2008-03-24T00:19:44 <dreimark> ThomasWaldmann: can you checkk if you have old session files on MMaster from me
2008-03-24T00:19:59 <dreimark> I guess you have
2008-03-24T00:22:05 <dreimark> (cache files)
2008-03-24T00:23:56 <dreimark> good night
2008-03-24T00:34:45 <ThomasWaldmann> dreimark: how?
2008-03-24T07:54:48 <the_unforgiven> Hi all.
2008-03-24T07:55:10 <the_unforgiven> I've been playing around with moin for about two weeks now - we plan to use it for our internal wiki
2008-03-24T07:55:24 <the_unforgiven> need some help regarding the same..
2008-03-24T07:56:03 <the_unforgiven> since it's going to be LAN-only wiki, we expect real big attachments from the users.
2008-03-24T07:56:34 <the_unforgiven> default way of handling file-uploads is via cgi.FieldStorage - which creates a temporary file, unlink()s the file and returns the file handle..
2008-03-24T07:58:37 <the_unforgiven> so, to move the file to the required area, we need to read through the whole file piece-by-piece and dump it in the required directory.
2008-03-24T07:59:38 <the_unforgiven> cgi.FieldStorage docs say that we could inherit from cgi.FieldStorage and override its make_file() method to have custom file handling - like dumping the file in the required directory to start off with..
2008-03-24T08:00:23 <the_unforgiven> I did that and got it working. But, the way I load it is by explicitly loading my class to create the form instance in the Request class(es)
2008-03-24T08:01:22 <the_unforgiven> is there a generic way of doing it? - just like we have wikiutil.importPlugin(), if we could allow pluggable hooks for cgi.FieldStorage replacements, that would make it very generic.
2008-03-24T08:08:08 <the_unforgiven> I could send the patches if this idea is approved - I don't know if this is already sorted out on the latest dev version. I've been using the stable 1.6.0 release.
2008-03-24T09:24:14 <ernestas> moin!
2008-03-24T09:37:18 <dreimark> moin ernestas
2008-03-24T09:45:22 * dreimark tests/debugs session on a local cgi wiki
2008-03-24T10:54:04 <dreimark> seems konqueror does not like how sessions are done in 1.7
2008-03-24T10:54:47 <dreimark> in 1.6 I don't have to login again for a new request
2008-03-24T11:20:12 <johill> xorAxAx: yes, save, of course, i.e. reject with an edit conflict
2008-03-24T11:20:32 <xorAxAx> johill: no, its not an edit conflict
2008-03-24T11:20:37 <xorAxAx> its a commit conflict :)
2008-03-24T11:22:12 <johill> hm?
2008-03-24T11:22:21 <johill> I was actually talking browser level
2008-03-24T11:23:01 <xorAxAx> ????
2008-03-24T11:23:08 <xorAxAx> how is the browser level related to the storage backend?
2008-03-24T11:23:13 <xorAxAx> you seem to be confusing something
2008-03-24T11:23:23 <johill> not at all, but there's a bug in the -storage branch that I cannot reproduce in the regular one
2008-03-24T11:23:37 <xorAxAx> yes, it doesnt unlock something
2008-03-24T11:23:44 <xorAxAx> didnt you fix that already?
2008-03-24T11:23:48 <xorAxAx> or is it yet another one?
2008-03-24T11:23:51 <johill> another
2008-03-24T11:23:54 <johill> I did this
2008-03-24T11:23:58 <johill> save page with content "a"
2008-03-24T11:24:01 <johill> open editor 1
2008-03-24T11:24:03 <johill> open editor 2
2008-03-24T11:24:06 <johill> in 1, save with "b"
2008-03-24T11:24:09 <johill> in 2, save with "c"
2008-03-24T11:24:14 <johill> -> three revisions, last is c
2008-03-24T11:24:20 <johill> (rather than edit conflict)
2008-03-24T11:24:21 <xorAxAx> thats broken, indeed
2008-03-24T11:24:44 <xorAxAx> dont forget to write it down somewhere, either on the wiki or in the todo/changes file
2008-03-24T11:24:47 <xorAxAx> on the branch
2008-03-24T11:25:29 <johill> it's in my personal todo list for now, I want to take a look
2008-03-24T11:25:32 <johill> btw
2008-03-24T11:25:35 <johill> I thought more about the locking
2008-03-24T11:25:48 <johill> and noticed that remove_revision/remove_item are totally useless
2008-03-24T11:25:58 <johill> i.e. lanius was right, you can only use them on a wiki that isn't even readable
2008-03-24T11:26:08 <xorAxAx> yes, that was a design decision
2008-03-24T11:26:12 <xorAxAx> see our discussions :)
2008-03-24T11:26:39 <johill> bad design decision, imho, since that means the admin will poke the filesystem/database directly rather than using moin
2008-03-24T11:26:49 <xorAxAx> ?
