18:58:42 <randomuser> #startmeeting Fedora Docs Office Hours
18:58:42 <zodbot> Meeting started Thu Apr  2 18:58:42 2015 UTC.  The chair is randomuser. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:58:42 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:58:46 <randomuser> late edition!
18:59:14 * Capesteve salutes randomuser
18:59:33 <randomuser> good day, Capesteve
19:00:06 <randomuser> what exciting things should we discuss today?
19:01:05 <Capesteve> I think smccann_afk needs some guidance as to what to do next with the virt intro guide
19:01:36 <Capesteve> randomuser: Have you checked it yet?
19:02:39 <randomuser> I was about to look for her branch
19:03:31 <Capesteve> I am trying to study tonight but I can scan it later if you need me to
19:04:10 <randomuser> your insight is always welcome!
19:04:20 <Capesteve> git repo pls
19:05:06 <randomuser> ssh://fedorapeople.org/home/fedora/mccann2/public_git/virtualization-getting-started-guide.git
19:06:24 <randomuser> Capesteve, what are you studying these days?
19:06:39 <Capesteve> Cloud, OpenStack
19:07:09 <randomuser> openshift?
19:07:26 <Capesteve> no, only OpenStack
19:19:10 <Capesteve> that book does not build for me
19:20:18 <randomuser> it doesn't here either, but...
19:20:57 <randomuser> https://bugzilla.redhat.com/show_bug.cgi?id=1199396
19:23:31 <randomuser> smccann_afk, you need to add an include
19:25:01 <randomuser> smccann_afk, the system requirements should probably align with what we are recommending in the hardware overview of the release notes; anaconda will not function with 256MB of ram without special options and even 512MB is a maybe
19:26:57 <randomuser> the network support section should link out to the networking guide for bridge setup instructions, and perhaps explain the distinction between a NAT virtual network and a bridge
19:27:41 <randomuser> the iptables DNAT mention should be about firewalld, yum commands should be dnf
19:29:19 <randomuser> imo the screenshots should reflect a FEdora ISO installation source; a rhel network installation tree doesn't seem as well targeted
19:36:34 <smccann> back... and timezone challenged cuz I thought office hrs were an hour ago.. :-P
19:37:53 <smccann> yum vs dnf - is that decided yet?  I lost the plot on that one
19:38:02 <randomuser> smccann, nah, you were right, I was just late starting it because of dayjob stuff
19:38:08 <randomuser> it's definitely happening
19:38:31 <randomuser> the discussion lately has been about whether /usr/bin/yum will redirect to /usr/bin/dnf or just not exist
19:38:44 <randomuser> and it looks like the plan is redirection
19:39:25 <randomuser> so while most yum commands should technically still function, the user will always get a message that "yum is deprecated, redirecting to dnf"
19:39:41 <smccann> ok gotcha
19:41:46 <smccann> so higher-level question - i'm doing virtmanager chapter, grundblom is doing boxes... who's doing the rest of the guide?
19:42:40 <randomuser> smccann, I suppose myself or lnovich.  If it's me, I don't know how much of the rest will stay
19:43:36 <randomuser> a couple paragraphs introducing the concept of virtualization and potential use cases should be enough for the targeted audience, imo
19:44:21 <randomuser> lnovich will probably want more detail :)
19:44:25 <smccann> yeah. I was debating whether large portions of the rest of the guide get moved into an 'appendix'. so it's there for the nosey newbie, but is obviously not required reading
19:45:14 <randomuser> if it isn't something a new user needs to know in order to get a fedora VM up and usable, it can probably go away
19:45:49 <randomuser> exceptions for obvious points of contention, best practices, basic groundwork
19:46:26 * randomuser has to go afk for a bit
19:46:43 <smccann> ok
19:46:45 <smccann> makes sense
19:47:22 <randomuser> I think you have the right idea for targeting this guide; scope creep comes easily if you aren't careful :)
19:47:31 * randomuser &
20:00:05 <Capesteve> got the book to build
20:06:40 <Capesteve> smccann: guide looks good
20:07:24 <Capesteve> would be nice to use ulink to make links to Fedora Virtualization Deployment and Administration Guide
20:07:56 <smccann> Capesteve - that guide hasn't been updated since f19 I think - is it still worth making links?
