- From: PhistucK <phistuck@gmail.com>
- Date: Tue, 4 Dec 2012 14:32:35 +0200
- To: Chris Mills <cmills@opera.com>
- Cc: "public-webplatform@w3.org" <public-webplatform@w3.org>
- Message-ID: <CABc02_KmfFL9QUqwZ5GLkHXbOWo1yYApBsSgJQ7sW_RPm=2EtQ@mail.gmail.com>
See my comments inline. ☆*PhistucK* On Tue, Dec 4, 2012 at 2:08 PM, Chris Mills <cmills@opera.com> wrote: > On 3 Dec 2012, at 12:03, PhistucK <phistuck@gmail.com> wrote: > > > See my comments inline. > > > > ☆PhistucK > > > > > > > > On Mon, Dec 3, 2012 at 11:37 AM, Chris Mills <cmills@opera.com> wrote: > > On 1 Dec 2012, at 10:53, PhistucK <phistuck@gmail.com> wrote: > > > > > I also think the core stuff should be finalized first. I think it > should the top priority. That is what everyone (no exception) uses - and it > is a huge mess. > > > > > > I am working on all of the DOM related reference pages for a while. > > > I am not nearly finished, but I think I am making some good progress. > > > However, I am not really adding content - I am really reorganizing it > (taking page summaries from manual MSDN listings and removing those > listings afterward, modernizing the MSDN examples, cleaning up weird 'out' > MSDN parameters, moving non semantic standards information to the standards > table and so on). > > > I sometimes add some usage information and compatibility information > (from caniuse, MDN and such), or write my own summary, but in the majority > of cases, this does not really happen. > > > > It is another thing that needs sorting out in a sensible way really. > > > > Lost you here, define "thing"? > > > > Problem with the site, task to be done... > Still lost, but feel free to ignore... > > > > > > > > I am also scared of moving pages, because of the Semantic Wiki cross > reference issues that it may cause (we really need a bot for fixing all of > the links after a move, especially since we have agreed upon a URL system), > so the old URL system is still very much in place (dom/objects/bla, > dom/methods/bla…). > > > > Julee and some others have been working on finalising the URL system I > believe. Where did that get to? I'd say once the mess is sorted out, we > will have something more solid and stable to work on. > > > > Apart form that, essentially it is just a matter of checking to make > sure there is nothing already at the URL you are intending to move > something to. I don't think it lets you just overwrite pages anyway. > > > > I am not scared of overwriting (though that simply has not crossed my > mind ;)), I am scared of broken references due to not removing the original > URL. I understand Semantic Wiki has some issues where the queries go crazy > or something, I do not remember. Some thread recently (a month or two) > mentioned it. > > Right. I'm not sure about that. It does create automatic redirects when > you move pages, so that references are not broken. As you saying that > doesn't work or messes up sometimes? > I do not remember the exact problem, but I think it has something to do with API Listing, method/property/event drawing and stuff like that. It messes stuff up sometimes. It breaks these connections, I think. Here - http://lists.w3.org/Archives/Public/public-webplatform/2012Oct/thread.html#msg39 http://lists.w3.org/Archives/Public/public-webplatform/2012Oct/thread.html#msg163 http://lists.w3.org/Archives/Public/public-webplatform/2012Nov/thread.html#msg152 > > > > > > > > > > > One idea that I had in mind, is that the standards table should come > from a list, with an option to either add to the list (like the current > type system), or an option to add a custom one only to the page itself. > > > The problem (sort of) is that standards have lots of editions, > especially working draft ones. So the links to them should be dynamic or > flexible somehow, with a automatic completion to the current version. > > > (I admit I am sinning by always linking to the latest version, which > can always change when it is a working draft (WHATWG HTML, W3C HTML5, DOM > Level 3 Events, DOM Level 4, WHATWG DOM).) > > > > I'd say that when quoting specs, really we need to link to both the > stable version of the spec, but then also the latest experimental version. > So you'd have something like > > > > the <ul> element (Stable:HTML 4.01 | In progress:HTML5). > > > > With "Stable:HTML 4.01" pointing to > http://www.w3.org/TR/html401/struct/lists.html > > and "In progress:HTML5" pointing to > http://www.w3.org/TR/html5/the-ul-element.html#the-ul-element > > > > Of course, if the item being talked about is part of a working draft or > other non-stable document, we would just need to link to the latest > experimental version > > > > so something like > > > > getUserMedia (In progress:Media capture and Streams). > > > > linking to > http://www.w3.org/TR/mediacapture-streams/#dom-navigator-getusermedia > > > > > > Would something like this work? I'd be happy to spec it out and write a > guideline. > > > > Yes and no, it can be broken if the latest version does not have it (for > example, PeerConnection turned into RTCPeerConnection. #dom-peer-connection > would lead to nowhere). > > Unfortunately, the right thing to do would be to link to the exact > version where something is specified, but it breaks the fixed list idea. > > Well, I was thinking of linking to the latest version it is mentioned in, > in cases like that. You obviously wouldn't link to a spec where the thing > isn't mentioned. If the method or whatever it is being discussed on the > page changed syntax in the way you describe, this would mean the page is > obsolete, and would need updating, so should be flagged as such. > > Does that mean we cannot create a fixed specification document list? > > > > > > > > > > > > > > > > On Sat, Dec 1, 2012 at 12:32 PM, Chris Mills <cmills@opera.com> wrote: > > > Hi all, > > > > > > I was just thinking about this. I've seen some discussion going back > and forth about adding documentation to WPD for stuff like Audio API and > CSS regions. That's great, but surely we should concentrate more for a bit > now on getting some of the existing HTML/CSS/JavaScript core stuff sorted > out? Getting that core stuff in place is surely a higher priority than > documenting nascent standards features that currently have limited browser > support. I am happy to work with Julee and the others to formulate a plan > for this. > > > > > > After I've got the high level page structures/IA/UX in a bit more > working order over the next couple of weeks, I am happy to start > contributing to the low level content myself. > > > > > > I'd say a general plan would be: > > > > > > * Split the content into areas of responsibility, e.g. CSS learning > articles (concepts plus tutorials), CSS property references, CSS selector > references, HTML learning articles, HTML element references, etc. > > > * Assign those areas to individuals who can take responsibility for > their tending > > > * Get people working on those areas over the next couple of months. > I'd say each domain area needs an editor and a proof reader, possibly a > demo writer as well, as when Lea gets Dabblet in place, we'll need to > install live demos on all the articles. > > > > > > Shall we discuss this at the general meeting on Monday? > > > > > > Chris Mills > > > Open standards evangelist and dev.opera.com editor, Opera Software > > > Co-chair, web education community group, W3C > > > Author of "Practical CSS3: Develop and Design" ( > http://my.opera.com/chrismills/blog/2012/07/12/practical-css3-my-book-is-finally-published > ) > > > > > > * Try Opera: http://www.opera.com > > > * Learn about the latest open standards technologies and techniques: > http://dev.opera.com > > > * Contribute to web education: http://www.w3.org/community/webed/ > > > > > > > > > > > > > > >
Received on Tuesday, 4 December 2012 12:33:52 UTC