2008-03-24T11:26:53 <xorAxAx> why is that?
2008-03-24T11:26:55 <johill> (because it's so much easier than putting the wiki into maintenance mode first just to delete one page)
2008-03-24T11:27:33 <xorAxAx> well, seriously flawed argument here - removing something directly will be more hazarderous than using the methods in any case
2008-03-24T11:27:46 <johill> yes
2008-03-24T11:27:52 <johill> but that's not the problem
2008-03-24T11:27:55 <xorAxAx> also note that the removal cannot be done such easily for db backends
2008-03-24T11:28:06 <johill> the problem is that we cannot offer removing in the UI
2008-03-24T11:28:18 <johill> only by a script or so
2008-03-24T11:28:23 <xorAxAx> yes, why not ...
2008-03-24T11:28:33 <xorAxAx> but it would even work in the ui
2008-03-24T11:28:40 <xorAxAx> with an additional action
2008-03-24T11:28:56 <johill> not quite, because multiple admins might actually then use the admin ui
2008-03-24T11:29:00 <johill> and that has the same races
2008-03-24T11:29:15 <xorAxAx> then it would corrupt the data, well
2008-03-24T11:29:31 <xorAxAx> multiple admins can use a script
2008-03-24T11:29:33 <johill> well, more likely moin would crash on a problem
2008-03-24T11:29:58 <xorAxAx> of course i guess you have a better design that can avoid these problems :-)
2008-03-24T11:30:27 <johill> it's non-trivial, but I think it's possible using something like RCU
2008-03-24T11:30:44 <johill> of course, the backend can't be a dumb backend like in this design but has to actually do the locking
2008-03-24T11:33:49 <xorAxAx> RCU?
2008-03-24T11:34:13 <johill> http://en.wikipedia.org/wiki/RCU
2008-03-24T11:35:06 <johill> but yes, it's non-trivial
2008-03-24T11:35:20 <johill> so for now I'd be happy if this design decision was actually documented
2008-03-24T11:35:32 <johill> and no, "see discussions" doesn't cut it
2008-03-24T11:36:03 <xorAxAx> we didnt specifically rule out RCU
2008-03-24T11:36:37 <johill> the design does rule it out because locking is done at a higher level
2008-03-24T11:37:12 <johill> unless you want to add a lot of new backend methods for it etc.
2008-03-24T11:37:28 <johill> no, it's fine, we don't need to solve that problem
2008-03-24T11:37:46 <johill> admins who don't put it into maintenance mode loose, at the cost of some reader processes just crashing
2008-03-24T11:38:04 <johill> (because they look up an item before it was deleted and access it after)
2008-03-24T11:41:47 <johill> I just fear that at some point somebody will run into it and think "oh that's easy to fix, just catch the exception", and then we'll end up with having workarounds in far upper levels
2008-03-24T11:45:54 * dreimark can reproduce the bug
2008-03-24T11:46:01 <johill> oh how?
2008-03-24T11:46:35 <johill> hm actually
2008-03-24T11:46:42 <johill> just make a bug page, I have to go now
2008-03-24T11:47:09 <dreimark> will do
2008-03-24T11:47:26 <johill> thanks
2008-03-24T11:47:27 <johill> bbl
2008-03-24T11:54:32 <lanius> i gues the solution for all this would be transactions?
2008-03-24T11:54:47 <xorAxAx> lanius: ?
2008-03-24T11:54:54 <lanius> to avoid locking
2008-03-24T11:54:59 <xorAxAx> the model already implicitly contains transactions
2008-03-24T11:55:12 <lanius> transactions on the backend level
2008-03-24T11:55:22 <xorAxAx> well, transactions doesnt mean anything :)
2008-03-24T11:55:38 <xorAxAx> maybe you want to look at the slides again that i sent you once :)
2008-03-24T11:56:18 <lanius> of course it means something :D
2008-03-24T11:56:48 <xorAxAx> "A database transaction, by definition, must be atomic, consistent, isolated and durable."
