[00:15] well the dummy package built, how ever ccicons failed again [00:24] I think it will try and fix it tomorrow night all [00:25] well its already morning here [03:41] godbyk, i imported translations from lp [03:41] daker: for the website? [03:41] yes [03:42] k [06:06] godbyk: ping [09:39] morning all [09:44] morning ubuntujenkins ^_^ [09:45] morning * :) [09:46] hello people are wake today [09:47] *awake [09:50] hello all [09:50] wake up? [09:51] not a good sign: 1169 untranslated yesterday became 1463 untranslated + 300 need reviews, now it is 1463 untranslated + 180 need reviews [09:52] where have they gone? !! [09:52] waiting for a day more, thus, #launchpad seems not very active [09:53] jcisio: we are trying our best tor work out what is wrong. #launchpad is usually very good [09:53] *to [09:54] sure I know ubuntujenkins [09:54] just see that it is not good timing, and UDS comes close [09:55] nice weekend all, bb [11:14] godbyk, any reponse from launchpad ? [13:06] daker: No response yet. I just re-broadcast my request for help. We'll see if anyone's around yet. [13:07] I think most the launchpad folks are en route to Belgium at the moment. [13:07] wgrant responded yesterday from an airport, but he said he doesn't know much about the translation part of Launchpad, so he couldn't help. [13:07] humphreybc too ? [13:07] Yep, humpreybc is on his way to Belgium right now. [13:07] humphrebc has arrived [13:07] Oh, has he? [13:07] yep [13:07] He IMed me when he got on the plane, but I haven't heard from him since. [13:07] http://twitter.com/humphreybc/status/13663010067 [13:08] Cool. I haven't checked my twitter feed yet today. [13:09] you haven't updated it in a while [13:10] as in 6 months [13:10] :) [13:10] heh.. yeah, I don't really tweet. [13:10] I don't blog much either. [13:10] I should write more. :) [13:12] godbyk: the reason i switched to twitter was because my family were worried i'd died because i stopped blogging. twitter is lazy blogging, only takes a couple seconds. [13:12] I'm going to get a glass of water and then sort through all these UDS sessions and figure out which ones I should attend (remotely) and when I can actually sleep. [13:13] synergetic: That's true. I have trouble compressing my thoughts into 140 characters. My IRC messages are often longer than that! ;-) [13:13] the only thing that makes me post at all is to let my family overseas know what i'm up to [13:13] My friends do tend to panic when they see I've been offline for a day or more, though. [13:13] godbyk: ah, but it forces you to revise until you are able to explain yourself succinctly [13:13] which is REALLY good practice [13:13] And my mother discovered Skype last week. Not a particularly good thing for me. :) [13:14] once upon a time you had telegrams and you had to get to the point. these are modern day public telegrams ^_^ [13:14] synergetic: True, though I suspect in most cases it causes people to instead simply dismiss the thought of posting anything of substance. [13:15] ooooh tough. why my mum discovered instant messenger, she'd message me, then again less than 30 seconds later "are you there???" and less than 5 minutes later she'd be typing caps lock about how i'm ignoring her and don't i want to talk to her [13:15] when i was in the other room cooking dinner or something [13:15] synergetic: Almost! If you had to pay per tweet, there'd be a lot less noise and I'd be more likely to read my twitter feed more than once a week. [13:15] i've heard twitter called the UDP of social media [13:16] Ha! My mom does that with phone calls, though I've weened her off of that. Now she's discovered Skype (with the webcam, no less!), so she's calling all the time again. [13:16] yeah. well. i have friends who follow me but who i don't follow because they tweet too much. and i'd tell them that i wasn't going to follow them for that reason and to not take it personally [13:16] dutchie: So true! [13:16] godbyk, i closed the website's translations [13:16] https://translations.launchpad.net/ubuntu-manual/lucid-e1/+pots/ubu-man-website-translations [13:17] mm yeah. my mum does'nt call me very much. but thank you for reminding me, it's mother's day! i should call her! [13:17] daker: That closed the manual's translations, too. [13:17] ah o.O [13:18] I spoke to my mom yesterday. She asked me (again) about Netflix and Hulu. [13:18] i just set it for the website template [13:18] Oh, and scanners. [13:18] daker: it affects the entire series, apparently. [13:18] daker: https://translations.launchpad.net/ubuntu-manual/ [13:20] should i set it as Open ? [13:23] I think it was Structured before. [13:24] https://translations.launchpad.net/ubuntu-manual/lucid-e1/+templates [13:24] now it's inactive [13:26] daker: Okay. You may want to email the list and let them know that if they want to translate the website to go to https://translations.launchpad.net/ubuntu-manual-website [13:27] oki [13:28] Hey, dutchie: https://bugs.launchpad.net/ubuntu-manual/+bug/577739 [13:28] i blame l[p [13:28] -[ [13:28] dutchie: Error: Missing "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. [13:29] hmm [13:29] I'm guessing it's po4a that should set it. [13:29] but I'm not really sure. [13:30] po4a has set Project-Id-Version containing PACKAGE and VERSION [13:31] * godbyk installs gtranslator [13:36] godbyk: my mum just keeps telling me that skype is broken - so i have to (over the phone) walk through how to fix it. and then it works and then she "breaks" the same way the next time. it's humourous to a point, but only to a point. skype needs a manual. [13:36] how does she break skype? [13:36] I think lots of things need