14:01:08 #startmeeting Fedora Docs 14:01:08 Meeting started Mon Jul 31 14:01:08 2017 UTC. The chair is bexelbie. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:08 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:01:08 The meeting name has been set to 'fedora_docs' 14:02:50 bexelbie, mtg already started? 14:03:04 it looks like it is just you and me :: 14:03:24 worrisome ... 14:03:38 I'd liek to try and publish the new site, assuming we can all agree on a fix for #5 in docs-fp-o 14:05:15 lemme see 14:06:33 bexelbie, I don't get it... Why is that an issue? 14:06:46 let me check my issue number :) 14:07:12 because if cross-references are broken then the links don't work 14:07:22 Maybe I didn't understand your response ... 14:07:40 are we talking about https://pagure.io/docs-fp-o/issue/5 14:07:41 ? 14:07:48 https://pagure.io/docs-reboot/docs-fp-o/issue/5 14:08:10 heh - no mallard :) 14:08:20 sorry, I should have provided a link to begin with 14:08:34 tf? two different projects? 14:10:02 hmm 14:10:12 the one you are referencing will be deprecated 14:10:20 the docs-reboot one is the one that is ready to go to production 14:10:43 Is there a simple way to script the conversion to hard links? 14:10:57 I have a rough idea but no functioning code yet 14:11:14 I have also not thought of a good way to leave them as is and do resolution at publish 14:11:31 so I am looking at trying to resolve the "how to do it one time and be done" method 14:11:32 I can basically only think of grepping for the referenced ID... 14:11:39 yep 14:11:51 we also have to grab the proper title and format it in place 14:12:05 Which isn't half bad, but I'm afraid there'll be some gotchas. 14:12:16 it sounds like, in principal, you're ok with explicit xrefs though as a way to move forward 14:12:17 ? 14:12:39 That shouldn't be hard, provided the sources follow the same formatting principles. 14:12:51 No, I don't much care one way or the other. 14:13:04 I'm wondering, how do OS folks deal with it. 14:13:15 They must be facing the exact same problem. 14:13:38 they use explict xrefs 14:13:42 as documented in #1 14:13:48 They publish using AsciiDoctor in downstream, so there must be a way to go from one to the other. 14:13:54 I am hoping they have some sanity checking scripts somewhere 14:14:17 ok, which #1 again? 14:14:25 https://pagure.io/docs-reboot/docs-fp-o/issue/1 -- pbbly not this 14:15:39 #1 in my comment on that issue 14:16:09 rkght 14:16:15 specifically they use the `xref:relative.path.file.adoc#section[link text]` structure 14:16:26 according to their contribution guide 14:16:33 that's ugly, but works 14:17:04 kinda takes the magic out of cross-linking... 14:18:23 I know 14:18:29 but I am not sure we should block while we rebuild magic 14:18:33 that is the core of the question 14:18:34 should we? 14:18:39 no 14:20:48 so we are unamious in this it sounds like :) 14:21:01 all two of us 14:21:12 I hope to find time to figure out a script to do a one-time rewrite 14:21:18 then I'll republish the site 14:21:25 I'll also work on queueing up the next steps 14:21:43 It looks like we have no one else who is going to raise a serious "for" or "against" ... so I'll keep the list informed 14:22:02 you got anything I could help with? 14:23:34 that script :) 14:23:40 yeah 14:23:47 that's what I'm thinking 14:23:50 also, text/feedback on the proposed landing page 14:23:52 I can do that 14:24:00 where's that? 14:24:05 https://bex.fedorapeople.org/fedora-docs-web 14:24:17 ah, I saw that 14:24:18 also read https://lists.fedoraproject.org/archives/list/docs@lists.fedoraproject.org/thread/7QCQS6JWXB7ZGKYOKLYYZYRYOE7RAZSF/ if you haven't already 14:24:50 that part htat says "Words words words" could use some editing :P 14:25:09 I think we will figure out a better landing page over time, especially once we have two versions shipped 14:25:18 Why? I thought that was intentional :-P 14:25:44 ok, to get the docs built locally: 14:26:17 - I get https://pagure.io/docs-reboot/docs-fp-o/ for builder,sh 14:26:25 - I get the book's repo 14:26:25 .hello2 14:26:30 x3mboy: x3mboy 'Eduard Lucena' 14:26:39 - I run builder.sh in the book's repo 14:26:48 Is that correct? 