2008-03-24T11:56:58 <lanius> yes
2008-03-24T11:56:59 <xorAxAx> just that sentence makes it interesting :-)
2008-03-24T11:57:21 <lanius> you create a transaction, perform the operations on it and then save the transaction
2008-03-24T11:57:28 <lanius> which would avoid locks again
2008-03-24T11:57:41 <xorAxAx> well, you already implemented atomic operations - one idea would be indeed to have an interface on the backend side that allows you to hand in a list of operations and expect transaction properties
2008-03-24T11:57:41 <lanius> the transaction then must assure that it is acid
2008-03-24T11:57:49 <lanius> yes
2008-03-24T11:57:53 <lanius> that is what i mean
2008-03-24T11:57:55 <xorAxAx> yes
2008-03-24T11:58:24 <xorAxAx> but this sounds a bit too simplistic
2008-03-24T11:58:33 <xorAxAx> because a transaction may also include reads
2008-03-24T11:58:44 <xorAxAx> if it was just about writes, it would be very easy :-)
2008-03-24T11:59:12 <xorAxAx> so a transaction is not some kind of container but a context
2008-03-24T11:59:21 <lanius> yes
2008-03-24T11:59:40 <lanius> makes it more difficult, right
2008-03-24T11:59:49 <lanius> didn't say it would be easy ;)
2008-03-24T12:00:06 <xorAxAx> and then you can also directly resort to couchdb or zopedb :)
2008-03-24T12:18:10 <ThomasWaldmann> moin
2008-03-24T12:22:52 <ThomasWaldmann> dreimark: i have also found others problem (which might occur or not). e.g. if the cookie is expired (or has invalid hash), we do nothing. if the browser also does nothing, this inhibits login.
2008-03-24T12:30:34 <dreimark> I do wonder currently why I can get "session cookie we got had invalid hash" at login time
2008-03-24T12:31:53 <dreimark> I think you do describe this behaviour
2008-03-24T12:32:45 <dreimark> on the other hand I thought at that point the cookie is created
2008-03-24T12:46:17 <dreimark> That is definitly wrong. At login time it should not read the cookie from the browser.
2008-03-24T12:53:11 <dreimark> if you delete your browser cookie then the oldest session file is deleted an a new one is created
2008-03-24T12:55:08 <dreimark> have to go too. bbl
2008-03-24T18:06:27 <dreimark> ThomasWaldmann: currently a deleted session is recovered from the browsers cookie
2008-03-24T18:17:20 <ThomasWaldmann> dreimark: i am currently doing some changes to moin_session
2008-03-24T18:17:26 <ThomasWaldmann> you can help testing soon
2008-03-24T18:19:22 <dreimark> ThomasWaldmann: I have one patch at http://paste.pocoo.org/show/35304/
2008-03-24T18:20:06 <dreimark> this patch is for the corrupt cookie login dilemma
2008-03-24T18:20:41 <dreimark> And I think if o the server side the seesions are deleted every user should be forced to login again
2008-03-24T18:21:53 <ThomasWaldmann> dreimark: see my last change for moin_session some time ago
2008-03-24T18:33:26 <ThomasWaldmann> (the problem is that there is not only moin_login, but also stuff like http auth - so one has to be very careful when doing stuff dependant on the login/logout buttons)
2008-03-24T18:35:54 <ThomasWaldmann> hey zenhase :)
2008-03-24T18:36:19 <zenhase> ah, it connected
2008-03-24T18:36:23 <zenhase> hi there
2008-03-24T18:37:03 <dreimark> byeah, but currently we can't force a user to login by deleting his session
2008-03-24T18:37:19 <dreimark> moin zenhase
2008-03-24T18:37:48 <ThomasWaldmann> dreimark: my (slightly bigger) patch will be available in a few minutes
2008-03-24T18:37:49 <zenhase> moinmoin dreimark
2008-03-24T18:38:05 <dennda> moinmoin/student/* == former gsoc student?
2008-03-24T18:38:13 <zenhase> yeah kindof
2008-03-24T18:38:37 <zenhase> my project last year didn't quite work out
2008-03-24T18:38:47 <dennda> oh
2008-03-24T18:38:58 <dennda> Do you intend to apply again?
2008-03-24T18:39:30 <dreimark> ThomasWaldmann: :)
2008-03-24T18:40:59 <dreimark> zenhase: there are some nice project ideas waiting to be done ;)
2008-03-24T18:41:30 <dennda> zenhase: why didn't it work out?
2008-03-24T18:43:09 <zenhase> dennda: in retrospective i would say it was mainly because i was stubborn and created myself some problems
2008-03-24T18:43:33 <zenhase> a case of not invented here
2008-03-24T18:44:03 <dennda> zenhase: do you intend to apply again this year?
2008-03-24T18:44:11 <zenhase> actually i did not intend
2008-03-24T18:44:22 <zenhase> i read today that applications start again today
2008-03-24T18:44:39 <dennda> Yes, that is correct. In about one hour from now
2008-03-24T18:44:45 <zenhase> and when i went to the moin soc 08 page i was quite interested how that would work out
2008-03-24T18:45:07 <zenhase> the ideas look nice and i recognize some of them from last summer :)
2008-03-24T18:45:36 <ThomasWaldmann> dennda: he had also problems because moin's send_page is quite a huge old mess
2008-03-24T18:45:57 <dreimark> zenhase: may be you are interested in reading http://moinmo.in/ReimarBauer/UnderConstruction
2008-03-24T18:46:01 <dennda> ThomasWaldmann: I assume you were his mentor?