20:08:53 <Capesteve> if you mention it, then seems like you can link to it
20:08:54 <Capesteve> Ask randomuser for those special non-versioned links
20:09:12 <Capesteve> that way if it is updated the links take you to the latest one
20:09:18 <smccann> ok
20:10:36 <smccann> what was the build problem with this guide btw that you and randomuser mentioned earlier?
20:10:57 <smccann> I looked at the bug but didn't follow if it was a problem I created or aught to fix or something else
20:11:21 <Capesteve> I had to delete the link to the entity file from the book info file
20:11:52 <Capesteve> I deleted the bookinfo ID before that but that did not seem to help
20:12:21 <Capesteve> normal I only have entity link in the .xml file named after the guide
20:12:31 <Capesteve> you have it in two files
20:13:18 <smccann> ok I'll take a look, thanks!
20:13:25 <Capesteve> ahh, I think the real problem is you have en-US in front
20:13:32 <Capesteve> which becomes en-US/en-US
20:14:54 <Capesteve> Yes!
20:15:31 <Capesteve> if you remove en-US on line 3 from the Book_Info.xml file, it works
20:15:45 <Capesteve> <!ENTITY % BOOK_ENTITIES SYSTEM "Virtualization_Getting_Started_Guide.ent">
20:15:59 <Capesteve> Not: <!ENTITY % BOOK_ENTITIES SYSTEM "en-US/Virtualization_Getting_Started_Guide.ent">
20:17:48 <smccann> hmm I think i did that w/ my first commit?
20:18:43 <smccann> so can you explain where/how you saw my content etc?  I'm a git about git so hoping I didn't screw it up (for the 3rd time :-)
20:19:06 <Capesteve> i cloned the repo, randomuser posted the url
20:20:07 <smccann> ok I'll hafta go clone myself and see what went wrong... I'm guessing my first commit didn't commit.
20:20:28 <smccann> It all builds fine for me locally, but I did make that en-US change (locally).
20:20:39 <Capesteve> do you mean you fixed it with your first commit?
20:20:58 <smccann> yeah if you look at the commits, the first one was supposed to fix that issue
20:21:38 <Capesteve> ^b5ac9d0 (Dayle Parker 2012-06-27 11:20:56 +1000  3) <!ENTITY % BOOK_ENTITIES SYSTEM "en-US/Virtualization_Getting_Started_Guide.ent">
20:22:40 <smccann> https://fedorapeople.org/cgit/mccann2/public_git/virtualization-getting-started-guide.git/commit/?h=smccann_virtmanager&id=284f7550cde7f686a1f01ade045dcf8a8cf154ae
20:23:14 <Capesteve> hmm, interesting
20:26:23 <Capesteve> I see HEAD is where randomuser last made a commmit
20:26:51 <Capesteve> I suggest we blame him for braking it somehow
20:27:30 <Capesteve> ahh, you have a branch
20:27:33 <smccann> I can live w/ that :-)
20:27:50 <smccann> oh sorry... um hey!  I have a branch!
20:27:54 <Capesteve> you have been committing to the branch
20:28:12 <smccann> yes.. i'm in my total newbie isolation box for now
20:28:13 <Capesteve> I was not expecting that, I am a bit slow
20:28:27 <smccann> myself and grundblom... and then 'someone' will make our work real... somewheres
20:28:44 <Capesteve> so I'll checkout the branch and build that then
20:29:59 <Capesteve> may I suggest, git cherry-pick that first patch to master?
20:32:05 <smccann> i don't think I have the authority to do that, but yes, I think the problem exists in master as I think both grundblom and I have a commit in our private branches to fix it
20:33:53 <Capesteve> earlier randomuser mentioned a missing include, now I see that he meant you need to add in include line in the en-US/Virtualization_Getting_Started_Guide.xml file for your new file
20:34:16 <Capesteve> otherwise it does not get built
20:34:48 <smccann> ah ok
20:37:04 <smccann> thanks for the help Capesteve  don't want to take too much of your OpenStack study time away! I'm just trying to get into that myself
20:37:16 <Capesteve> kk
20:44:21 <randomuser> I've been putting the entity declaration in every file; I guess because I didn't think it through
20:50:56 <grundblom> hey everyone just got back, wow, there is a lot to read here.
20:51:23 <smccann> hey grundblom!
20:51:37 <randomuser> hey grundblom !
20:51:45 * randomuser just got back too, sort of
20:52:05 <smccann> hey randomuser!
20:52:06 <smccann> :-)
20:53:19 * jhradilek waves.