manuals. I share the sentiments as a lot of the people mentioned in that article you linked me to. [13:37] * godbyk gives up on gtranslator because the interface is so absolutely miserable. [13:38] she breaks it by switching into the phone tab rather than the skype tab - so she keeps trying to call my mobile phone rather than skype [13:38] i because it's still a big greet CALL button so she doesn't notice the change [13:38] are I see [13:38] one says CALL PHONE and the other just says CALL so i'm not surprised she gets confused [13:39] makes sense [13:39] but she doesn't have skype credit so she gets upset that skype won't let her call [13:40] godbyk: my grandfather buys the manual for everything. it's really funny because he's got just stacks of books on how to use windows, how to use google, how to use a digital camera, everything. [13:40] and he just pores over them - i can't switch him to ubuntu until there's a comprehensive manual [13:41] aha :) [13:41] synergetic: Ah, yeah. I have a grandfather who's the same way. Unfortunately, most of the manuals he buys are absolutely abysmal. [13:41] godbyk: oh, they ARE horrible, i've had a thumb through them. but he's so damned independent that he won't ask me or my brother for suggestions [13:42] he'll just get whatever comes up first on amazon [13:42] has the skype gui been made open yet? [13:42] Ugh. My grandmother (on the other side of my family) is using Ubuntu. She hasn't had too many issues, but I never hear from her when she has problems. [13:43] I have to hear from other family members that she was complaining that this or that wasn't working. [13:43] I've finally got my mom trained that I'll help her with her Ubuntu issues, but she's on her own with her Windows problems (as I haven't used it in so long, I don't know the answers anymore). [13:44] yeah. only person in my family who uses linux is me - my dad is stopped by the skypephone he has which has windows-only drivers but he gives CDs to everyone [13:45] my mum refuses to switch because she's had XP since it was new and she doesn't want to learn something new when what she's got serves her fine [13:45] I have my mother, grandmother, and brother using it. (Though my brother doesn't use his computer at all at the moment.) [13:45] and my grandfather is determined to having the newest version of whatever MS produce, because that's what most people use [13:45] My mom has a netbook with windows on it, too, now. [13:46] I got those family members to switch because I told them that while I was happy to help with with any linux issues they had, I couldn't help them with windows problems anymore. [13:46] yeah. my brother did ubuntu for a month or so but then went back to windows for his games [13:47] Needless to say, there's been a drastic reduction in support calls over the past couple years. :) [13:47] that's i think why my family don't ask me about computers - because i just tell them to switch to ubuntu. [13:47] ^_^ [13:47] Games! My other brother, who would otherwise probably like linux, has stuck with windows solely because he plays so many games. [13:47] exactly [13:48] And my two sisters are using windows because they have to run software for school and work that's windows-only. [13:48] ah! the mum of one of my old flatmates got her a 10" netbook, put ubuntu NBR on it for her, set it up so she could quickly get to email and internet [13:48] she was over the moon, because it was such a simple interface [13:48] nice! [13:48] I haven't played with ubuntu unr yet, but the idea of it appeals to me. [13:49] it's easier to give people who've really used computers much ubuntu to start with - it's the people who use windows just because that's what they've always used that're hard to convert [13:49] i'm using NBR with the computer i'm on just now (EeePC901) and it's absolutely fabulous [13:49] Agreed. [13:49] Usually I've had those folks switch to Ubuntu when they're having so many issues with Windows. [13:49] my bf uses NBR on his media computer because it's easy to see from a distance [13:50] And usually they just need Internet a and email. [13:50] yeah - i switched to kubuntu because i'd had so many issues with windows, then switched to ubuntu when they went to KDE4 - i really hate KDE4 [13:51] I used to use KDE eons ago, but there were so many preferences and options scattered all over the place. I came back to GNOME. [13:51] I haven't tried KDE4. [13:52] the friends i convert at uni switch because they use open office anyway and then they can also get things like Xournal and can more easily get at things on campus through SSH and FTP [13:53] Cool. [13:53] uhm. i know there was some talk about wanting to do quick getting-started guides - i have a bi-fold flier i made for an install day last year if it would help the cause any [13:54] i should have it kicking around here -somewhere- [13:54] Cool. Yeah, I'm not sure what all humphreybc has in mind for those. [13:54] I think we're also going to try to work with the learning team on some of that material, too. [13:55] cool. it's a PDF if there's anywhere you guys put communal documents? or i can just put it in my dropbox public folder or something in the meantime [13:56] maybe for inspiration if nothing else, somewhere to start from? [13:56] we don't have a repository yet for it, but if you'd like to put it in your dropbox or email it to me I can put it up someplace. [13:57] righto ^_^ [14:12] godbyk: took two tries but i sent it >.