2008-03-24T18:46:18 <ThomasWaldmann> no
2008-03-24T18:46:21 <zenhase> ThomasWaldmann: thanks for defending me, but i still think it was also mostly an issue on my behalf :)
2008-03-24T18:46:31 <ThomasWaldmann> but mentoring is always a bit of teamwork
2008-03-24T18:46:37 <zenhase> yeah
2008-03-24T18:46:53 <zenhase> i think here at moin you are not really fixed at your mentor
2008-03-24T18:47:02 <zenhase> everybody helps
2008-03-24T18:47:31 <dreimark> the mentor has to do all the forms ...
2008-03-24T18:47:50 <dennda> I hope to find that out on myself this year :)
2008-03-24T18:47:51 <zenhase> thomas and xorAxAx, being the most irc-active imo, were always there if you needed sth
2008-03-24T18:48:04 <zenhase> dennda: what are you going to do? :)
2008-03-24T18:48:07 * xorAxAx will be less often there in the next months :)
2008-03-24T18:48:30 <dreimark> xorAxAx: can you update the irc statistics ?
2008-03-24T18:48:53 <dennda> zenhase: I'm going to apply for the "Extend Storage Backend" and the doodle task
2008-03-24T18:48:59 <xorAxAx> dreimark: no, my db isnt uptodate
2008-03-24T18:49:00 <zenhase> doodle task?
2008-03-24T18:49:09 <dennda> just search the page for doodle
2008-03-24T18:49:12 <grzywacz> hi
2008-03-24T18:49:46 <zenhase> hi grzywacz :)
2008-03-24T18:50:12 <dreimark> hmm, neagulm is missing
2008-03-24T18:50:13 <grzywacz> oy zenhase :)
2008-03-24T18:50:32 * grzywacz waves to xorAxAx
2008-03-24T18:50:56 <zenhase> hmm
2008-03-24T18:52:17 <zenhase> yeah, with neagulm the last years students would all be here
2008-03-24T18:52:59 <xorAxAx> hi my ex-student :)
2008-03-24T18:54:15 <grzywacz> :)
2008-03-24T18:54:39 <xorAxAx> have been enjyoing wesnoth 1.4 campaigns lately
2008-03-24T18:55:27 <dennda> only three students last year?
2008-03-24T18:55:33 <zenhase> dennda: 4
2008-03-24T18:55:55 <grzywacz> Applications start in like 1 hours?
2008-03-24T18:55:58 <grzywacz> *hour
2008-03-24T18:56:23 <zenhase> seems so
2008-03-24T18:56:25 <dreimark> dennda: http://moinmo.in/GoogleSoc2007
2008-03-24T18:57:07 <xorAxAx> in about 1 hour, yes
2008-03-24T18:58:18 * dennda shivers
2008-03-24T18:58:23 <zenhase> oh, i didn't check my gmail-account for some time ... seems i am still subscribed to those MLs from last soc
2008-03-24T18:58:36 <zenhase> dennda: don't be afraid? why the shivers?
2008-03-24T18:59:55 <dennda> zenhase: I hope there are enough slots to hit
2008-03-24T19:00:23 <xorAxAx> dennda: if there arent, we havent encouraged enough students to apply :-)
2008-03-24T19:00:55 <zenhase> how many of the people are applying students?
2008-03-24T19:01:00 <zenhase> +here
2008-03-24T19:01:02 <dennda> xorAxAx: well, you'll always have less slots than applications, won't you? :)
2008-03-24T19:01:16 <xorAxAx> yes
2008-03-24T19:01:18 * dennda is going to apply
2008-03-24T19:01:54 <dennda> zenhase: you sound like you were, too :)
2008-03-24T19:03:12 <zenhase> as said, i currently don't think i will :)
2008-03-24T19:03:34 <dennda> gnome had 25 slots last year, if I counted correctly
2008-03-24T19:03:35 <dennda> quite a few
2008-03-24T19:04:01 <zenhase> well, it's a huge project
2008-03-24T19:04:20 <dennda> it is
2008-03-24T19:04:28 <zenhase> and it broke just 2h ago on my home box :o
2008-03-24T19:04:59 <dennda> hehe
2008-03-24T19:05:46 <zenhase> oh, i almost didn't see kikka :o
2008-03-24T19:06:10 * zenhase pokes kikka
2008-03-24T19:06:29 <dennda> Will the ranking process be an open one or will it be closed to the public?
2008-03-24T19:06:30 <zenhase> ok, jabber says he's offline
2008-03-24T19:06:44 <dreimark> wait some hours, you are to early
2008-03-24T19:07:11 <xorAxAx> dennda: will be closed
2008-03-24T19:08:07 <kikka> zenhase!111
2008-03-24T19:08:09 <dennda> hm ok
2008-03-24T19:08:16 <kikka> zenhase: Working on gsoc, yepp?