20:53:25 <smccann> oh randomuser - Capesteve sez I should ask you how to ulink to the nonversion location for the virt deploy guide
20:53:49 * smccann waves all-round for everyone!
20:54:09 <randomuser> smccann, I'm not sure there is one, let me check
20:54:45 <Capesteve> smccann: Have you cloned any other guides? you could grep them for examples of ulink
20:55:29 <Capesteve> there are some in your guide already
20:55:33 <smccann> Capesteve I have the deploy guide so yeah I can probly find an example of ulink somwheres come to think of it
20:55:42 <smccann> doh
20:55:58 <randomuser> smccann, we have some redirect set up, ie docs.fp.o/networking will automatically redirect the user to the latest version of the networking guide.
20:56:22 <randomuser> but there's nothing like that [currently] for the virtualization administration and deployment guide
20:56:35 <Capesteve> smccann: look in en-US/Products
20:56:36 <randomuser> and maybe there shouldn't be, until the guide is updated
20:57:12 <Capesteve> smccann:  grep -r ulink en-US/Products.xml
20:57:28 <smccann> so no links then for now to deploy guide?
20:57:38 <randomuser> no versionless links
20:58:07 <Capesteve> so the question is randomuser, should smccann puts some ulinks in
20:58:11 <smccann> so put f19 links in?
20:59:11 <randomuser> if it were me writing it, I would consider manpages, upstream documentation, and other vetted docs, and make the best choice for the given topic
20:59:45 <Capesteve> so the remove mention of old docs
20:59:49 <randomuser> linking to a draft guide for an eol fedora release doesn't seem ideal - but it may still be the best resource
21:00:04 <Capesteve> my point was if you have them mentioned then why not make the title a link
21:00:25 <Capesteve> if they are so bad then why mention them
21:00:55 <Capesteve> but if its border line then ignore my suggestion
21:03:39 <smccann> ok gotta run.. thanks all for the help!
21:03:47 <Capesteve> kk
23:25:26 <kendell> did I miss the meeting, again?
23:31:26 <randomuser> kendell, I didn't close the meeting.... so, no!
23:31:32 <kendell> lol ok good
23:32:20 <randomuser> it's a free-for-all anyway, did you have anything?
23:32:49 <kendell> just wondered if anyone besides ben has had a chance to look at my updated guide and if they had time to either docbook it or help me do it.
23:33:29 <kendell> I could actually fork the repo but am hesitant to do that, it's just about done besides all the xml and markup stuff
23:33:50 <randomuser> any edits since the mail?
23:33:54 <kendell> nope
23:34:23 <kendell> I've been concentrating on getting critical accessibility bugs fixed so f22 doesn't ship with an inaccessible gdm
23:34:57 <randomuser> kendell++
23:35:56 <kendell> I've also been helping out on ask fedora when I can, although I'm still trying to figure out how to keep track of questions I've written answers for
23:36:31 <randomuser> nice!
23:36:49 <randomuser> I've been thinking that ask fedora might be good docs training wheels
23:36:57 <kendell> most of the questions there seem to be about nonfree drivers, codecs, or hardware issues I don't know how to solve
23:37:16 <randomuser> yeah, i think there are a total of six unique questions there :P
23:37:53 <kendell> usually nvidia issues. Dunno why nvidia gives linux so much trouble. That, and hybrid graphics. I thought linux's troubles with hybrid graphics had been largely solved
23:38:20 <randomuser> the nvidia thing is hard because nvidia ships a broken installer
23:38:36 <kendell> tsk tsk. Bad nvidia. Bad bad nvidia
23:38:48 <randomuser> AMD does it too :)
23:39:06 <kendell> this is going to sound obvious, but they need to give higher priority to their linux stuff
23:39:59 <randomuser> it's mostly just packaging issues, really, and people can avoid them by using sanely packaged, readily available drivers
23:40:22 <kendell> nods, rpmfusion. Although I don't know how to get around nvidia/amd drivers not being signed, breaking secure boot
23:40:29 <randomuser> you can'
23:40:34 <randomuser> can not.
