
2008-07-28T18:14:31 <johill> but that means news is slow ;)
2008-07-28T18:14:33 <waldi> xorAxAx: no idea about the WantedPages output
2008-07-28T18:14:37 <xorAxAx> johill: not at all
2008-07-28T18:14:40 <waldi> -HjælpTilRedigering/SubPages: HjælpTilRystemPagesInDutchGroup
2008-07-28T18:14:40 <waldi> +HjælpTilRedigering/SubPages: HjælpTilRedigering
2008-07-28T18:14:42 <johill> if it only has stuff that isn't stored elsewhere
2008-07-28T18:14:43 <xorAxAx> johill: news doesnt change
2008-07-28T18:14:48 <dennda> waldi: Gesundheit
2008-07-28T18:15:02 <xorAxAx> waldi: thats the only differnece?
2008-07-28T18:15:34 <xorAxAx> johill: news stays the same like it is now
2008-07-28T18:15:48 <johill> an optimisation
2008-07-28T18:15:58 <xorAxAx> everybody is optimal
2008-07-28T18:16:02 <xorAxAx> what is not? :)
2008-07-28T18:16:03 <waldi> xorAxAx: no
2008-07-28T18:16:10 <dreimark> ThomasWaldmann: the widget.browser should be configurable for POST too
2008-07-28T18:16:14 <waldi> xorAxAx: this is one change i just have in fron of me
2008-07-28T18:16:32 <xorAxAx> there is no pgae "HjælpTilRedigering/SubPages"
2008-07-28T18:16:32 <dreimark> and the javascript src link optional removable
2008-07-28T18:16:34 <xorAxAx> waldi:
2008-07-28T18:16:37 <xorAxAx> on master17
2008-07-28T18:16:49 <xorAxAx> neither on master16
2008-07-28T18:17:03 <waldi> xorAxAx: this is the name of wanted page
2008-07-28T18:17:19 <xorAxAx> ah
2008-07-28T18:17:31 <waldi> but in my version not even HjælpTilRystemPagesInDutchGroup exists, but the old version sees that page
2008-07-28T18:17:51 <xorAxAx> hmm
2008-07-28T18:18:14 <xorAxAx> old data dir?
2008-07-28T18:18:29 <waldi> symlinked
2008-07-28T18:18:35 <xorAxAx> to where?
2008-07-28T18:18:39 <xorAxAx> well, if that is really the case it looks broken in the old code
2008-07-28T18:19:04 <johill> xorAxAx: anyway, state item seems wrong, if anything it should be new API
2008-07-28T18:19:18 <johill> if only because of atomicity requirements
2008-07-28T18:19:18 * dreimark we have got a storm warning for the next hours
2008-07-28T18:19:45 <johill> but I'm going to head out now to get some food
2008-07-28T18:19:46 <waldi> xorAxAx: between the two branches
2008-07-28T18:20:00 <dennda> dreimark: good luck
2008-07-28T18:20:24 <xorAxAx> johill: we should continue this talk tomorrow when you are fit :)
2008-07-28T18:22:24 * dreimark hopes one summarizes the talk - results - questions afterwards
2008-07-28T18:23:15 <xorAxAx> dreimark: we are basically thinking about the rename op and that RC doesnt show it at all in the new model
2008-07-28T18:25:00 <CIA-53> Bastian Blank <bblank@thinkmo.de> default * 4107:e207fdf5f2df 1.8-dom-bblank/MoinMoin/Page.py: Page - Produce pagelinks during the initial converter step
2008-07-28T18:25:06 <dreimark> in mercurial a rename does not show you the old history of the file.
2008-07-28T18:25:43 <dreimark> if you found out that the file has an older one with an other name you can get the history for that file
2008-07-28T18:26:49 <waldi> hg is not able to produce the complete history of a file?
2008-07-28T18:27:28 <johill> dreimark: can arnica rotate based on exif?