< touchpad fail [14:12] * ubuntujenkins succes ccicons has built [14:14] synergetic: received it. thanks! [14:14] ubuntujenkins: awesome! how are the other packages doing? [14:15] I am unistalling everything to work out what gets pulled in but is not in the ppa.That way i can wok out what else i need to add to the ppa for pre lucid ubuntu [14:19] I think we will have a ppa by next week end easily [14:19] cool [14:41] hello [14:43] Hey, c7p [14:44] hello c7p [14:49] godbyk, someone want to join https://launchpad.net/~ubuntu-manual-website/+member/xuacusk8 [14:49] daker: okay [14:50] daker: He's the Asturian translation editor. [14:54] godbyk, try to see that https://translations.launchpad.net/ubuntu-manual-website/trunk/+lang/fr [14:54] there are 2 templates [14:54] ????!!! [14:55] I see two templates, yeah. [14:55] there is only one pot file in the branch [14:55] daker: there are two: [14:55] ./test/includes/languages/ubu_man_website_translations.pot [14:55] ./main/includes/languages/ubu_man_website_translations.pot [14:55] ah [14:58] how to remove the first one ? [14:59] https://translations.launchpad.net/ubuntu-manual-website/trunk/+lang/fr [14:59] should i remove the pot file ? [14:59] I think removing the pot file would do it, but I'm not sure. [15:00] remember that our current live site is the test/ directory. [15:01] i will remove the one in the main folder [15:01] okay [15:08] who has lucid with no latex/texlive installed? I would like some other testers for the ppa [15:10] my lucid box has the upstream TL2009 installed. [15:12] I might do an install on my external disk to test [15:14] virtual machine? [15:14] they are always slow, but that may mean that i can do work at the same time. [15:18] godbyk, i can't setup the Priority [15:18] https://translations.launchpad.net/ubuntu-manual-website/trunk/+templates [15:20] daker: what do you want to set it to? [15:20] the inactive translation has 0 priority, and the other translation has priority 1. [15:22] yes that's enough ? [15:23] Probably. [15:23] I think the priorities only matter in relation to the other translations for that project. [15:44] Wow the branch history files is big than the website it self [15:47] what has anyone listened to todays full circle podcast? they said we released a quick start guide [15:50] ubuntujenkins: yeah, I noticed. maybe the docs team released something? [15:51] I am not aware of the doc team releasing something like that. [15:52] neither am I [15:56] for the dummy package ubuntu-manual-tex do we want bzr as a dependency? That way people have it so they can pull the branch. Or it is not relevant to the latex stuff and we don't want it. The two arguments thoughts godbyk ? [15:57] hmm.. [15:57] maybe have the -tex package depend only on tex-related stuff [15:57] we could have another dummy package for our tools like bzr, I guess. [15:58] that was more my thoughts but what other tools do we need? [15:58] po4a [15:59] thats need to make the manual so i put it in the -tex dummy package [16:22] godbyk: I am right in thinking this suggests i am missing a font? http://paste.ubuntu.com/430607 [16:23] I can't find anything in synaptic called pzdr or pzd [16:23] ubuntujenkins: well, it means that tex can't find it. [16:23] ubuntujenkins: it comes from the zapfding package, I think. [16:24] or the collection-fontsrecommended collection of packages. [16:25] I am trying texlive-fonts-recomended [16:28] yey as soon as i add texlive-fonts-recomended to the dummy package we have lucid package support [16:29] cool [16:30] I should at it to a ppa under the project rather than under my name [16:31] yeah, that'd probably be better. [16:31] I assume you have access/rights to do so? [16:32] there is only one way to find out [16:35] hmm well i can't make a new ppa the existing one is called ppa:ubuntu-manual/beginnersmanual . I need a team admin to make a new one [16:36] I can delete the ppa that is there but not make one by the looks of it [16:36] ubuntujenkins: let me see what I can do [16:38] where would I go to create a ppa? [16:38] on this page https://edge.launchpad.net/~ubuntu-manual under the existing ppa there should be an "add ppa option" I think ben is the only one who can [16:38] he appears to be the only group admin [16:39] Should I just re-enable the Ubuntu Manual PPA that's there? [16:40] I can do that but the name of is "beginnersmanual" is that ok for the tex stuff [16:40] ew. [16:40] nah. [16:40] there are other admins, though. [16:41] excactly [16:41] looks like it doesn't include me, however. [16:41] dutchie: you around? [16:42] godbyk: look at https://edge.launchpad.net/~ubuntu-manual/+members and click the status heading only ben is admin [16:42] godbyk: yes? [16:42] dutchie: hook me up with some admin rights on the https://edge.launchpad.net/~ubuntu-manual/+members and/or help us set up a new ppa [16:42] dutchie: are you able to make a new ppa on https://edge.launchpad.net/~ubuntu-manual ? [16:43] ubuntujenkins: he's just the the only admin on that page. [16:43] it's only sorting that single page of the list. [16:43] are makes sense [16:45] hmm [16:46] can't work out how to adminify people [16:47] dutchie: on the page https://edge.launchpad.net/~ubuntu-manual/+members you should see a yellow circle with a ! click that and you can change it then [16:47] https://edge.launchpad.net/~ubuntu-manual/+member/godbyk [16:48] i can't view it [16:48] from here: https://help.launchpad.net/Teams/CreatingAndRunning it sounds like only owners can make admins. [16:48] heh.. I can't view that page either. [16:48] Now I want to know what it says about me! :) [17:04] once we have a ppa I have made a script to transfer the packages [17:05] dutchie: are you able to make a ppa please? [17:05] sure [17:05] what should we call it? [17:05] godbyk: ^ [17:06] what all is going to be in the ppa? [17:06] same