[05:47] check this out: http://video.google.com/videoplay?docid=7707585592627775409# [07:32] does anyone know when the uk english version will be released? An e-mail asked on the ubuntu-uk mailing list [08:00] hey godbyk how is it going ? [08:01] hey c7p I will upload the pictures as soon as i can [08:02] ubuntujenkins: ty, sorry for the yesterday inconvenience [08:03] no problem c7p [08:34] IlyaHaykinson: hi! docbook example leads to: XML Parsing Error: undefined entity Location: http://www.netapt.com/~ilyah/ubuntumanual/ubuntu-desktop.xml Line Number 180, Column 66: [08:34] it complains about the ampersand [08:35] thorwil: I didn't say it was perfect! :) [08:35] It doesn't like × I'm guessing? [08:35] if so, just replace it with 'x' for now. [08:35] Or was there another ampersand I missed in there? [08:35] godbyk: well, one can always use view source [08:36] right [08:36] I just converted it by hand to give us an idea of what docbook could do. [08:36] what the tags looked like, [08:36] and to make sure it covered all the stuff we needed. [08:36] a long time ago i tried to write a document with docbook, rendering to pdf via fop [08:36] I think the only stuff not in that example is the terminal code. but that's pretty easy for docbook, I imagine. [08:37] I haven't figured out the docbook toolchain yet. it seems overly complicated. [08:37] that experiment stopped when i realized i would need some not so trivial layout to add images. so it was killed for the same reason as my earlier flirt with latex [08:38] I figured if nothing else, we could just run it through a script to convert to our tex code, tweak the result (for typographic perfection), and run it through xelatex to generate the pdf. [08:38] thorwil: view-source on it to read the code [08:38] oh, nod. [08:40] godbyk: yes, conversion via latex is likely easier and might lead to better typesetting. doubtful that this fop business saw the same level of refinement in that regard [08:42] sorry folks. gotta go sleep. getting up early. l8r. [08:43] IlyaHaykinson: I just woke up. [08:43] Silly UDS in UTC+2. :-/ [08:45] heh [08:46] Anyway, g'night, IlyaHaykinson. [08:59] g' morning godbyk how do the translated pdfs go? [09:01] Hey, c7p. [09:01] humpreybc's been chasing launchpad developers around UDS trying to get answers for us. [09:01] currently, they suspect it may be due to a launchpad bug that's already on their to do list. [09:01] I'm trying to get more details, though. [09:02] Ben gotta catch them all! [09:02] we had a problem with lp but now the translation is 100% [09:02] i think we can move on (or not?) [09:03] c7p: Probably. [09:04] If you've finished the translation, I'll rebuild the PDFs and you can start editing (fixing spelling errors, grammar errors, typos, etc.) [09:05] we have worked on these errors with the team the previous week, also i made a change yesterday so from this aspect we are ok [09:06] Oh, okay. So you think you're done editing all the strings, then? [09:06] we are walking always about the greek translation [09:06] Morning, synergetic. [09:06] morning, godbyk ^_^ [09:06] godbyk: yes :) [09:06] c7p: Awesome! Okay. [09:07] I'll write up some notes today to walk you through the next steps then, c7p. [09:09] godbyk: nice :), if it is needed i can tell you some bugs that i found out on the pdf, when i compiled the pdf successfully yesterday [09:09] c7p: Sure - fire away! [09:14] c7p and translators: Here's the latest about the Launchpad translations issue: [09:14] godbyk: firstly the outer page's svg wasn't generated in greek as it used to. Secondly on the "chaplinks" e.g 'Chapter : Working with Ubuntu' are on pdf as 'Κεφάλαιο ??: ??.' (only the Chapter is translated). Thirdly the heading is kind weird (i think you remember the situation) [09:14] We can work around the bug if I disallow translators from editing the .po files directly. [09:15] godbyk: sorry for the intervention, you may go on [09:15] That's about it. [09:15] It's a bug in Launchpad. [09:15] The problem occurs when Launcpad reads the .po files. It will overwrite some of the strings in Launchpad under some circumstances. [09:16] To work around the bug, we can tell Launchpad to ignore the .po files. [09:16] This means that you can't edit the .po files directly anymore, though -- you *must* do all the work directly in Launchpad. [09:16] I'll email this to the list and see what others think, too. [09:17] c7p: If the \chaplink command is showing ??s then it means that the build died somewhere along the way. xelatex has to be run multiple times to resolve the cross-references. [09:18] godbyk: good [09:21] godbyk: also the pictures are a bit distorted but this happens to all the translated pdfs that I've seen [09:21] c7p: Distorted in what way? [09:22] (I'm rebuilding all the PDFs with the latest translations right now. They'll be done in about 30 minutes.) [09:22] they aren't very clear, no big deal [09:22] I think that's an issue with evince. [09:23] godbyk maybe it's evince, I ll check it [09:43] yap i can confirm that it's an evince issue [09:58] c7p: In the Greek translation, string 1001 has a bug. The \gls command shouldn't be translated. [09:59] let me correct it [10:01] done, sorry for the inconvenience [14:53] Anyone know when UDS ends? [14:54] Red_HamsterX: It's over on Friday. [14:54] Ah. Thanks. :) [14:55] Did you get a chance to quickly review my whims about rewriting the Qs sevrer using Python (WSGI)? [14:55] I just want to make sure that wouldn't pose an insurmountable problem on your end. [14:55] I haven't yet. (Been wrapped up listening to UDS sessions) [14:55] If you go to wiki.dreamhost.com and search for python or wsgi, you can see what options are available. [17:02] I kinda want to try building it in Django, but fear that may be total overkill. [17:03] I wish I had more control over my web host. [17:03] But since I'm only paying US$7 a month or whatever, I can't complain too much. [17:04] I think I'm paying a bit less (WebFaction), but I signed up for five years. [17:05] I'm pretty happy