23:40:49 <kendell> nods, that's an upstream issue, out of fedora's hands
23:40:59 <randomuser> well, you could roll your own keys and sign things, maybe
23:41:28 <kendell> that's a possibility. I don't have any nvidia hardware here at all, and only one amd card which the foss drivers work absolutely fine for
23:42:39 <kendell> I'm tempted to bug the gnome people to make gnome lighter on resources, so it doesn't suck quite as bad on low resource hardware
23:43:11 <randomuser> hahahaha
23:43:19 <randomuser> you would not be the first person to state that concern
23:43:21 <kendell> bet that'll get nowhere though, so I'm not bothering
23:43:42 <kendell> I mean, when windows beats gnome on the same hardware ... that's kind of sad
23:44:00 <randomuser> yeah, but windows is only windows
23:44:08 <kendell> huh?
23:44:11 <randomuser> but Fedora is not only GNOME
23:44:20 <kendell> I ... don't get it
23:44:36 <randomuser> you don't have to use gnome to use fedora
23:44:41 <kendell> oh ok
23:44:59 <kendell> well, you practically do if you're blind, since the mate version doesn't have orca yet, but still
23:45:17 <randomuser> you can't use orca with mate?
23:45:28 <kendell> sure you can. It's just not installed on the mate live images by default
23:45:38 <randomuser> ahh... have you filed a bug for it?
23:45:59 <kendell> nope. I'm actually not sure how. Where do I file, bugzilla.redhat?
23:46:22 <randomuser> yeah... I'm not sure which component would be best, though
23:46:28 <kendell> me either
23:47:04 <randomuser> a mail to devel@ might do the job; list out the spins and their accessibility state, provide a clear path for improvement, and I'd venture you'll see some action
23:47:37 <kendell> not a bad idea. All that's needed is to include orca on accessible desktops, once orca is turned on it just works
23:47:42 <randomuser> hey, looking at this accessibility guide
23:48:10 <randomuser> the second paragraph of section 2 talks about the absence of GUI dependency
23:48:54 <randomuser> would it make it more appealing, to a visually impaired user, to discuss how GUIs are built using toolkits with inherent accessibility support?
23:48:57 <kendell> yeah, you don't have to have a gui to run linux, and the cli is accessible. The problem is, fedora's kernels don't include the speakup kernel modules, and espeakup isn't in the fedora repositories, so unless that's fixed you need gui if you want accessibility
23:50:11 <kendell> I can provide an overview. i don't know much, but I know the basics. A lot of the stuff in the first couple of chapters I left unaltered. I most concentrated on gnome and mate accessibility, and updating installation of software to mention gnome software and changed yum to dnf since it's replacing yum
23:50:52 <randomuser> kendell, to me, that portion of the copy leaves room for the reader to infer that they will have to use second-class applications
23:51:24 <kendell> you know, it looked like that to me too. It sounded a bit like a sales pitch, carefully glossing over the areas that weren't so great at the time
23:51:33 <randomuser> when in reality, if I understand it correctly, applications QT and GTK have buttons and widgets etc that just work with orca
23:51:59 <randomuser> applications built on QT and GTK, i mean
23:52:06 <kendell> for the most part, yes. Unless the dev has mangled something or uses custom widgets, and even then gtk has methods to get names from those if it can.
23:52:32 * randomuser nods
23:52:46 <kendell> qt accessibility was nill back when this guide was last written. In fact, kde is all but inaccessible to orca, and I'm not sure how good their builtin accessibility tools are
23:52:50 <randomuser> it would seem preferable then to elaborate on that point
23:53:22 <kendell> good point. Where should it go? in the second chapter?
23:53:38 <randomuser> rewrite that paragraph
23:53:44 <kendell> ok, looking now
23:54:04 * randomuser chuckles a bit
23:54:31 <kendell> lol I totally overlooked that. That does sound like a sales pitch. And a bad one
23:54:53 <randomuser> you know, we've discussed using phrases like "look for this" or "for more info, see" because of the potential offense to blind users
23:55:01 <kendell> that? Oh disregard that. Now, did I tell you about linux's many accessible applications? rofl
23:55:04 <randomuser> avoiding using them, that is
23:55:57 <kendell> it doesn't  offend me. There's really no blind friendly way to say read more documentation if you want to know more. You can't very well say something like, head on over to http://example.com and have a listen
23:56:09 <kendell> well, you could, but ...
23:56:12 <randomuser> heh
23:56:17 <randomuser> anyway
23:56:39 <randomuser> the next paragraph talks about fedora is popular, industry professionals use it, etc; IMO we could tout the community there instead
23:57:12 <kendell> is that paragraph really accurate? I mean, I know it's affiliated with redhat and all, but do a lot of businesses use it?