stuff as in https://edge.launchpad.net/~ubuntujenkins/+archive/ubuntu-manual/+packages and anything needed for earlier versions of ubuntu [17:07] should we just call it ubuntu-manual and put all our packages in there? what are the benefits of having multiple PPAs? [17:08] still needs a name [17:08] We could have a release ppa and a daily build one if we wished [17:08] and a display name [17:09] how about ubuntu-manual for the release versions of stuff and ubuntu-manual-daily for the daily builds of things? [17:09] what about the tex stuff? I am not building that daily [17:09] put that in ubtuntu-manual [17:09] yeah, just drop that in ubuntu-manual, I'd think. [17:10] unless you think it should be segregated into its own ppa. [17:10] and the display name? [17:10] Ubuntu Manual Releases and Ubuntu Manual Daily Builds? [17:11] Ubuntu Manual Releases is fine don't worry about dailies yet [17:11] sounds like a plan. [17:11] we have nothing for the dailies [17:11] I figure we can always change it up later if we really need to [17:11] yeah, nothing for dailies yet. [17:11] yep [17:11] though I could start building daily PDFs. [17:12] not sure if it'd be helpful for anyone or not. [17:12] not helpfull untill i have made the ubuntu-manual program [17:12] https://edge.launchpad.net/~ubuntu-manual/+archive/ubuntu-manual [17:13] thansk dutchie [17:13] np [17:16] and i am able to upload stuff to it [17:17] cool [19:01] how are we going to promote / suggest the ppa? [19:01] I should mail the list i assume [19:01] hey SunK how are you? [19:03] I'd mail the list and have a couple people test it out. [19:03] what is the correct way to remove upstream tex if they have it? [19:04] I've been just suggesting 'apt-get remove texlive-*' but that leaves behind a few things in packages named otherwise. [19:04] shouldn't our packages have larger version numbers, so it'd just be an upgrade? [19:05] I ment the one the website curently suggests. I will double check but it think i set the version numbers higher. [19:05] there is a thing on ppa versioning [19:06] *the way of instaling tex that the website suggests [19:06] https://help.launchpad.net/Packaging/PPA/BuildingASourcePackage#versioning [19:08] ok i need to change the version numbers [19:09] I finish my matlab then i will fix it === Daker is now known as daker [19:23] ping godbyk [19:26] IlyaHaykinson: pong [19:27] i wanted to start the discussion about file formats for maverick [19:27] sure [19:27] i've been looking at DocBook, and i think it could do everything we need, especially if we greatly limit the subset of allowed tags [19:28] from there, we can write a converter into our LaTeX commands [19:28] for the web site, we can make the system manipulate a database of docbook snippets [19:29] be with you in just a moment. [19:29] ok [19:29] skyping with the parents. [19:29] :-/ [19:41] (still on skype) daker forwarded this link to me the other day: http://doc-book.sourceforge.net/homepage/ [19:42] okay, off skype now. [19:42] brt [19:42] I think the docbook wiki site looks like what we want (heavily modified, of course) [19:43] we'd need to give it a UI overhaul, change it so the wiki is more moderated [19:43] but it stores the underlying data in docbook format. [19:43] you might also look at mallard: http://projectmallard.org/ [19:43] that's what the gnome docs are headed to. [19:44] in any case, I think we will want an online GUI editor so that no one has to mess with the underlying markup language. [19:45] tinymce ? [19:46] http://tinymce.moxiecode.com/ [19:46] I'll look at it. [19:46] brb [19:49] back now. [19:49] am back [19:49] godbyk, what about tinymce ? [19:50] daker: How does it store the text? html? [19:50] there 2 choices plain text(with tags) or html [19:50] are* [19:51] we'd have to modify it to use some custom tags, too. [19:51] (for screenshots, for instance.) [19:51] we can [19:53] i think mallard is a little more immature than docbook, which has a lot more semantic markup possibilities [19:53] Once we have an idea of what we're doing, I think we're going to have to get some more people to help with this new website. daker may go crazy trying to do it all by himself. :) [19:53] mallard is younger than docbook, that's for sure. [19:53] godbyk, :D [19:54] but since all the gnome docs (and presumably the ubuntu docs) are headed that direction, if it's robust enough for our needs, it would make interoperability a bit easier. [19:54] which one are the docs team are using? mallard or docbook? [19:54] most gnome docs are in docbook though [19:54] (though it would hopefully not be too difficult to translate the mallard xml to and from the docbook xml) [19:54] for example, here's evolution: http://svn.gnome.org/viewvc/evolution/trunk/help/C/evolution.xml?revision=37459&view=markup [19:54] well, mallard is what yelp will be using. [19:55] so when you go to System > Help and Support.. that stuff will (eventually) be mallard-based. [19:55] and mallard claims to specifically be for topic-oriented (non-linear) help [19:55] yeah, that's true. [19:55] honestly, mallard seems like it's not much better than authoring in a subset of HTML [19:56] there are very few tags, which means that the semantic meaning is often lost [19:56] fair enough. [19:56] whereas authoring in docbook appears to be more complex because of the hundreds of tags, but on the other hand carries over a lot of meaning [19:56] does docbook handle everything we need? [19:56] so far as far as i can see, yes. [19:56] okay. [19:56] maybe except for screenshotTODO [19:56] and not sure about todo itself [19:57] surely it's not too difficult to add tags... [19:57] but it even has