2008-03-24T19:08:17 <zenhase> oh, i woke him :)
2008-03-24T19:09:02 <zenhase> kikka: no, but i wanted to watch the applicants this year and the proposals
2008-03-24T19:09:36 <kikka> Ah!
2008-03-24T19:10:02 <zenhase> you did some ghopping didn't you? :)
2008-03-24T19:10:28 <kikka> I never finish anyth! ;)
2008-03-24T19:10:40 <zenhase> :o
2008-03-24T19:11:28 * kikka saw the olpc-gsoc meeting last evening. :)
2008-03-24T19:14:18 <dreimark> kikka: do they have plans to continues with 1g1g ?
2008-03-24T19:15:42 <kikka> dreimark: I don't think so. :/
2008-03-24T19:16:59 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2605:ae2d9a64c294 1.6/ (MoinMoin/auth/__init__.py docs/CHANGES): moin_session: delete invalid cookies, create a new session if we get a valid user from prev. auth methods at the same time
2008-03-24T19:17:02 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2606:9f1e99ad34c5 1.6/wiki/underlay.tar: merged main
2008-03-24T19:17:42 <ThomasWaldmann> dreimark: they did not finish their first g1g1 yet, afaik
2008-03-24T19:18:45 <ThomasWaldmann> please review/test those changeset ^^^
2008-03-24T19:19:33 <kikka> Hey ThomasWaldmann! :)
2008-03-24T19:20:06 <ThomasWaldmann> hi kikka :)
2008-03-24T19:21:02 * dreimark pulls
2008-03-24T19:21:41 * ThomasWaldmann hopes it has no regression(s)
2008-03-24T19:21:59 <ThomasWaldmann> johill: if you have time, please have a look
2008-03-24T19:22:10 <dennda> don't you have unit tests to help detect those regressions?
2008-03-24T19:22:42 <dennda> (at least some of them)
2008-03-24T19:22:43 <ThomasWaldmann> dennda: we don't have full coverage
2008-03-24T19:22:56 <dennda> oh
2008-03-24T19:22:59 <dreimark> dennda: we have but for some it is not easy to setup a unit test
2008-03-24T19:23:00 <dennda> that should be fixed
2008-03-24T19:23:04 <dennda> ah ok
2008-03-24T19:23:41 <ThomasWaldmann> the main point of that change is to a) delete invalid session cookies (so there should never be a dead end situation as before that cs)
2008-03-24T19:24:12 <dennda> ThomasWaldmann: does that fix the login problem?
2008-03-24T19:24:41 <dennda> s/does that/is that supposed to/
2008-03-24T19:24:59 <ThomasWaldmann> and b) even if we receive a invalid cookie, we will create a new cookie/new session, if we get a valid user obj from some previous auth method (overwriting the invalid cookie)
2008-03-24T19:25:16 <ThomasWaldmann> dennda: if you are talking about the same login problem, yes :D
2008-03-24T19:25:30 <dennda> ThomasWaldmann: I hope there aren't that many :)
2008-03-24T19:25:59 <ThomasWaldmann> 1.6.2 changes only lists 2 :P
2008-03-24T19:31:17 <dreimark> ThomasWaldmann: b) is meant with login or with browser reload?
2008-03-24T19:31:20 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2607:7899d9d75762 1.6/ (3 files in 2 dirs): added dummy mig scripts for 1.6.2
2008-03-24T19:32:03 <ThomasWaldmann> dreimark: that depends on the previous auth method
2008-03-24T19:32:34 <ThomasWaldmann> if it is moin_login, you will have to enter username/password and click login
2008-03-24T19:32:59 <dreimark> I supposed that this was meant, ok.
2008-03-24T19:33:07 <ThomasWaldmann> if it is http basic auth, processing will happen with any request anyway
2008-03-24T19:34:20 <dreimark> in 1.7 it is easy to get whose sessions are left over
2008-03-24T19:35:13 <dreimark> new session creations works too. (the old file stays in the dir)
2008-03-24T19:36:59 <dennda> zenhase: you are heinrich wendel?
2008-03-24T19:37:36 <dennda> http://code.google.com/soc/2007/moin/about.html states only three tasks
2008-03-24T19:38:20 <kikka> dennda: zenhase is away, but he's not heinrich wendel.
2008-03-24T19:39:09 <dreimark> dennda: no - lanius is
2008-03-24T19:39:16 <dennda> ok
2008-03-24T19:42:17 <kikka> I should tidy up my room... :/
2008-03-24T19:42:48 <dennda> Let the garbage collector do it for you
2008-03-24T19:43:00 <ThomasWaldmann> xorAxAx: the leading blanks for the Description in setup.py are intended/valid?
