18:31:37 <randomus`> #startmeeting Fedora Docs Workshop
18:31:37 <zodbot> Meeting started Thu Mar 26 18:31:37 2015 UTC.  The chair is randomus`. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:31:37 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:31:58 <randomus`> #chair yruseva
18:31:58 <zodbot> Current chairs: randomus` yruseva
18:33:21 <yruseva> um, wait, what?
18:33:32 <yruseva> randomus`? :)
18:33:48 <pbokoc> yeah, we decided you're the new docs project lead
18:33:51 <pbokoc> have fun
18:34:05 * randomus` runs away
18:34:39 <yruseva> haha I only opened xchat to check some network names to join from my personal laptop
18:34:52 <yruseva> well, um
18:34:59 <yruseva> thanks for the honour
18:35:04 <yruseva> free ice cream for everyone :D
18:35:27 <pbokoc> what would you have us do, oh dear leader?
18:35:42 <randomus`> best leader evah!
18:36:24 <pbokoc> https://theboeskool.files.wordpress.com/2014/12/kim-jong-un-7.jpg
18:39:07 <jsmith> All in favor of randomus` being named the Benevolent Doctator for Life, raise your hands and salute him!
18:39:47 <pbokoc> I've been doing that every time I see his name on IRC... you mean I'm the only one?!
18:40:11 <yruseva> pbokoc, we didn't want to tell you...
18:40:14 <jsmith> pbokoc: At least the only one willing to admit it in public...
18:40:59 <randomus`> you're compulsively saluting? this could be fun...
18:41:12 <yruseva> :D
18:41:13 <pbokoc> oh, no, I just raise my hands
18:41:31 <randomuser> oh, ok
18:41:40 <pbokoc> I mean, how do you even salute with both your hands raised
18:42:14 <randomuser> I raise both elbows whenever I see pbokoc online
18:42:57 * smccann thought randomus was the more interestic nick... very roman
18:43:11 <randomuser> http://media.giphy.com/media/KJmORsw9VYFQQ/giphy.gif
18:43:30 <pbokoc> :D
18:43:49 <randomuser> I assumed this happened in your office, all day
18:44:43 <pbokoc> no, we just have a giant ball we occasionally throw around... you know, the kind normal people sit on
18:45:35 <randomuser> that's fun too
18:46:12 <randomuser> so, did everyone get a chance to read the email I sent to the list about standardized metadata?
18:46:20 <randomuser> and further, did you care?
18:47:51 <pbokoc> oh, I missed that
18:48:32 <randomuser> there's even a glaring omission, just to help you get engaged
18:49:09 <pbokoc> ಠ_ಠ
18:50:37 <randomuser> jsmith or bcotton will probably pick it out away, so I feel the need to pretend it was deliberate
18:51:36 <pbokoc> oooh, I found it! I think Mr. Obvious Pseudonym is not a real person!
18:53:54 <randomuser> ehh... no
18:56:09 <pbokoc> so anyway, I'm not sure where you're going with that... where would we use the information?
18:56:34 <randomuser> any tools that format and present the documentation could use it
18:58:18 <pbokoc> right, so we wouldn't necessarily depend on publican to build the website, or we could, I don't know, package all docs and have some system to sort everything and create a splash page?
18:59:16 <randomuser> yes, exactly
19:00:06 <pbokoc> well, that sounds cool - and not too hard or time consuming to do, so why not
19:00:39 <randomuser> yeah, I already started building the 'some system'
19:00:44 <pbokoc> even though... I think the bigger a certain document is, the less useful the description would be
19:01:19 <pbokoc> if you have something like the System Administrator's Guide that has a few hundred pages and covers 30 wildly different topics...
19:01:41 <randomuser> having avenues for short articles should help scope creep, I would think
19:05:57 <pbokoc> true
19:06:56 <pbokoc> we'll need a definite list of acceptable values for a lot of fields though
19:07:28 <randomuser> indeed
19:09:19 <randomuser> especially the taxonomy stuff, or it would get out of control
19:10:11 <pbokoc> yeah, nothing better than having each tag used exactly once because there's a difference between "network" and "networking" :))
19:11:03 <randomuser> ohh... good point
19:11:23 <randomuser> I dunno that I want to make something that has user-visible tags, tbh
19:11:51 <randomuser> maybe just drop them in the html header, or use them for some mythical search feature
19:12:15 <pbokoc> I don't think there's any harm in that, but not much to be gained either
19:15:39 * randomuser groans
19:15:51 <randomuser> tags would be very difficult to control
19:16:25 <randomuser> you'd end up with a whole layer of request/approval for every new article
19:16:36 <pbokoc> I don't think they'd be more difficult than the taxonomy stuff... we'll need to automatically check that too, anyway
19:17:16 <randomuser> sure, but with the taxonomy stuff, we could probably sit here for an hour or two and knock out a draft that would cover almost everything
19:17:39 <randomuser> but tags could be a set of attributes specific to that one article
19:21:28 <pbokoc> I guess it would depend on how easy or hard it would be to add a new tag
19:21:43 <pbokoc> if it's easy, and the only barrier is that not everyone can do it, then we could add them on the go
19:21:54 <randomuser> I can't think of a way to do it without human involvement
19:22:19 <pbokoc> someone makes a new article, goes through existing tags, and finds that there aren't any existing tags specific enough - so they ask someone with access to create a new one
19:22:34 <pbokoc> but yeah, I can't think of a way to do it without a human "gatekeeper"
19:22:36 <randomuser> that would work at this scale, sure
19:23:28 <pbokoc> I guess we'll have to wait for Skynet
19:23:38 <randomuser> but if we get a 50 or a hundred or more people writing, and one percent write or edit one article each day, and half of those article edits involve adding tags...