markup equivalent to \menu{foo\thenbar} [19:57] nice. [19:57] well, adding tags -- of course not hard. though if possible, we should strive to stick with the standard [19:58] if we use docbook as our base, we'll need to make sure we have great tools to hide that underlying markup. [19:58] i think a lot of it could be hidden by allowing text->docbook transformation [19:58] so we have some control over the tags used (for consistency) and make it as easy as possible for people to write and edit content without getting caught up in all the markup. [19:59] going from docbook to tex is easy enough.. an xslt could do that. [19:59] yeah; writing markup is certainly not as friendly as writing LaTeX, even [19:59] we'll just have to add in an editing step on the generated .tex file to make sure everything looks great typographically. [19:59] (though that step is already there. [19:59] nod. [19:59] I did it with the english version [19:59] and we're going to have to do it for the translated versions, too.) [20:01] team [20:01] need help [20:01] how to work on translating the website? [20:01] Hey, shrini. What's up? [20:01] godbyk: hi [20:01] how are you? [20:02] took so long to hang here [20:02] :-) [20:02] shrini: Head over to https://translations.launchpad.net/ubuntu-manual-website, select your language. [20:03] godbyk: there's a question of storage, too; one way is to have each "page" on the site be a file in bazaar; we'd allow editing on a per-paragraph basis ( in docbook) and just make the change directly to the underlying xml file [20:03] then queue up the change for review [20:03] another option is to have a database of paragraph / step snippets [20:04] IlyaHaykinson: I think we may want to have a database sitting in the someplace to assist with searching/indexing. [20:04] and have _all_ work use this system, perhaps generating a complete xml file as a result [20:04] right. [20:04] godbyk: thanks [20:04] the downside, though, is that when authoring we'd need a lot more tools [20:04] got it [20:04] to allow writing multiple paragraphs at once [20:04] shrini: no problem. let us know if you have any other questions or run into problems. [20:04] godbyk: do you have experience with xslt? [20:05] sure godbyk. thanks a lot [20:05] thorwil: a tiny, tiny bit. [20:05] thorwil: at this point, I'd have to relearn it all, I think. it's been quite a while. [20:05] got my brother to work on translating to tamil [20:05] he finished exams [20:05] shrini: awesome! [20:05] i've done a bunch of xslt work a while ago too... [20:05] been ~4 years though, probably [20:06] wow.. it's probably been about that long for me, too. [20:06] ok. otherwise i would have to warn you that xslt is a ... thing of its own [20:06] all my transforms were pretty simple, too, as I recall. [20:06] * godbyk makes a note that thorwil is now the resident xslt expert. [20:07] meh. i was doing transforms from XML into a text-based EDI data format (ANSI 837), yuck. [20:07] oh, [20:07] godbyk: i only looked at it long enough to become frustrated [20:08] IlyaHaykinson: I think the database-backend approach may give us more benefits, and we can always generate xml files from that. [20:08] File backends almost never turn out to be a good idea. [20:08] especially when they're as linear as xml is. [20:09] yes, though we would need to then accommodate a lot of different types of data [20:09] i.e. paragraphs, lists of steps [20:09] where each would be a stored unit [20:09] i suppose each could then be a unit of translation, too [20:11] right. [20:12] if we had our own translation stuff, we wouldn't have to worry about launchpad's shortcomings (just our own). [20:12] but it'd allow us to do nicer string manipulations (comparisons, parsing, etc.) [20:13] we should use a unit tested no-sql distributed cloud storage solution with an enterprise level object-node mapper with a collaborative realtime editing ajay/comet low-latency frontend [20:14] thorwil_: Whatever, Mr. XSLT! [20:15] lol [20:16] we could just use pootle [20:16] especially so if we write the new system on top of Django or something like that [20:16] Something fancy we could do fairly easily: Pull all the translated strings from the software we're documenting. When a translator is working on the section about F-Spot, for instance, we could auto-translate the \menu, \button, and other GUI elements using the strings from the F-Spot translations. [20:16] yeah, pootle looks fairly comparable to launchpad's translations stuff. [20:16] nifty idea. [20:17] thorwil_: true, could use a nosql store, though perhaps then we'd lose some of the indexing / searching approaches [20:17] instead of suggesting paragraphs at a time, we could actually do sentence segmentation and suggest each sentence. this would help when we, say, combine two paragraphs into one para. [20:17] that said, both mongodb and couchdb have some indexing support [20:18] i think paragraphs are more logical units, though. [20:18] we could also show the diff between the suggested fuzzy string's source and the source of the current to-be-translated string, so the translator can see what change she'd have to make in her translation. [20:18] IlyaHaykinson: it was just a buzz word collection :/ [20:18] there are so many sentences... [20:18] thorwil_: ;-) [20:18] I think paragraphs are a good unit, yes. [20:19] the sentence segmentation is primarily to help with fuzzy translation string matching. [20:19] well, we could just segment out things that are marked up [20:19] to reduce the amount of time the translators spend retranslating nearly identical text over and over again [20:19] sentences are not useful as unit when it