14:27:02 if you want to build a single repo's content - the README tells you to how 14:27:09 basically you only need builder if you are making the whole site 14:27:18 o/ x3mboy 14:27:23 bexelbie, o/ 14:27:25 right 14:27:29 I see 14:27:30 Sorry I'm late 14:27:32 otherwise you only need the repo that containst he content you want to work with 14:27:34 no worries 14:27:40 it's just us - not much of a meeting ... 14:28:15 OK, I'll try to script it either tomorrow (unlikely) or on Wed (that should work). 14:28:35 Which means we miss Aug 1st... but... 14:28:40 So, what's the current state? 14:29:14 rkratky, Aug 1 was ambitious 14:29:41 bexelbie, I like ambitious :-) 14:29:55 x3mboy, we seem to be good to go with two major blocks 14:30:04 Anyway, gotta go AFK. I've got my AI. 14:30:13 1. Release notes isn't fully converted (I am doing this now) and we need some text for landing page 14:30:20 2. the cross-refs are broken 14:30:28 for #2 rkratky is going to try to script a solution on wednesday 14:30:34 if we don't get a better option :) 14:30:39 rkratky, ty!! 14:30:56 bexelbie, huh, you mean you're converting RN to AB, right, not adoc? 14:31:04 The only significant feedback I have seen is about translations .. and that is phase 2 14:31:13 rkratky, yes, I am making it build iwth asciibinder right now 14:31:20 in pagure.io/docs-reboot/release-notes 14:31:47 bexelbie, cool, for a second there I was afraid you were converting from DB again 14:31:56 That on on your fp.o looks really nice :D 14:32:17 Ok, so how can I help? 14:32:34 want to write a script? 14:32:49 you could either do it before rkratky can or help him 14:33:04 we also need to queue up a list of repos that need to be marked deprecated/moved 14:33:13 I was planning to try and put my intern on that if no one stepped up 14:33:43 There is also an infra ticket that will be needed - but it is mostly drafted in my head and would take as long to explain as to do myself 14:33:51 though i'll get you to sanity-check my idea 14:33:55 if you want 14:35:06 Ok, well for scripting I prefer to help than do it myself 14:35:25 So, tell me what is the idea 14:35:32 For the ticket 14:35:49 ok, so I believe what we need to do, is the following 14:36:23 1. rename docs.fp.o to docs-old.fp.o and have the content in the CDN, etc. moved over. This should no longer autopublish from anywhere. Publishing will require an infra ticket - only one more update is expected 14:36:37 2. At some point delete hte F26 content from docs-old (then get infra to republish) 14:36:57 3. Create a new docs.fp.o and docs-stage.fp.o that publishes from a web-publishing repo (this requires a CI job that isn't done yet) 14:37:17 4. Rename pagure.io/docs-reboot/ pagure.io/fedora-docs (this will probably require help in the DB in pagure) 14:37:34 5. move all the all pagure repos that we don't use to somewhere like pagure.io/fedora-docs-old/ 14:37:37 wdyt? 14:39:15 For 1, It's ok but this will require redirecting from docs.fp.o untul the new docs site is ready (I think) 14:40:07 I was thinking that we would ask them to do it all as one ticket (or at least 1-3). 4 and 5 are a separate and unrelated ticket 14:40:39 For 2, ok, For 3 ok, For 4, ok (there is no point to maintain the old repo if we will work on the new site) and for 5 that will be also when the new site is ready 14:43:19 so it sounds like my plan is mostly sane :) 14:43:34 Yes, it looks ok 14:43:49 The scripting need is in https://pagure.io/docs-reboot/docs-fp-o/issue/5 14:43:53 fi you want to look at it 14:44:44 basically we need to convert a bunch of things like look like `<>` to `xref:../relative/path/to/file.adoc#FOO[Title from file.adoc for FOO]` 14:46:51 Ok, I will look at it 14:52:15 sweet - feel free to ping with questions 14:52:22 I am going to end this "meeting" 14:52:25 anyone else? 14:54:46 #endmeeting