2008-03-24T19:43:40 <xorAxAx> umm, not intended but surely valid. the submission looks fine in pypi (besides the maintainer/author issue)
2008-03-24T19:44:39 <ThomasWaldmann> platform: is only used for documentation?
2008-03-24T19:45:17 <dreimark> MAX_STORED_SECRETS = 2 works too :)
2008-03-24T19:45:33 <dreimark> and a third browser kills a session of an other browser too
2008-03-24T19:48:33 <dreimark> deleted session files are recreated if the cookie is valid
2008-03-24T19:59:27 <ThomasWaldmann> can anyone who not has visited master.moinmo.in for a while try to log in there and check if there are issues?
2008-03-24T20:00:48 <starshine> how long a while?
2008-03-24T20:00:53 * starshine probably hasn't.
2008-03-24T20:02:26 <ThomasWaldmann> more than 24h
2008-03-24T20:02:51 <ThomasWaldmann> dreimark: did you also check anon sessions?
2008-03-24T20:04:28 * starshine tests the lost password thingy ;P
2008-03-24T20:08:49 <starshine> well it works this way at least
2008-03-24T20:09:08 <dreimark> ThomasWaldmann: currently. they seem not to create a session
2008-03-24T20:09:23 <starshine> can you check if I have a created session currently or not
2008-03-24T20:09:29 <starshine> I should, I would think.
2008-03-24T20:12:01 <dreimark> starshine: you have if you can go to various pages and aren't logged off
2008-03-24T20:13:12 * dreimark remembers need some changes of config for anonn sessions
2008-03-24T20:15:35 <ThomasWaldmann> starshine: if you have time, you could read docs/CHANGES (1.6.current) section if there is weird wording in it.
2008-03-24T20:16:38 <starshine> dreimark: ok, passes.
2008-03-24T20:17:05 <starshine> general thoughts on frontpage: SyntaxReference now redirects, we should probably fix the link so it bears the new name :)
2008-03-24T20:19:12 * starshine fixifixi
2008-03-24T20:20:09 <dreimark> ThomasWaldmann: auth = [authmodule.moin_login, authmodule.moin_session, authmodule.moin_anon_session]
2008-03-24T20:20:20 <dreimark> gives no sessions for anon user too
2008-03-24T20:21:37 <starshine> it's supposed to get rid of the session if I logout by choice, right?
2008-03-24T20:21:53 <dreimark> right
2008-03-24T20:21:53 <ThomasWaldmann> starshine: i already made the underlay for 1.6.2
2008-03-24T20:22:27 <starshine> awww :(
2008-03-24T20:22:42 <ThomasWaldmann> but will be in 1.6.3 / 1.7 soon :)
2008-03-24T20:23:00 <ThomasWaldmann> dreimark: did you set the anon cookie lifetime?
2008-03-24T20:23:07 <starshine> in that case I guess I should be more thorough and hunt up the rest of the oldlinks?
2008-03-24T20:23:26 <dreimark> oops no, doing now
2008-03-24T20:23:34 <ThomasWaldmann> starshine: I rather meant that docs/CHANGES file in the moin dist :)
2008-03-24T20:24:10 <ThomasWaldmann> but if there is some cleanup on moinmaster BEFORE we fork off the 1.7 master wiki, it is for sure better than afterwards
2008-03-24T20:24:33 <starshine> good, I'll do that right away
2008-03-24T20:24:44 <starshine> hubby is using the link so this might not be best moment for a pull :)
2008-03-24T20:24:59 <ThomasWaldmann> you need more bw :)
2008-03-24T20:25:09 <starshine> yes, yes I do.
2008-03-24T20:25:19 <ThomasWaldmann> no neighbor with a good, open wlan :D
2008-03-24T20:26:32 <starshine> it's referenced by english name in most of the In(lang)Group pages.
2008-03-24T20:27:07 * starshine doublechecks she's in UTF-8 ..
2008-03-24T20:28:07 <dreimark> ThomasWaldmann: added anonymous_cookie_lifetime = 2
2008-03-24T20:28:15 <dreimark> now I get with every request a new session
2008-03-24T20:28:37 <starshine> for the i18n strings, you want the string left alone and the english translation fixed to translate it to be correct?
2008-03-24T20:28:40 <ThomasWaldmann> starshine: http://hg.moinmo.in/moin/1.6/raw-file/tip/docs/CHANGES
2008-03-24T20:29:58 <starshine> ah link is free, pulling now.