2008-07-28T18:27:35 <dreimark> I found no way to see it if the file is renamed. you have to go to the old version of the old filename and then you get the old history
2008-07-28T18:28:28 <xorAxAx> hg had been storing renames for years
2008-07-28T18:28:32 <xorAxAx> s/had/has/
2008-07-28T18:28:33 <dreimark> johill: normally yes it works with my camera but not with each one. I can read the tag and rotate
2008-07-28T18:28:53 <dreimark> xorAxAx: it stores them but how can you show the old history
2008-07-28T18:29:03 <dreimark> withouit knowing the old name
2008-07-28T18:29:14 <xorAxAx> what is "the old history"?
2008-07-28T18:29:24 <dreimark> the changes before the rename
2008-07-28T18:29:30 <PawelPacana> dreimark: -f
2008-07-28T18:29:34 <johill> now something brok, my theme is gone
2008-07-28T18:29:44 <xorAxAx> PawelPacana: indeed, first option in -h output
2008-07-28T18:29:48 <PawelPacana> --follow follow changeset history, or file history across copies and renames
2008-07-28T18:30:21 <dreimark> hmm, the webinterface doesn't know that
2008-07-28T18:30:32 <PawelPacana> dreimark: that is only limitation of ui
2008-07-28T18:30:46 <dreimark> ok, I know it now too
2008-07-28T18:32:40 <dreimark> xorAxAx: I don't get listed it in my hg -h version
2008-07-28T18:33:14 <xorAxAx> dreimark: hg log -h
2008-07-28T18:33:26 <johill> dreimark: doesn't rotate here
2008-07-28T18:33:35 <TheSheep> re
2008-07-28T18:33:52 <TheSheep> xorAxAx: how about having an empty revision like with page deletion?
2008-07-28T18:33:53 <dreimark> johill: can you send me an example image, I can look into it later today or tom.
2008-07-28T18:34:08 <johill> I'll just fix it
2008-07-28T18:34:16 <dreimark> :)
2008-07-28T18:34:25 <xorAxAx> TheSheep: umm, thats pretty bad because it lets history be noncontiguous in a single item
2008-07-28T18:34:32 <xorAxAx> TheSheep: also its rather denormalised
2008-07-28T18:34:47 <xorAxAx> because the revs dont need to know about renames, an item doesnt even need to know its name storage wise
2008-07-28T18:34:47 <dreimark> did hg know and can list files which were renamed only ?
2008-07-28T18:35:29 <TheSheep> xorAxAx: something like: rename an item, delete the (nonexistant now) old name
2008-07-28T18:35:52 <TheSheep> xorAxAx: which would create an item with only one, empty revision, under the old name
2008-07-28T18:36:04 <TheSheep> xorAxAx: of course, that's littering teh namespace
2008-07-28T18:36:12 <xorAxAx> huh
2008-07-28T18:36:24 * dreimark has to leave want to be at home before it starts to rain
2008-07-28T18:36:38 <dreimark> bbl
2008-07-28T18:36:44 <xorAxAx> TheSheep: thats breaking the semantics because you cant use the old name for a new or renamed page anymore
2008-07-28T18:37:12 <johill> dreimark: where's that code?
2008-07-28T18:37:15 <TheSheep> xorAxAx: yes, unless there is a special case, but that's ugly
2008-07-28T18:37:46 <xorAxAx> TheSheep: i still like my model :)
2008-07-28T18:38:00 <TheSheep> xorAxAx: then how about an 'empty edit' on the item, one that has the smae data and metadata?
2008-07-28T18:38:11 <xorAxAx> i think the special item can even be nameless
2008-07-28T18:38:32 <xorAxAx> TheSheep: you mean without any data? still gives you noncontinuous history
2008-07-28T18:38:52 <TheSheep> xorAxAx: how so? the whole item is moved, like today
2008-07-28T18:39:07 <xorAxAx> not monotonic revs
2008-07-28T18:39:11 <johill> dreimark: I don't see the code, so I guess I cannot fix it, try http://johannes.sipsolutions.net/Photos/Ottawa%202008
2008-07-28T18:39:16 <xorAxAx> revs with data, empty rev, revs with data
2008-07-28T18:39:29 <TheSheep> xorAxAx: you get that when you change ACLs anyways
2008-07-28T18:39:32 <xorAxAx> people/code looking at the empty rev will be confused
2008-07-28T18:39:35 <xorAxAx> TheSheep: ?