19:24:01 <randomuser> that's like 4,375 new tags per minute
19:24:03 <pbokoc> well, that would be a good problem to have :)
19:24:09 <pbokoc> yes
19:24:21 <randomuser> we'd have to hire an army of primate tag wardens
19:24:39 <pbokoc> as I'm saying, Skynet
19:24:57 <pbokoc> or one of us could just go full Lawnmower Man
19:25:33 <randomuser> I nominate ignatenkobrain
19:26:44 <randomuser> anyway, what we do with the tags isn't as crucial at this stage as the question of whether or not tags would be valid in the proposed specification
19:29:21 <pbokoc> so where do you propose the metadata be stored? As a separate file in each repo? Some sort of header hidden in a comment in each document's source?
19:29:52 <randomuser> packaged with the documentation somehow
19:30:15 <randomuser> using in a comment would sort of defeat the purpose of abstracting this information from the document's markup format
19:30:32 <pbokoc> yeah, that's true
19:30:59 <pbokoc> but on the other hand, we could do fun stuff like having different sets of metadata for different parts of a single big document
19:31:02 <randomuser> but, I do have a halfway written thing to extract the title, subtitle, and abstract from ie Book_Info.xml
19:31:29 <randomuser> hey.. now that is an interesting idea
19:31:52 <jsmith> randomuser: Let me guess -- you're not using XPath to get that data :-p
19:32:00 <pbokoc> that would be really helpful for e.g. the cookbook
19:32:29 <randomuser> jsmith, BeautifulSoup :) it's adequately simple.
19:33:38 <pbokoc> ...but then again, someone would have to maintain that
19:34:08 <randomuser> pbokoc, it might be doable with... say, a markup: attribute and something to signal which section the metadata describes
19:34:36 <randomuser> then $tool that knows about $markup would know how to find that section in that markup
19:36:16 <pbokoc> yeah, that's what I'm thinking
19:36:31 <randomuser> https://github.com/immanetize/anerist/issues/2
19:36:41 <pbokoc> but I don't want to be the person to tell Capesteve to go make 30 different configs each pointing to a different part of the sysadmin guide :)
19:37:08 <Capesteve> ?
19:37:12 <randomuser> actually, you got me thinking bow
19:37:15 <randomuser> bowbowbow
19:37:18 <pbokoc> oh, hello Stephen :)
19:37:39 <randomuser> someone asked about how linking between docs would work in this thing, and I hadn't figured it out
19:38:05 <randomuser> but, if $tool can make anchors for named sections, it can also produce a list of anchors that a given document provides
19:38:48 <randomuser> so if I wanted to link from my very short firewall tutorial to the Networking Guide, I could discover the link by browsing to that section of the NG and pressing some magic button
19:39:03 <randomuser> #chair
19:39:07 <randomuser> #chair pbokoc
19:39:11 <randomuser> #chair Capesteve
19:39:21 <randomuser> I have to go back to work, you kids have fun
19:39:26 <randomuser> share your popcorn, Capesteve
20:09:10 <pbokoc> #endmeeting
20:09:31 <pbokoc> no? oh well
20:32:45 <randomuser> #endmeeting harder
20:33:05 <pbokoc> zodbot stopped responding at some point before you made me chair
20:33:34 <randomuser> oddly, I still get fedmsg signal about the meeting ending
20:33:47 <pbokoc> it still logs the meeting though: http://meetbot.fedoraproject.org/fedora-docs/2015-03-26/fedora-docs.2015-03-26-18.31.log.txt
20:34:18 * pbokoc wonders if having that link appear in the meeting log inside the link will make the thing explode
20:34:32 <randomuser> fedora-notifs> randomus`'s meeting titled "Fedora Docs Workshop" ended in #fedora-docs https://meetbot.fedoraproject.org/fedora-docs/2015-03-26/fedora-docs.2015-03-26-18.31.html
20:34:32 <randomuser> <fedora-notifs> randomus`'s meeting titled "Fedora Docs Workshop" ended in #fedora-docs https://meetbot.fedoraproject.org/fedora-docs/2015-03-26/fedora-docs.2015-03-26-18.31.html
20:34:40 <randomuser> hm
20:35:00 <pbokoc> I was just gonna point that out :-D
20:35:02 <randomus`> #endmeeting