| 10:17 | <zcorpan_> | can someone with Edge check http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3699 ? |
| 12:19 | <Ms2ger> | > Steven Pemberton gives a talk entitled: "After HTML5, What Next?" |
| 12:19 | <Ms2ger> | Not sure I want to know |
| 12:30 | jgraham | expects it involves Namespaces. And RDF. |
| 12:33 | <zcorpan_> | philipj: it's not clear to me what you want the shape of the parser to be like |
| 12:35 | <philipj> | zcorpan_: basically, first write it as if interleaved cue and style blocks were supported, then add a check right before actually using the contents of a style block, so that supporting interleaved blocks is just a matter of removing that check |
| 12:36 | <philipj> | that we there won't be too much to think about in terms of test changes and the like if that change is made |
| 12:36 | <philipj> | s/we/way/ |
| 12:37 | <philipj> | zcorpan_: was that clarifying, or is it how to actually make that true that's not clear? |
| 12:38 | <zcorpan_> | i guess my problem is that the current algorithm being flat and jumping around makes it insane, but i suppose i can change it to use a different style |
| 12:40 | <philipj> | once one gets safely to step 17 (cue loop) it should be a matter of checking for "STYLE" before assuming that what we have is a cue, right? |
| 12:41 | <philipj> | of course cue loop will then be a terrible name |
| 12:42 | <zcorpan_> | do you want the algorithm to share the steps to "slurp data until a blank line" between STYLE and cue? |
| 12:44 | <philipj> | zcorpan_: that sounds nice on the face of it, especially if that could also be used while looking for a blank line immediately after the WEBVTT line |
| 12:44 | <philipj> | void setMetadata(MediaMetadata metadata); |
| 12:44 | <philipj> | oops |
| 12:44 | <philipj> | zcorpan_: if you could also handle NOTE and explicitly ignore it, it would make it obvious how to avoid emitting console warnings for them in implementations |
| 12:46 | <philipj> | and in all of this block handling, I suppose that lines with --> or that start with STYLE/NOTE, should always start a new block even without a newline? |
| 12:47 | <zcorpan_> | yeah --> should attempt to create a cue |
| 12:50 | <zcorpan_> | philipj: i'm not sure it's a good idea to revamp the parser in this PR though |
| 12:58 | <philipj> | zcorpan_: perhaps not, but if it is going to happen at all, it would be nice to do it before adding STYLE, or at least that's the order I'd like to see it implemented, otherwise it's extra churn |
| 12:59 | <philipj> | it's pretty unlikely that I'll be the one actually implementing this, though, so my word shouldn't carry too much weight if you think I'm just wrong :) |
| 13:01 | <zcorpan_> | i suppose i can revamp and then rebase the PR |
| 18:01 | <nikkibee> | hello |
| 18:01 | <botie> | hey, nikkibee |
| 18:01 | <nikkibee> | I'm looking to do an internship with outreachy and I was interested in this project :) |
| 18:01 | <nikkibee> | I've already built the git repo like it says here https://wiki.mozilla.org/Outreachy/2016/December_to_March#Contribute_to_the_HTML_Standard.21 |
| 18:02 | <nikkibee> | I've looked at the bugs marked easy but I'm not sure where I could start |
| 19:20 | <Mek> | hmm, how do specs generally deal with race conditions between "in parallel" algorithms... From the definition of "in parallel" it seems any possible interleaving is allowed, but any algorithm that checks some state and then updates it in a later step would have potentially bad race conditions then... |
| 19:38 | <wanderview> | nikkibee: you probably want to talk to annevk (who is probably traveling to the TPAC conference right now) |
| 19:38 | <nikkibee> | yep, I saw on the project page that annevk is the mentor for it |
| 19:39 | <nikkibee> | so are they going to be busy with the conference until next week? |
| 19:43 | <nikkibee> | http://www.w3.org/2015/10/TPAC/ this is the conference, right? |