2008-07-28T18:39:44 <xorAxAx> acls dont delete your data :)
2008-07-28T18:39:49 <TheSheep> xorAxAx: acls are only metadata, the item data doesn't change
2008-07-28T18:39:52 <xorAxAx> and its still denormalised as i pointed out above
2008-07-28T18:40:31 <TheSheep> "18:39 < xorAxAx> acls dont delete your data :)
2008-07-28T18:40:36 <TheSheep> can you elaborate on that?
2008-07-28T18:40:58 <xorAxAx> well, i wonder why you are talking about acls now
2008-07-28T18:41:04 <xorAxAx> they are unrelated to renames
2008-07-28T18:41:12 <xorAxAx> also they dont stich holes into the revision sequence
2008-07-28T18:41:18 <TheSheep> (by 'empty edit' I mean something you would get by hitting 'edit' in a wiki and then 'save' immediately, if there weren't a check for that)
2008-07-28T18:41:33 <xorAxAx> yes, thats duplicated data as i suggested above. thats ugly because of that
2008-07-28T18:41:42 <xorAxAx> (you were basically repeating my former idea)
2008-07-28T18:41:42 <TheSheep> xorAxAx: why duplicated data?
2008-07-28T18:41:53 <TheSheep> xorAxAx: I'm trying to get it across
2008-07-28T18:41:57 <xorAxAx> because you would copy the data file for every rename which is a bit unexpected
2008-07-28T18:42:09 <TheSheep> xorAxAx: not really
2008-07-28T18:42:13 <xorAxAx> ?
2008-07-28T18:42:21 <TheSheep> xorAxAx: it's teh same as changing alcs or other metadata
2008-07-28T18:42:28 <TheSheep> acls
2008-07-28T18:42:31 * TheSheep clams down
2008-07-28T18:43:09 <TheSheep> xorAxAx: it's the same as only changing revision metadata without touching the data, it has to be supported
2008-07-28T18:43:27 <TheSheep> xorAxAx: PawelPacana has gone to great trouble to support it
2008-07-28T18:44:24 <xorAxAx> ah, indeed
2008-07-28T18:47:01 <xorAxAx> well, then the denormalisation issue remains but its not a big one
2008-07-28T18:47:15 <xorAxAx> and the wiki lacks any place to store global state currently
2008-07-28T18:47:33 <xorAxAx> every wiki instance has a special id ... where would the backends store that?
2008-07-28T18:47:47 <TheSheep> in configuration
2008-07-28T18:48:00 <xorAxAx> no, its automatically generated
2008-07-28T18:48:12 <xorAxAx> and opaque to the user and admin
2008-07-28T18:49:00 * TheSheep doesn't like opaque magical thingies
2008-07-28T18:49:49 <xorAxAx> well, its a GUID
2008-07-28T18:50:00 <xorAxAx> not really magical
2008-07-28T18:50:10 <CIA-53> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com default * 4375:13e13664883f 1.8-storage-cdenter/MoinMoin/action/diff.py: storage: diff-action: Rewrite diff-action to use the storage API directly rather than editlog
2008-07-28T18:50:25 * xorAxAx is gone
2008-07-28T18:50:51 <dennda> Hm there is some bug wrt revision numbering in the upper layers
2008-07-28T18:51:49 <ThomasWaldmann> like?
2008-07-28T18:51:56 <ThomasWaldmann> +1?
2008-07-28T18:52:05 <dennda> dunno for sure yet
2008-07-28T18:52:13 <ThomasWaldmann> find out :)
2008-07-28T18:52:21 <dennda> sure :)
2008-07-28T18:52:35 <dennda> if you want to see it you can try to show the diff between two revisions
2008-07-28T18:52:54 <dennda> (or at least the result of the bug)
2008-07-28T18:53:19 <xorAxAx> hmm, you changed the diff code - so its a bug in the code you just changed?
2008-07-28T18:53:32 <dennda> no
2008-07-28T18:53:38 <xorAxAx> but where else?
2008-07-28T18:54:00 <xorAxAx> ah, indeed
2008-07-28T18:54:06 <xorAxAx> diff uses page probably
2008-07-28T18:54:11 <dennda> it does
2008-07-28T18:54:14 <xorAxAx> and should stop doing so :)
2008-07-28T18:54:24 <dennda> oh
2008-07-28T18:54:41 <xorAxAx> or is there a reason to let Page alive except legacy?