2008-03-24T20:30:04 <dreimark> ThomasWaldmann: the page trail has only the current page and each request is a new session file
2008-03-24T20:30:35 <ThomasWaldmann> ok, i'll check it
2008-03-24T20:31:07 * dreimark sets up a http_auth
2008-03-24T20:32:23 <zenhase> dennda: no i am florian krupicka
2008-03-24T20:33:13 <dennda> ok
2008-03-24T20:33:16 <dennda> google is late :)
2008-03-24T20:33:16 <zenhase> dennda: as said, my project did not really work out, so we cancelled it midterm
2008-03-24T20:33:36 <zenhase> perhaps that is the reason it does not appear on the page anymore
2008-03-24T20:35:20 <dennda> ok
2008-03-24T20:35:48 <zenhase> why is google late?
2008-03-24T20:35:58 <zenhase> still no application form? :)
2008-03-24T20:37:52 <ThomasWaldmann> dreimark: anon sessions work for me
2008-03-24T20:39:09 <dreimark> hmm, I do start with a clean session again
2008-03-24T20:39:16 <dreimark> (-dir)
2008-03-24T20:40:27 <ThomasWaldmann> from MoinMoin.auth import moin_login, moin_session, moin_anon_session
2008-03-24T20:40:27 <ThomasWaldmann> auth = [moin_login, moin_session, moin_anon_session]
2008-03-24T20:40:27 <ThomasWaldmann> anonymous_cookie_lifetime = 1
2008-03-24T20:40:27 <ThomasWaldmann> moin_session_verbose = True
2008-03-24T20:43:01 <dreimark> not so much different from mine
2008-03-24T20:43:06 <dreimark> anonymous_cookie_lifetime = 2
2008-03-24T20:43:11 <dreimark> moin_session_verbose = True
2008-03-24T20:43:15 <dreimark> import MoinMoin.auth as authmodule
2008-03-24T20:43:28 <dreimark> auth = [authmodule.moin_login, authmodule.moin_anon_session, authmodule.moin_anon_session]
2008-03-24T20:44:41 <dennda> zenhase: yes :)
2008-03-24T20:45:04 <xorAxAx> yes, google is late
2008-03-24T20:45:05 <xorAxAx> 20:41:29 <@lh> sangprabo: because the web app is a festering pile of happiness
2008-03-24T20:45:46 <ThomasWaldmann> lol
2008-03-24T20:45:49 <dennda> nicely said
2008-03-24T20:46:19 <dreimark> last time it was something around 3h later
2008-03-24T20:47:40 <ThomasWaldmann> dreimark: i just put what i posted into wikiconfig and it worked. doesn't that work for you?
2008-03-24T20:47:49 <dreimark> yours work for me too
2008-03-24T20:49:02 <dreimark> ThomasWaldmann: mine doesn't the import seems to cause the problem
2008-03-24T20:49:09 <dennda> xorAxAx: < mhilmi> scorche: you dont think I would get special attention if the header of my proposal said "FIRST" :P <-- :D
2008-03-24T20:49:15 <xorAxAx> :-)
2008-03-24T20:49:33 <zenhase> first post!
2008-03-24T20:50:52 <dennda> there's a nice xkcd on that :)
2008-03-24T20:51:08 <zenhase> on what? first-post-ism?
2008-03-24T20:51:26 <dreimark> ThomasWaldmann: why does it make this difference to use import MoinMoin.auth as authmodule ?
2008-03-24T20:51:48 <dreimark> that way it is done in multiconfig
2008-03-24T20:55:06 <dennda> zenhase: yes
2008-03-24T20:55:28 <ThomasWaldmann> dreimark: your method also works for me
2008-03-24T20:55:51 <ThomasWaldmann> you indented it correctly?
2008-03-24T21:03:14 <dreimark> yes, http://paste.pocoo.org/show/35316/ that seems not the problem
2008-03-24T21:04:26 <dreimark> and it is reproducable
2008-03-24T21:07:08 * dreimark tries to debug this behaviour
2008-03-24T21:07:24 <ThomasWaldmann> auth = [authmodule.moin_login, authmodule.moin_anon_session, authmodule.moin_anon_session]
2008-03-24T21:07:29 <ThomasWaldmann> duplicate
2008-03-24T21:09:04 <dreimark> yeah it was a good idea to post it ;)
2008-03-24T21:10:39 <dreimark> ThomasWaldmann: works :)
2008-03-24T21:31:59 * ThomasWaldmann runs make dist
2008-03-24T21:32:30 <dreimark> ThomasWaldmann: wait
2008-03-24T21:32:43 <dreimark> I have problems with http_auth
2008-03-24T21:33:08 <ThomasWaldmann> like?
2008-03-24T21:34:23 <dreimark> doing from MoinMoin.auth import moin_login, moin_session, moin_anon_session, http
2008-03-24T21:34:32 <dreimark> auth = [moin_login, moin_session, moin_anon_session, http]
2008-03-24T21:35:31 <ThomasWaldmann> if you do it like that, http auth will likely work, but not establish a session
2008-03-24T21:36:15 <dennda> Is the storage-code already in use, actually?