23:57:18 <randomuser> that's something we can tangibly offer; community support, etc
23:57:32 <randomuser> like you said, it's a comforting pitch
23:57:52 <kendell> yeah, it does ... sigh. If I read this, I'd be reinterpreting every other line.
23:58:15 <kendell> community support is a big plus.
23:58:45 <kendell> and what's that whole section 508 mandate bit? I know for a fact most disability laws are week at best, and most businesses ignore them with impunity
23:59:14 <randomuser> so that section, rather than referring ambiguously to industry professionals that might help, can have directly actionable references and links to ask.fp.o or the IRC wiki page
23:59:41 <kendell> that's a really good idea. I'll put in links, and whoever docbooks it can turn those links into real ones, I guess?
23:59:52 <randomuser> yes
00:00:09 <randomuser> I mean, add context too, don't just paste in links
00:00:35 <kendell> yeah. I was thinking something like, to learn more on fedora's community support, visit http://example.com
00:00:53 <randomuser> sure, however you want to deliver it
00:01:04 <kendell> after I rewrite that paragraph. That sounds a bit more like a pitch redhat would write than fedora, which focuses more on the community than business
00:01:40 <randomuser> the section 508 stuff is interesting IMO, but not if there aren't prevalent VPATs for software in Fedora
00:01:44 <randomuser> ...and I think there is not.
00:02:17 <kendell> I've never heard of this vpat thing. Most accessibility laws focus and are affected most by windows. Linux falls through the ccracks, even though it's often just as accessible, and often more
00:02:50 * randomuser nods
00:02:57 <kendell> come to think of it, i've never heard of any hearing impaired software for linux. Is there any?
00:03:18 <randomuser> yeah, there's all kinds of stuff
00:03:46 <kendell> I've seen a few things that depend on old deprecated at-spi v1.x, which orca no longer works with, but it might still work for what it's intended to do
00:04:34 <kendell> I need to add info on refreshable braille displays, which orca and speakup both work with very well.
00:04:53 <kendell> too bad I don't have a couple thousand bucks to plunk down on one to test with linux
00:05:57 <randomuser> hey, if you learn enough to be the guy coding application support for them, someone might buy one for you
00:06:39 <kendell> they really shouldn't be that expensive, imo. They're fragile and not built any better than most consumer grade hardware but if you're blind, get used to being gouged for special hardware and software, assuming you ue windows
00:07:10 <randomuser> it seems like the speakup section could use more context
00:07:11 <kendell> if you use linux, you're extremely lucky. All accessibility software that I know of is free software, both speech and in beer
00:07:27 <randomuser> like, "speakup is not available in fedora because..."
00:07:37 <randomuser> and have a definitive answer to complete that sentence
00:08:21 <kendell> yeah, I need to expand on that. I need some leverage with someone over at pulse audio to get on whatever bug is keeping speakup form working very well with pulse audio. The problem is this. Espeakup is the thing that talks, and both it and speakup run as root, whereas pulse audio runs as user process, so speakup can't get access to the audio card.
00:08:21 <kendell> I don't know how to fix that
00:09:01 <kendell> I hope to get that rectified, but until speakup works reliably with pulse audio without all sorts of hacks to get it to speak there's not much point
00:11:10 * randomuser nods
00:11:29 <randomuser> if it doesn't work, the guide should either explain the shortcomings or just not mention it
00:11:50 <kendell> I'm just not sure how to go about it. The speakup people are ... unfriendly at best, telling you to use another console screen reader, as if there were any. The pulse audio people claim it's not their problem, go bug someone else, and around and around
00:12:30 <randomuser> yeah, I dunno.  Only talking about representing the situation accurately in the guide.
00:12:55 <randomuser> one more point for you - the orca section is all in one massive paragraph
00:13:03 <kendell> that I can do. I wish I could offer a solution, but aside from compiling your own kernel, and that's well beyond something the average person can do, there isn't one, unless speakup gets added into the fedora kernels
00:13:33 <randomuser> that should probably be broken up somehow
00:14:05 <kendell> kernel is already separated into kernel core and kernel modules, so ... maybe kernel modules speakup, or something?
00:17:23 <randomuser> nah, kernel-modules is still mainline only, it is just to make cloud images lighter
00:17:42 <randomuser> I've got to go, talk to you later
00:17:49 <randomuser> #endmeeting