2008-07-28T18:56:27 <johill> don't think so, except maybe that it makes sure there's metadata properly
2008-07-28T18:58:55 <dennda> I can try to throw it out. Let me do that
2008-07-28T19:00:25 <dennda> I hate magic numbers
2008-07-28T19:08:02 <dennda> hm well, at one point Page.url() is used
2008-07-28T19:12:05 <xorAxAx> johill: hmm
2008-07-28T19:12:16 <johill> could be a shim
2008-07-28T19:12:19 <xorAxAx> johill: do you have to copy the metadata manually when creating a new rev?
2008-07-28T19:12:50 <xorAxAx> dennda: well, time for stepwise refactoring :) ask zenhase for details on how to collaborate
2008-07-28T19:12:50 <dennda> shim?
2008-07-28T19:12:57 <dennda> zenhase: ping
2008-07-28T19:13:02 <xorAxAx> dennda: you could e.g. wrap the item into the page just for getting the url for now
2008-07-28T19:13:16 <xorAxAx> and then somebody else can remove this wrapping and replace the url call with something sane
2008-07-28T19:13:30 <dennda> xorAxAx: You mean pageobj.__item = itemobj?
2008-07-28T19:13:32 <xorAxAx> i.e. Page.from_item(item).url() or something like that
2008-07-28T19:13:54 <xorAxAx> and then somebody can grep for from_item later on
2008-07-28T19:14:43 <dennda> Why zenhas_e? I thought his job was to do some wsgi refactoring? /me reads his application
2008-07-28T19:14:58 <xorAxAx> dennda: he has introduced some url class
2008-07-28T19:15:17 <dennda> I see
2008-07-28T19:15:51 <xorAxAx> AFAIK
2008-07-28T19:16:02 <dennda> He will know
2008-07-28T19:20:24 <moon> xorAxAx: did you know why FCKConfig.FormatSource value in moinfckconfig.js set false?
2008-07-28T19:23:05 <xorAxAx> moon: hmm, no, but probably it introduces converter issues. given that you have juts found that - how could it have ever worked in safari? :-)
2008-07-28T19:23:15 <xorAxAx> i would just try setting it
2008-07-28T19:23:33 <dennda> PawelPacana: What do I need to do if I want to try your backend? (I untared underlay.tar, created a folder for data and user and created a wikiconfig_local.py accordingly)
2008-07-28T19:24:09 <zenhase> dennda: pong
2008-07-28T19:24:18 <moon> xorAxAx: it's about html source indentation I think. anyway I think set true is better.
2008-07-28T19:24:34 <zenhase> still haven't got anything for my hand
2008-07-28T19:24:38 <zenhase> :(
2008-07-28T19:24:43 <xorAxAx> moon: ah, if its just whitespace, nobody should care, yes
2008-07-28T19:25:49 <dennda> zenhase: I am rewriting parts of action/diff.py. It uses Page-objects right now. I want to replace them by normal Item objects from the backend. However, Page.url() is used at some point. I need to somehow work around that. xorAxAx said you wrote some URL class?