| 19:43 | <nikkibee> | looks like I won't be able to really reach him in time for the outreachy deadline :X |
| 19:43 | <wanderview> | nikkibee: yea, the conference is all week,.. and its in Japan so he will probably online at opposite times from right now |
| 19:44 | <nikkibee> | gotcha |
| 19:44 | <wanderview> | nikkibee: when is the deadline? I would recommend email... his address is one his mozillians profile: https://mozillians.org/en-US/u/annevk/ |
| 19:44 | <nikkibee> | the deadline for the application is november 2nd |
| 19:45 | <nikkibee> | I could email him and hope he can respond, but I'd also need help on getting started enough for the application process from other people here |
| 19:45 | <nikkibee> | I tried out a team where they were in a far-apart timezone and I couldn't do anything with just a few comments over night |
| 19:49 | <wanderview> | nikkibee: I guess I would recommend starting with an email to annevk... depending on which issues you are interested in he might be able to direct you to someone in this timezone who knows about it |
| 19:50 | <wanderview> | nikkibee: the timing is tough, though, since a lot of people from this channel are at that conference :-\ |
| 19:50 | <nikkibee> | ah :X |
| 19:50 | <nikkibee> | I'll still try an email |
| 19:50 | <nikkibee> | I've been having a rough time with finding a project so far so now I'm just trying a lot at once |
| 19:52 | <wanderview> | hopefully we can get it worked out... annevk is normally pretty responsive |
| 19:57 | <nikkibee> | that's good! |
| 19:57 | <nikkibee> | thanks for letting me know how it is right now wanderview, I'll get started on that email |
| 21:15 | <zcorpan> | nikkibee: maybe i can help |
| 21:17 | <zcorpan> | nikkibee: how about starting with this one https://github.com/whatwg/html/issues/265 ? |
| 21:18 | <zcorpan> | nikkibee: also add yourself to the spec's Acknowledgements section in the PR |
| 21:21 | <nikkibee> | zcorpan: hi! that was one I was interested in doing |
| 21:21 | <zcorpan> | Mek: yes, but specs should not be doing that. they should check state before starting "in parallel" and then update state with "queue a task" |
| 21:22 | <nikkibee> | I don't know what you mean by the spec's Acknowledgements section. do you mean when I make a PR? |
| 21:22 | <zcorpan> | yeah. at the end of the spec there's a list of names of people who have contributed |
| 21:23 | <zcorpan> | we want people who contribute to add their name to the list as part of their first contribution, but this might not be properly documented yet |
| 21:23 | <Mek> | zcorpan: unfortunately that's not what all specs are doing... http://www.w3.org/TR/service-workers/#unregister-algorithm for example "in parallel" looks up a registration and then checks if it fonud one |
| 21:24 | <nikkibee> | zcorpan: I can certainly do that |
| 21:25 | <nikkibee> | https://github.com/whatwg/html-build I git cloned this and followed the build instructions. what do I do next? I'm not sure if I need to make a fork of the https://github.com/whatwg/html directory and start there, or if I can make my fork from what I have locally now, or what |
| 21:26 | <zcorpan> | Mek: ok. i'm not familiar enough with service workers to be able to tell if it's a bug or not, but certainly you can raise an issue about it |
| 21:26 | <zcorpan> | nikkibee: you should fork whatwg/html and make a PR in that repo |
| 21:27 | <Mek> | well, that's just one example of it (and it's definitely a bug, since it means two unregister could both return true when only one should) |
| 21:27 | <nikkibee> | I will do so |
| 21:28 | <Mek> | zcorpan: in some other algorithms it has "Run the following steps atomically". I suppose adding such language in more spots would also work to fix the issue... |
| 21:29 | <Mek> | actually maybe it's not a bug in this case |
| 21:30 | <zcorpan> | gotta go. good luck nikkibee :-) and good night |
| 21:30 | <nikkibee> | thanks! goodnight- oop |