comes to translations [20:19] i.e. menu names would be in some sort of tags. [20:19] it sucks that they currently have to retranslate the entire paragraph when we, say, add or remove a single comma. [20:20] well, we could add change detection. [20:20] yeah. [20:20] so that if it's a minor change, we don't flag the para for retranslation [20:20] perhaps merely for review [20:20] basically, any small thing we can do to help the translators would go a long way. [20:20] also, we would match units within the paragraph [20:20] and highlight the change, so it's obvious. [20:21] yes, that's what I was referring to with the sentence-segmentation talk. [20:21] so if the text is "....blah..." then the could have an id attribute [20:21] and then we would auto-translate those based on the previous translation [20:21] it's rather obvious what you could do to ease translations, but it gets very complicated regarding implementation soon [20:21] but i think this could work at the para level, not just sentences [20:22] IlyaHaykinson: I agree. [20:22] thorwil_: perhaps. but anything worth doing is complicated. :) [20:22] In general, software should be a heck of a lot smarter than it currently is. [20:22] As programmers, we too often take the easy way out. [20:23] Sometimes citing lack of information/data to make the right decision (instead present an option to the user). [20:23] Sometimes just claiming it's too hard or it'll take too long. [20:23] But in the end, it all results in the same thing: dumb software. [20:23] godbyk's suddenly come over all philosophical [20:23] [20:24] godbyk: my fingers are itching to have a go with smalltalk/seaside or scala/lift, but what i have in mind would be a job for 2 or 3 competent developers, at least [20:24] I hear ya. [20:24] I think if we're setting out to do this thing, we'd best do it well. [20:25] Otherwise, we'll be just another website with documentation on it. [20:25] and all the technological choices that look good to me also cut out a lot of people who would run with php, rails or django [20:25] Aside from failing to attract contributors (developers, editors, authors, translators, et al.), it's also *boring*. :) [20:26] i wouldn't mind having a go with it being in, say, django. [20:26] though yes, a scala-based system could be nifty [20:26] thorwil_: We're also currently hampered a bit by my web host. We can't do anything CPU intensive. And any non-standard software means I have to compile it myself and let it reside in my home directory. (I don't have root access.) [20:26] It we start getting too crazy, we'll need to find another hosting solution. [20:26] vserver at least :/ [20:27] Right. [20:27] I'm willing to manage a vserver, but we'd have to figure out how to pay hosting fees. [20:27] i don't think this would be too big of a problem [20:27] i have a machine we could use, for some time at the very least [20:27] IlyaHaykinson: duly noted. [20:27] one can have one for very few euros/month, but i have no clue about actual needs (cpu/ram/space) [20:27] datacenter hosted; a bit old but usable. [20:27] and it's a phyiscal box [20:28] physical, even [20:28] nice! [20:28] So are we fairly convinced that docbook is the way to go for the underlying markup of the master doc? [20:29] also, after benjamin builds connections with the ubuntu folks, he should be able to swing us a canonical-owned server or something like that [20:29] godbyk: i think that if xml is of advantage, docbook is the way [20:29] IlyaHaykinson: That'd be quite nice. [20:29] i think so far, docbook seems the best [20:29] Sounds like a plan then. [20:30] So what's next? [20:30] apt-get install doc-book [20:30] :) [20:30] heh [20:30] not sure thats the right name [20:30] no hyphen [20:30] and there are a ton of docbook* packages. [20:31] oh, cute. there's a docbookwiki package, too. [20:31] theres no docbook-full [20:31] * thorwil_ goes back to drawing [20:32] hmm so how do we write docbook? I only learnt latex since starting the project [20:33] ubuntujenkins: well, docbook is somewhat akin to html. [20:33] but more semantic markup. [20:34] I don't think i will get time to help write this release but I would like to know how to do it so that i can help others [20:34] http://svn.gnome.org/viewvc/evolution/trunk/help/C/evolution.xml?revision=37459&view=markup [20:34] as an example of a complete manual in docbook [20:34] http://www.docbook.org/tdg5/en/html/ch02.html for a more structured introduction [20:35] wow that looks harder to lean than latex would our translators have to learn it? [20:37] ubuntujenkins: that's why we're gonna have to write a web-based editor that hides all the markup and has pretty buttons to click on. [20:37] kinda like: http://tinymce.moxiecode.com/examples/full.php [20:37] but with semantic markup instead of low-level formatting. [20:38] sounds good to me i wasn't really reading irc much [20:38] * ubuntujenkins hmm I have just filled our ppa to its limit [20:39] uh oh [20:40] didn't they just increase the limits the other day, too? :) [20:40] its all good, I will just remove them and then add them from my ppa. [20:41] its they way that i am changing the version number that is causing it [20:43] * ubuntujenkins now to wait untill my package deletion request is done [20:45] actually i'm not sure we need to write a web editor [20:45] i think we can allow just text-bsaed entry [20:48] be back in an hour [22:34] godbyk: ping [22:34] IlyaHaykinson: pong [22:35] to continue the discussion about docbook... [22:35] sure [22:35] i think we have two different modes of editing [22:35] one is for writing lots of content (the style of writing/editing we had for