2008-07-28T19:26:35 <xorAxAx> dennda: he probably hasnt yet done anything regarding page.url but you both work into the same direction
2008-07-28T19:26:35 <zenhase> hmm
2008-07-28T19:26:49 <xorAxAx> just saying that the last step of moving it is not so clear anyway
2008-07-28T19:27:07 <zenhase> dennda: try request.href or request.abs_href if you need host-part too
2008-07-28T19:27:17 <PawelPacana> dennda: you need develompment version of mercurial, you can clone it from http://selenic.com/repo/hg
2008-07-28T19:27:25 <zenhase> dennda: it takes *domainparts and **kwargs for querystring
2008-07-28T19:27:26 <dennda> PawelPacana: urgs, ok
2008-07-28T19:27:44 <zenhase> dennda: that is in my tree though :o
2008-07-28T19:28:14 <zenhase> hmm
2008-07-28T19:28:22 <PawelPacana> then just: 'make local' and later add this directory to your PYTHONPATH
2008-07-28T19:28:35 <xorAxAx> zenhase: well, given that the new item concept needs a new gui, new action concept, new disptaching, new url generation, things need to be done :)
2008-07-28T19:28:44 <xorAxAx> not necessary in the strict context of soc, but sometime
2008-07-28T19:29:20 <dennda> page_url = wikiutil.escape(currentpage.url(request), True) <--
2008-07-28T19:29:30 <zenhase> great
2008-07-28T19:29:36 <TheSheep> dennda: I can make a tarball with it if you like
2008-07-28T19:29:36 <zenhase> dennda: http://werkzeug.pocoo.org/documentation/utils#werkzeug.utils.Href
2008-07-28T19:29:37 <dennda> PawelPacana: Time for virtualenv. I wanted to try that anyways :)
2008-07-28T19:29:48 <zenhase> dennda: this is the object which is on the request-objects
2008-07-28T19:29:57 <johill> http://johannes.sipsolutions.net/Photos/Ottawa%202008
2008-07-28T19:30:12 <johill> dreimark: would appreciate fixing the rotation :)
2008-07-28T19:31:16 <dennda> Why did you take so many balcony-photos?
2008-07-28T19:31:22 <xorAxAx> johill: who is the woman with the bird?
2008-07-28T19:31:30 <xorAxAx> dennda: probably he was drunk at the hotel :)
2008-07-28T19:31:42 <johill> suzie, hpa's wife
2008-07-28T19:31:53 <xorAxAx> whoever hpa is. sounds like an architecture
2008-07-28T19:31:55 <johill> I liked the balconies across me
2008-07-28T19:32:01 <johill> one of the kernel.org admins
2008-07-28T19:32:25 <johill> different styles, lots of junk, lots of plants etc
2008-07-28T19:33:24 <johill> it's not her bird
2008-07-28T19:33:42 <johill> some random guy we met on the bike tour had it
2008-07-28T19:33:52 <johill> (not one of the group)
2008-07-28T19:34:16 <xorAxAx> ah
2008-07-28T19:34:30 <dennda> Didn't you say it was a conference? Looks more like holidays to me :)
2008-07-28T19:35:08 <xorAxAx> well, at the busy parts you are not relaxed enough to take photos usually
2008-07-28T19:35:20 <xorAxAx> but indeed, dont show such photos to your SO, johill
2008-07-28T19:35:24 <johill> conference centre was so gloomy I didn't want to take pictures
2008-07-28T19:35:32 <johill> no light
2008-07-28T19:35:49 <johill> no natural light anyway
2008-07-28T19:36:01 <xorAxAx> except if you dont want to account for any disimbursements in the context of formular N
2008-07-28T19:37:37 <xorAxAx> cf. VI R 94/01
2008-07-28T19:38:54 <xorAxAx> hehe, its about comdex, how boring
2008-07-28T19:40:27 <dennda> zenhase: I didn't get it yet. How do you suggest to remove that .url() occurrence?
2008-07-28T19:40:51 <xorAxAx> dennda: he just said that he has been strating to use that class currently
2008-07-28T19:40:52 <dennda> johill: Was it worth the trip?
2008-07-28T19:41:04 <xorAxAx> dennda: but he didnt do anything in the page realm probably
2008-07-28T19:41:29 <dennda> I assume you still suggest wrapping the thing?
2008-07-28T19:41:53 <dennda> (i.e. hand the work to somebody in the future)
2008-07-28T19:43:02 <xorAxAx> i suggest to literally write the code i wrote
2008-07-28T19:43:13 <xorAxAx> (also adding a method in Page)
2008-07-28T19:43:16 <zenhase> hmm
2008-07-28T19:43:47 <xorAxAx> with a docstring that says "everytime you see this function used, people will want to consider refactoring the caller when there is a new model in the system how stuff will work"
2008-07-28T19:44:25 <dennda> ok, I can live with that :)
2008-07-28T19:44:49 <CIA-53> Byeongweon [tasyblue@gmail.com] default * 3942:79956dda436a 1.8-guieditor-mbyeongweon/wiki/htdocs/applets/moinfckconfig.js: set gui editor html source formated
2008-07-28T19:44:49 <CIA-53> Byeongweon [tasyblue@gmail.com] default * 3943:561cd7560781 1.8-guieditor-mbyeongweon/wiki/htdocs/applets/moinFCKplugins/ (3 files in 3 dirs): make moinmoin fckeditor plugin dialog select first text element on load
2008-07-28T19:46:27 <CIA-53> Bastian Blank <bblank@thinkmo.de> default * 4108:900fe4c68ece 1.8-dom-bblank/MoinMoin/converter2/creole_in.py: Creole input converter - Cleanup rules generation
2008-07-28T19:56:39 <ThomasWaldmann> re
2008-07-28T20:19:35 <dennda> xorAxAx: Actually, you said Page.from_item(itemobj)...