with it, though. [17:05] I'm on Dreamhost. They're not too bad. [17:10] PHP is easy, but ugly and not as powerful as Python (IMO). Plus, it means maintainers would need to know two languages to hack on a single project. [17:10] Python CGI works, but it feels counter-intuitive and messy. [17:11] WSGI would work quite well, but DreamHost supports it through an extension that still isn't willing to call its WSGI support stable. [17:11] Django is just something I want to learn. [17:12] WSGI is probably the best choice overall, though, in terms of resource use versus maintainability. [17:12] Also, CGI is slower. [17:12] Yeah, it seems that dreamhost doesn't support WSGI quite yet. [17:12] (well, certainly not 'natively' via apache modules) [17:12] Speed isn't really a factor for this project, though. [17:13] A few milliseconds won't matter. [17:13] As long as our server isn't being hammered by requests. [17:14] Google's App Engine supports WSGI, so we could probably make the public service available there. [17:14] And use yours for UMP-specific activities. [17:14] Red_HamsterX: if python, then WSGI. i'd chose pylons or turbogears over django for being designed for WSGI from the start [17:14] And mine for development builds. [17:15] Of those, which is most suited to a small service, thorwil? [17:15] Database manipulation isn't terribly important to this design. [17:16] Red_HamsterX: pylons [17:16] turbogears is stuff added on top of pylons [17:17] database business is exactly what is different on the app engine [17:20] Pylons looks like fun. [17:20] Sold. :) [17:22] Mako seems to be a nice template language. it gets around inventing weird constructs like used in django's template language by basically just using python [17:23] though that means you have to trust template authors, which is why this is considered not-designer-friendly :} [17:23] Do you have experience with Pylons, in case I need to ask you for the term used to express an abstract idea? [17:24] (So I can Google more good-like) [17:27] Red_HamsterX: not past looking at their documentation [17:32] http://www.mutualinformation.org/2010/03/why-i-switched-to-pylons-after-using-django-for-six-months/ [17:33] Fair enough. [17:33] That's all I had as motivation for wanting to learn Django, too. [17:33] (Pretty-looking documentation) [17:39] Red_HamsterX, godbyk: are you able to grok this? http://en.wikipedia.org/wiki/Operational_transformation [17:42] thorwil: works for me. [17:45] Makes sense here, too, based on the abstract and some of the sample models. [17:46] Looks similar to database concurrency problems. [17:47] this is the basis for what google wave and etherpad do, afaik [17:47] thorwil: sorry, I misread grok as a different verb. [17:47] I'll read it in just a moment. [17:48] What relevance does this have to UMP? [17:48] Aside from being an always-cool research domain. [17:48] s/cool/fun/ [17:48] s/cool/fun/g [17:48] * Red_HamsterX messes with reality. [17:49] Red_HamsterX: etherpad like editing would be the ticket to a great platform for UMP [17:49] For writers or for users? [17:49] I don't see the application. :( [17:50] Red_HamsterX: say we have our document source on such an online platform. multiple writers could edit without locks or merge conflicts [17:51] of course this would also need an elegant solution to markup/tagging [17:52] Red_HamsterX: goodby to manual bzr pull and push [17:53] and the more granular the locking and edits are, the easier it is to avoid conflicts. [17:53] especially if we make it stupidly simple to make small changes -- that'll help a lot [17:53] then we'll have to have some videos of the diff animation, of course. :-) [17:53] heh [17:54] Wouldn't such a dynamic environment pose a great problem where versioning is concerned? [17:54] There's a group at my university doing some work in the CSCW field. [17:55] Nah, you'd just have a ton of very small-change revisions. [17:55] And that wouldn't pose a storage concern? [17:56] shouldn't be too bad. [17:56] plain text compresses pretty easily. [17:56] Red_HamsterX: at some point it might. a solution could be to reduce the granularity of the history for old edits [17:57] So, after a milestone has been passed, start compacting changes to day/week/month-level diffs? [17:58] It kinda raises the question of whether diffing would even be the right solution. Maybe whole-file snapshots (compressed) would be more logical. [17:58] Red_HamsterX: aside of our specific needs, there absolutely should be a wiki engine with such immediate collaborative editing. one that also does wysiwyg to not force users to learn strange markup [18:00] i worked with the ubuntu wiki a lot, in the artwork section. feels like it stoneage. it's actually a rather high barrier to entry [18:01] saw lots of ways how people get things wrong [18:03] I've found it to be pretty average, feature-wise. [18:03] A decent tool for anyone experienced with other wikis. [18:03] But, yeah, not very novice-friendly. [18:05] You're think with all the research that's been done on operational transformation, that conflicts with bzr et al. would occur far less often. [18:06] s/'re/'d/ [18:07] Hey, look.. Gobby's mentioned in that article, too. [18:07] (Gobby's UI is horrible!) [18:11] guess what arrived in the post today :) [18:11] a real doll? [18:12] no [18:13] hmm, a real manual? [18:13] yes \o/ [18:14] lol [18:15] dutchie: cool. print alright? would you perhaps be equipped, able and willing to create some _sharp_ hi-res photos of it? of the kind one could put in a portfolio? ;) [18:16] able and willing, yes. equipped, no [18:16] dang [18:19] * thorwil -> dinner [19:08] how do i get to know shipping cost from lulu? [19:10] it says when you order [19:10] but i'm not sure before then [20:00] ubuntujenkins mentioned the coupon "FREESHIP" in a response he just posted to the mailing list. === ianto is now known as DavidMiliband [22:53] Okay, off to bed. [22:53] UDS may kill me yet. Starting my day at 1 a.m. is just weird. === DavidMiliband is now known as ianto