lucid-e1) [22:35] one is for making corrections on the website, typically one para at a time [22:36] the second one is better i think [22:36] well, i think both are necessary [22:36] at different times. [22:36] Right. [22:36] if we have small units on which we allow people to work (paragraphs for example), we allow the second use case. [22:36] but make the first one rather tough [22:36] because then we need the ability to insert, append, prepend sections [22:36] and to select section types [22:37] and rearrange sections [22:37] etc. [22:37] Yep [22:37] and every long piece of text becomes a pain to write [22:37] on the other hand, if you treat the entire document as one big unit [22:37] For writing content, it should be a free-form editor that allows for multiple paragraphs. [22:37] The finer details and translations can be handled per paragraph, I think. [22:37] then writing lots of text becomes easy, but smaller scale editing is a (text processing) pain [22:37] so how about this... [22:38] we store data at the paragraph (or other similarly sized section) level [22:38] but we allow a one-time import of large amounts of text which we parse into paragraph units [22:38] we'd probably also need a high level interface for shifting content around [22:38] right, and then provide some ability to rearrange stuff [22:38] and modify sections/chapters [22:39] and also a way to replace several paragraphs at once with another larger-file reimport [22:39] That sounds like it'd work, IlyaHaykinson [22:39] we'll likely need a few different types of "units" [22:39] paragraphs, steps, screenshots, terminal examples [22:39] so ok. next problem. [22:40] i think it's way over-ambitious to create a tinymce-like editor for docbook [22:40] i'd just come out and say that it won't happen. [22:40] so, what about a slightly-content-aware plain text editor [22:40] so basically, let's say you're editing a paragraph that looks like the following: [22:40] we have enough over-ambition on this team ;) [22:41] "abc def FileSave ghi jkl" [22:41] when you try to edit it, we just show you "abc def File Save ghi jkl" [22:41] I looked at the TinyMCE and they do have something that allows it to output bbcode in addition to html. [22:41] then, we look at what you've changed. if you change things before or after the nodes, then we allow the change to go through [22:41] So it may be easy to have it dump docbook, too. [22:41] I'm not sure. [22:42] if you change File or Save then we match up the tokens to the original ones [22:42] if we can [22:42] and if we don't know what happened, we turn it into plain text [22:42] _BUT_, at the same time, each change is stored as a revision [22:42] that gets a bit tricky, because then we have to track things like' should this empty element exist? or did they mean to remove it? [22:42] right. [22:42] but we provide a level of perceived risk with each changeset [22:43] here's the tinymce bbcode example: http://tinymce.moxiecode.com/examples/example_09.php# [22:43] so let's say we know for sure that "abc" changed to "zzz" but nothing else got touched [22:43] in that case, we allow the change to take place [22:43] I'll have to download the tinymce code and look at it to see how easy/hard it'd be to bend it to our will. [22:43] however, if the change affected the original layout, we mark it as high risk and require docbook-level recoding [22:44] so have human-editor intervention? [22:44] well, the thing is, i don't know if we should allow "the public" to write pure docbook code [22:44] for handling markup? [22:44] i would rather take changes that have no markup [22:44] and re-mark it up [22:44] or, as i was suggesting, apply some smarts and in some cases re-mark it up automatically if we can be sure that the changes didn't affect the markup [22:44] if they can't do it easily (gui/wysiwyg), then yeah, it's probably better if a human editor intervenes with the docbook markup [22:45] right. [22:45] and we can always allow an option, while editing, to edit the raw markup code (for those who know it) [22:45] this way the general public can submit changes without knowing any of the complexity of docbook [22:45] which is the biggest problem with docbook, imho. [22:46] I agree. [22:46] (there are so many tags that while semantic knowledge is retained, human editing becomes tough) [22:46] that's why I was hoping to avoid docbook for as long as possible. :) [22:46] nod. but if we can make two rules [22:46] 1) we only use a subset of docbook as per a style guide, and [22:46] 2) mere mortals (non-UMP members) are not expected to know docbook [22:47] then i think we'll be able to use the full power of docbook [22:47] without incurring the cost of complexity for casual editors [22:47] over time, as we have more development time, maybe we can allow some super-simple markup in the editor [22:47] but initially we'd just take raw text [22:48] that sounds like a plan. [22:48] I've looked at the bbcode plugin for tinymce [22:48] and it's just doing a find/replace on the generated html to convert it to bbcode tags. [22:49] nod. i figured it'd be something like that. [22:49] tinymce is really oriented at HTML, from what i can see [22:49] I'll also have to look at how easy/difficult it'd be to add our own semantic markup buttons to the editor. [22:49] yeah, it is. [22:50] ok, let me put together a quick document of this idea -- i'll mail something out later tonight [22:50] i think we can make it work. [22:50] okay, sounds good. [22:51] I think we should definitely jump in and get started on this stuff -- especially if we're going to try to pull it off in time for maverick. [22:51] something else I'd like to see in