2008-07-28T20:19:48 <dennda> How do you do such a "static method" on the class Page?
2008-07-28T20:23:31 <xorAxAx> dennda: @staticmethod?
2008-07-28T20:23:43 * ThomasWaldmann cleans browser.py
2008-07-28T20:23:44 <dennda> 2.3
2008-07-28T20:24:05 <xorAxAx> @foo\ndef bar... means def bar...\nbar = foo(bar)
2008-07-28T20:24:05 <moinBot> xorAxAx: Error: "foo\ndef" is not a valid command.
2008-07-28T20:30:13 <dennda> xorAxAx: thanks for the hint. useful thing
2008-07-28T20:31:56 <TheSheep> dennda: with the new backed, will deleted pages be displayed as nonexistent?
2008-07-28T20:32:35 <TheSheep> xorAxAx: isn't that new in python2.4?
2008-07-28T20:32:42 <dennda> it is, TheSheep
2008-07-28T20:32:54 <dennda> (answering the latter question)
2008-07-28T20:33:19 <TheSheep> dennda: but can you show its history?
2008-07-28T20:33:27 <dennda> As for the deletion: There is currently no implementation for deleting anything
2008-07-28T20:33:40 <dennda> Could you describe your usecase?
2008-07-28T20:34:28 <TheSheep> dennda: I have a page, say FrontPage, someone comes and uses DeletePage action on it, I came later and want to bring it back
2008-07-28T20:34:52 <TheSheep> dennda: so I go to RecentChanges, click on 'history', then on restore
2008-07-28T20:35:43 <TheSheep> dennda: it works like that now
2008-07-28T20:36:08 <dennda> How the DeletePage action is handled isn't dependant on the storage layer, is it?
2008-07-28T20:36:36 <TheSheep> dennda: http://moin.sheep.art.pl/SampleDeletedPage?action=info
2008-07-28T20:37:02 <dennda> If you look at the API as is, you will notice that there are no "delete" methods
2008-07-28T20:37:15 <TheSheep> dennda: what I'm concerned with is how do you tell, from the upper layers, whether the page exists or not
2008-07-28T20:37:57 <TheSheep> dennda: do you have to fetch the item's data and see if it's empty?
2008-07-28T20:39:02 <TheSheep> dennda: because this check is done for every internal link on every page, might be a little expensive like that
2008-07-28T20:40:56 <dennda> afaik, we didn't discuss the higher level implementation of pages being deleteted yet, so don't take my word for granted: What's wrong with backend.has_item("SampleDeletedPage")?
2008-07-28T20:41:18 <TheSheep> dennda: it will return true, because the item itself exists
2008-07-28T20:41:39 <TheSheep> dennda: just the last revision is empty
2008-07-28T20:42:13 <TheSheep> dennda: or it will return false, and then the PageInfo action will also say there is no such page
2008-07-28T20:44:02 <dennda> Well, when you delete a page you could just create a new revision with no content and some DELETED = True metadata
2008-07-28T20:44:57 <dennda> I would feel more comfortable though, if johill told me whether I am right saying that or not. :) (But I don't want to disturb his recovery-process right now)
2008-07-28T20:45:04 <TheSheep> shouldn't this be handled in the backend?
2008-07-28T20:45:12 <TheSheep> yeah
2008-07-28T20:45:18 <TheSheep> I just wanted to make sure
2008-07-28T20:45:20 <TheSheep> thank you
2008-07-28T20:45:55 <dennda> Hm, what would a delete_item method in the backend do?