2008-03-24T21:36:36 <ThomasWaldmann> dennda: the storage branch is not merged yet
2008-03-24T21:36:51 <dennda> ThomasWaldmann: The goal is to get it merged into 1.7?
2008-03-24T21:37:03 <dennda> (1.7 main branch, that is)
2008-03-24T21:37:14 <ThomasWaldmann> dennda: 1.8 rather
2008-03-24T21:37:34 <ThomasWaldmann> for 1.7 the goal is to release it soon
2008-03-24T21:39:23 <dennda> ah ok
2008-03-24T21:39:33 * dennda is just diving further into the backend code
2008-03-24T21:40:04 <dennda> just wanted to know if anyone ever tested it
2008-03-24T21:41:21 <ThomasWaldmann> yes, but not in production use
2008-03-24T21:42:29 <dreimark> ThomasWaldmann: http://moinmo.in/ReimarBauer?action=AttachFile&do=view&target=traceback.html
2008-03-24T21:44:32 <ThomasWaldmann> dreimark: it is http.http
2008-03-24T21:46:50 <ThomasWaldmann> xorAxAx: the name change in setup.py now creates MoinMoin-1.6.2.tar.gz
2008-03-24T21:47:10 <ThomasWaldmann> (and a directory MoinMoin-1.6.2 inside)
2008-03-24T21:47:17 <dennda> is it considered "complete" or is the current storage system from that branch still unfinnished?
2008-03-24T21:47:36 <xorAxAx> ThomasWaldmann: hmm
2008-03-24T21:47:44 <ThomasWaldmann> dennda: read the ideas page
2008-03-24T21:48:15 <xorAxAx> ThomasWaldmann: moin itself is fine, but the pypi package is called differently. so i guess you could change it back to moin and use MoinMoin only for register
2008-03-24T21:48:27 <xorAxAx> (hacky)
2008-03-24T21:48:47 <dennda> ThomasWaldmann: I meant something else but I think I answered it myself, nevermind :)
2008-03-24T21:49:05 <dreimark> ThomasWaldmann: yeah
2008-03-24T21:52:12 <dreimark> ThomasWaldmann: sorry it was my first setup of http_auth on ubuntu. (it is in principle easier as on SuSE)
2008-03-24T21:52:39 <dreimark> but therefore I did not recovered the things I should know
2008-03-24T21:58:47 <dreimark> ThomasWaldmann: I do have the same trouble with the gui editor on MM
2008-03-24T22:01:04 <dennda> boy... the first thing that really needs to be done is updating the developer api documentation of the storage thing
2008-03-24T22:01:26 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2608:5e7adef9a89c 1.6/ (MoinMoin/__init__.py MoinMoin/version.py README docs/CHANGES): updating some (c) lines and docs/CHANGES
2008-03-24T22:01:30 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2609:95a0990c9c60 1.6/ (4 files in 3 dirs): fix some usages of underlay.tar.bz2 to use underlay.tar, use 'moin' for setup.py
2008-03-24T22:49:50 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2610:ac1d2222bfa5 1.6/ (10 files in 6 dirs): release 1.6.2 (bumping version numbers)
2008-03-24T22:51:28 <CIA-39> Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2611:6a49ee06febc 1.6/.hgtags: tagged release 1.6.2
2008-03-24T22:56:16 <xorAxAx> student application is open now! please apply :-)
2008-03-24T23:01:40 <dennda> xorAxAx: lemme finish the application first :)
2008-03-24T23:01:56 <dennda> i'm still reading parts of the storage backend
2008-03-24T23:02:31 <xorAxAx> hehe, you have some days left
2008-03-24T23:02:40 <dennda> "quoting" a wiki name, in moinmoin terms, means returning a file system representation of the potentially unicode-wiki-name?
2008-03-24T23:03:37 <ThomasWaldmann> dennda: if it is not urlquote then yes
2008-03-24T23:03:47 <dennda> FSquote
2008-03-24T23:04:17 <dennda> ok
2008-03-24T23:45:08 * starshine back
2008-03-24T23:50:48 <ThomasWaldmann> wb starshine
2008-03-24T23:56:20 <dennda> http://paste.pocoo.org/show/35352/
2008-03-24T23:56:28 <dennda> is it me or should it be just the other way around?
2008-03-24T23:57:44 <dennda> taking a look at one implementation in backends/filesystem.py that proves my point
2008-03-24T23:58:25 <ThomasWaldmann> looks weird
2008-03-24T23:59:21 <ThomasWaldmann> maybe written by a sleep-deprived student :P
2008-03-24T23:59:35 <dennda> yeah :)
2008-03-24T23:59:52 * dennda thinks he just found his first moinmoin-documentation-bug
MoinMoin: MoinMoinChat/Logs/moin-dev/2008-03-24 (last edited 2008-03-23 23:30:02 by IrcLogImporter)