our editor (plain text or otherwise) is some as-you-type checks against our style guide rules. [22:51] check for spelling, grammar, usage, etc. as much as possible. [22:53] well, perhaps instead of as-you-type, just at time or whatnot [22:53] i.e. a quick check once you send the snippet to the backend [22:53] Does anyone know, the UDS schedule (for example: http://summit.ubuntu.com/uds-m/track/community/) -- are those times UTC or Belgium local? [22:53] Sure. [22:53] Just something that can be done at the time of editing/writing. [22:54] So we can avoid revisiting that text if possible. [22:54] i'm sure they're local. brussels is UTC+2, so otherwise the sessions would start at 7am [22:54] er, 11am. [22:54] That's what I figured. [22:56] well, this week will be fun for me, then, I guess. :) [22:56] you're going to stay up to participate by irc? [22:56] I'm gonna try. [22:56] i don't think they have live streaming... [22:56] There are a few sessions I want to hit up. [22:56] I thought they were supposed to have live streaming. [22:57] I think that they have live streaming icecast.ubuntu.com iirc [22:57] Some UX sessions and some sessions related to the manual (docs and translators). [22:58] I don't think i can make many due to lots of lectures and tones of uni work :( [22:58] AHA [22:58] i know what happened to the translations [22:59] what? [22:59] dutchie: do tell! [22:59] the translation focus got set to lucid-e2 a little while back, and they must have been done on that [22:59] what?! [22:59] then when I switched it back to lucid-e1, they didn't get backported [22:59] or they did, but only as suggestinos [22:59] +1 to godbyk [22:59] Factoid '1 to godbyk' not found [22:59] don't know how it happened [23:00] i was looking at them for some reaso [23:00] ubuntujenkins, the link doesn't work [23:00] weird. [23:00] I wonder how/when it got switched to lucid-e2. [23:00] most of the translators reported that the flood of suggestions occurred just a day or two ago. [23:00] when did you change the focus back to e1, dutchie? [23:00] week or two ago [23:00] hmm [23:01] maybe it's not [23:01] :( [23:01] daker: I just rememberd that I don't actually know it have a look on the wiki somewhere [23:01] (Getting more hate mail from translators.) [23:01] yep seen it [23:01] I haven't got a useful response on #launchpad yet. [23:02] wgrant wrote me back and said he didn't know anything about the translation side of launchpad. [23:02] other than that, no responses. [23:04] The translations look really messed up right now: https://translations.launchpad.net/ubuntu-manual [23:04] Sort by status. [23:04] I think there used to be a lot more of them that were farther along than that previously. [23:04] I think that it would be easier to paste the po file in etherpad and have the translators translate in that [23:05] I would agree with that godbyk [23:05] wtf? [23:05] this is getting frustrating. [23:06] I'll ping #launchpad with our problem again, but I'm not holding my breath. [23:06] stupid uds [23:07] anyway, i have an early morning [23:07] night [23:07] We should send humpreybc out to track the launchpad guys down and demand answers. [23:07] g'night, dutchie [23:07] night dutchie [23:08] godbyk, yes [23:09] godbyk: i'm going to work on a doc to describe the docbook proposal; do you want to try to craft a mini-sample of some short chapter (or other section) from our current manual in docbook format? [23:09] it may help to have an example [23:09] IlyaHaykinson: I can try, yeah. [23:10] if you have time, that'd be awesome. [23:10] I'm just glancing through the pootle site right now... Wondering if we wouldn't be better off cutting our losses and migrating translations to our own installation of that or something. [23:11] i looked at pootle earlier. seemed like it can totally do our work. [23:11] the creative commons uses it [23:12] http://translate.creativecommons.org/ [23:13] cool [23:13] the interface looks nice [23:13] I think it'll take a bit of time to get it set up properly. [23:14] What would we lose by ditching Launchpad for translations? [23:14] auth [23:14] Less tightly-integrated bzr? [23:14] auth [23:14] karma [23:14] yeah, but, well, screw karma [23:14] shared strings with other projects (though we don't have any of those, I bet) [23:15] if it means getting rid of these headaches [23:15] no doubt [23:15] hm, well, we can always query launchpad for other translations [23:15] we just couldn't be a source for them [23:15] but, well, neither is any of the gnome stuff [23:15] shared strings isn't much of an argument as I could probably seed those strings for pootle anyway [23:17] nod [23:17] all right. I'll add that to my list of things to look into this week, I guess. [23:17] lemme see if I can find the docbook tags and whatnot and convert some tex for ya [23:22] http://www.docbook.org/tdg5/en/html/ch02.html is a good start [23:23] thanks [23:23] I have entirely too many windows open. [23:23] I think I'm going to cheat and just create a couple more workspaces. :) [23:24] IlyaHaykinson: are you wanting just some interior text or do you want the document structure and stuff, too? [23:25] more the text, i think [23:25] IlyaHaykinson, http://www.writingopensource.com/forums/tech-and-tools/online-docbook-editor [23:26] https://edit.php.net/ [23:26] here is the branch http://cvs.php.net/viewvc.cgi/doc-editor/ [23:28] nod. looks unmaintained though [23:29] oh, daker: we can probably remove the vrac mirror now that things have settled down. we can add it back in later if we need to. [23:29] oki [23:30] hi! [23:30] hello xuacu [23:40] alright, gtg, bbl [23:40] night IlyaHaykinson [23:51] night all