2008-07-28T20:46:34 <TheSheep> dennda: set that flag, in your case
2008-07-28T20:46:49 <TheSheep> dennda: in case of hg repo it could do hg rm
2008-07-28T20:47:07 <TheSheep> or also just set some flag
2008-07-28T20:47:50 <dennda> I don't think it wouldn't do anything that you, as the user of the API, can't do with the API
2008-07-28T20:48:03 <dennda> I don't think it would do anything that you, as the user of the API, can't do with the API
2008-07-28T20:48:11 <dennda> this way :)
2008-07-28T20:48:27 <TheSheep> dennda: the point is, you want to do different things in different backends
2008-07-28T20:48:33 <dennda> (since you don't want to delete a page forever)
2008-07-28T20:48:46 <TheSheep> dennda: hg rm won't delete it forever
2008-07-28T20:49:25 * PawelPacana bbl
2008-07-28T20:53:52 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3905:0248e23277ef 1.8/MoinMoin/ (9 files in 5 dirs): DataBrowserWidget: remove toHTML/render methods, change all callers to use .format()
2008-07-28T20:54:02 <ThomasWaldmann> waldi: ^^
2008-07-28T20:54:20 <waldi> okay
2008-07-28T20:56:57 <ThomasWaldmann> waldi: that was only used by Despam action. did i miss something or how did you find it?
2008-07-28T20:57:09 <waldi> WantedPages
2008-07-28T21:00:55 <ThomasWaldmann> do you have more infos?
2008-07-28T21:17:17 <johill> DELETED = True sounds right to me, we don't actually ever remove anything
2008-07-28T21:17:21 <johill> at least for now
2008-07-28T21:22:48 <dennda> puh... :)
2008-07-28T21:23:07 <dennda> thanks johill
2008-07-28T21:27:06 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3906:bef1e62eebe5 1.8/MoinMoin/widget/browser.py: DataBrowserWidget: make API conform to base class
2008-07-28T21:30:21 <zenhase> re
2008-07-28T21:30:26 <zenhase> got myself an upgrade
2008-07-28T21:30:54 <zenhase> now my right wrist has a nice inline skating protector to keep my wrist straight and unencumbered
2008-07-28T21:31:12 <zenhase> bit harder to type, but at least i can type again at a reasonable speed
2008-07-28T21:45:23 <ThomasWaldmann> waldi: i don't see an error with DBW when using WantedPages, but it crashes rather early in wikiparser/compatformatter
2008-07-28T21:46:13 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 3907:1008dd34b71c 1.8/MoinMoin/ (8 files in 4 dirs): DataBrowserWidget: make callers use correct API (render method)
2008-07-28T21:46:59 <waldi> /MoinMoin/converter2/macro.py:39: DeprecationWarning: Macro WantedPages used request.write: <form action="/WantedPages" method="GET" name="dbw.form"><div><div><table id="dbw.table-48"><tbody><...
2008-07-28T21:47:03 <waldi> warn(message, DeprecationWarning)
2008-07-28T21:50:30 <ThomasWaldmann> that must be somewhere when it renders pages for getting pagelinks.
2008-07-28T21:50:51 <xorAxAx> waldi: you probably want to dump the traceback
2008-07-28T21:50:59 <xorAxAx> if some debugging flag is enabled
2008-07-28T21:57:47 <dennda> woot. diff kinda works now
2008-07-28T22:07:01 <ThomasWaldmann> dreimark: where's melita?
2008-07-28T22:07:19 <waldi> xorAxAx: and it should kill the attempt with an exception
2008-07-28T22:08:00 <CIA-53> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 4183:47564f7f5bd5 1.8-dom-bblank/ (72 files in 32 dirs): merge moin/1.8 main branch
2008-07-28T22:18:27 <ThomasWaldmann> "Improved the garbage filter in Paste from Word dialog." < from FCKeditor 2.6.3b news
2008-07-28T22:20:02 <xorAxAx> :)
2008-07-28T22:26:37 <dennda> Is it possible to bind the standalone server to a different adress?
2008-07-28T22:29:37 <xorAxAx> yes
2008-07-28T22:29:43 <xorAxAx> well documented
2008-07-28T22:31:38 <zenhase> port and interface are the config-options
2008-07-28T22:31:45 <zenhase> in wikiserverconfig.py
2008-07-28T22:31:54 <dennda> thanks
2008-07-28T22:32:17 <xorAxAx> dennda: reading and augmenting documentation is also expected from you
2008-07-28T22:33:29 <dennda> Sure, sorry
2008-07-28T22:52:19 <ThomasWaldmann> zenhase: nice breakdown in wsgiapp
2008-07-28T22:57:52 <CIA-53> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com default * 4376:c3d4144651df 1.8-storage-cdenter/MoinMoin/action/diff.py: storage: diff-action: Further changes. Still not finished
2008-07-28T22:57:56 <CIA-53> Christopher Denter <moin GUESSWHAT the DASH space DASH station ROUNDTHING com default * 4377:3db24b8f7271 1.8-storage-cdenter/MoinMoin/ (Page.py action/diff.py):
2008-07-28T22:57:56 <CIA-53> storage: Page.py / action/diff.py:
2008-07-28T22:57:56 <CIA-53> * Adding "static method" on Page-class that wraps a Page object around an item.
2008-07-28T22:57:56 <CIA-53> * Starting to remove Page-class usage from action/diff.py (in parts via the above mentioned static method for now)
2008-07-28T22:57:58 <CIA-53> * Further fixes / adjustments in action/diff.py
2008-07-28T22:58:00 <CIA-53> * Basic diff functionality works now.
2008-07-28T23:36:42 <zenhase> ThomasWaldmann: what broke down? :o
2008-07-28T23:36:55 <zenhase> ThomasWaldmann: oh wait, did you meant my splitting?
2008-07-28T23:37:34 <dennda> I interpreted that as a compliment
2008-07-28T23:37:54 <ThomasWaldmann> zenhase: yes, your splitting into nice little pieces and middlewares
2008-07-28T23:38:00 <zenhase> thank you
2008-07-28T23:38:24 <zenhase> it is already a lot easier to grasp this way :)
2008-07-28T23:38:42 * xorAxAx hands the walkes false friend award to ThomasWaldmann
2008-07-28T23:38:43 <xorAxAx> :-)
2008-07-28T23:38:59 <xorAxAx> waalkes even
2008-07-28T23:39:37 <ThomasWaldmann> xorAxAx: that's correct, not a false friend :P
2008-07-28T23:39:53 <zenhase> yes, i just misinterpreted :)
2008-07-28T23:40:28 <ThomasWaldmann> ambigue somehow, though :)
2008-07-28T23:40:57 <xorAxAx> i cant find the meaning you intended, which german word should it be? :)
2008-07-28T23:41:08 <ThomasWaldmann> very amigue even (looking at dict.leo.org)
2008-07-28T23:41:15 <ThomasWaldmann> +b
2008-07-28T23:41:36 <ThomasWaldmann> just translate it via dict.leo.org to see all german meanings
2008-07-28T23:41:42 <xorAxAx> i am using dict.cc
2008-07-28T23:41:43 <xorAxAx> locally
2008-07-28T23:41:45 <zenhase> there are a lot :o
2008-07-28T23:41:46 <xorAxAx> and neither of them fits
2008-07-28T23:42:04 <ThomasWaldmann> "Aufteilung" "Aufglieferung" fits well
2008-07-28T23:42:07 <zenhase> and it's everything between structuring and totally wrecking :)
2008-07-28T23:42:16 <xorAxAx> ok, didnt see Aufgliederung :)
2008-07-28T23:43:15 <ThomasWaldmann> but could've been also understood as "Versagen" or "Panne" or "Zusammenbruch" X)
2008-07-28T23:43:23 <xorAxAx> i still seem to infer from google results that native speakers prefer it for destructive evens
2008-07-28T23:43:41 <xorAxAx> for example, i cannot find "revenue breakdown" in docs written by native speakers
2008-07-28T23:53:21 <ThomasWaldmann> http://www.youtube.com/watch?v=OUL5ZOX98JU MoinMoin: MoinMoinChat/Logs/moin-dev/2008-07-28 (last edited 2008-07-27 22:30:02 by IrcLogImporter)
