Welcome to my last WHATWG Weekly report.
The week was relatively quiet. Most of the activity centered around the Last Call for HTML5 and other documents at the W3C.
In the WHATWG email list, Kenneth Russell make a request to use CORS for cross-domain image and video requests. CORS is Cross Origin Resource Sharing, and the request was to add an attribute to image, video, and audio elements. Ian Hickson did add this attribute.
Markus Ernst posted a request for a content-style attribute for contenteditable elements, and David Dahl posted a spec and API for an easy to use cryptography API called DOMCrypt. "Easy to use" and "cryptography" are normally terms I don't see being used together in the same sentence.
Narenda Sisodiya posted a demo of a slidecasting technique. Included in a later email was snippets of code in how the demo was created.
In the W3C, the big news was around the poll for the Last Call publication of HTML5 and various other documents. If I understand how the poll process works, a quorum of members was reached, and all documents achieved the necessary Yes votes to proceed.
There are formal objections against two publications: the HTML/XHTML Compatibility Authoring Guidelines Specification, and the HTML5: Techniques for providing useful text alternatives guide. The formal objections were that these were guidelines, not technical specifications, and should be published as Notes not Recommendations. However, as noted by Wayne Carr in the poll, a "W3C recommendation is specification or a set of guidelines", in which case guidelines do have a place in the recommendation track.
The Accessibility Task Force did not Formally Object to the publication of the HTML5 Last Call draft, as they stated they would earlier in the week. Instead, the Task Force, and the WAI_PFWG agreed to a conditional support for publication of HTML5. No, I don't know what this means, either, other than no Formal Objection was made to the publication of HTML5.
Next up is the Last Call Review Process. In my opinion, and not reflective of WHATWG, I don't expect the review process to go swimmingly. Using Bugzilla to track Last Call comments seems counter-intuitive, and Bugzilla does have accessibility issues. In addition, the current process makes it extremely difficult to shepherd a tracking issue and change proposal through the process unless you're a member of the group, yet Last Call is supposed to be a time to capture Last Call comments from outside the group.
However, what is, is. This is the process, folks will just have to muddle through.
I do want to note that there are some very thoughtful comments in the HTML WG Last Call poll. If you have a few minutes to spare, I recommend reading through all of them. Somewhat surprised, though, that only about a quarter of the HTML WG members voted.
That's all I have to report. Thanks to the WHATWG folks for giving me the opportunity to participate in the WHATWG weblog and Weekly Reviews.
In the WHATWG email lists:
In April, Kyle Huey proposed a canvas.toBlob function, and Ian Hickson added the feature this week (HTML5 Tracker item for change).
Jer Noble posted a note providing feedback on Mozilla's Full Screen API based on Webkit's implementation of the functionality. This triggered a lively and fairly extended discussion.
Aryeh Gregor noted that pressing Enter in contenteditable generates different behaviors in the different browsers, asking which is the preferred behavior.
Philip Jägenstedt brought up concerns about video...script race conditions and consistent HTML5 video event firing. This is not a new concern, as witness the detailed write up of the problem by Simon Pieters in October, 2010.
Boris Zbarsky questioned the interaction between <wbr> and CSS white space. He questioned the <wbr> requirement that the element trumps the white-space property and provides a line breaking opportunity—something not consistently applied among the browsers.
Over at the W3C:
The W3C has been extremely busy this week with roll outs of new drafts, last calls, and various other announcements. I'm going to defer to Karl Dubost at the W3C Blog for coverage of the individual items in detail, but among the announcements were ones for LC for the second edition of SVG 1.1, the rollout of the first draft for the RDFa interfaces, as well as an announcement of several companies joining the Open Web Platform effort.
The HTML WG co-chairs passed down a decision on Issue 131, having to do with Canvas caret focus API. The co-chairs provided a document diff to be applied, and as of May 15th, the diff was applied and is now waiting verification. Any further concerns about the issue will now be addressed as LC comments.
I'm assuming this HTML5 tracker item is the application of the change.
Janina Sajka passed on a request from the HTML-A11Y Task Force to expedite reconsideration of the Issue 80 decision. This request has to do with asking that title not be used as an acceptable alternative for alt with an image.
The HTML5 co-chairs have published a Last Call Timeline. May 24 marks the beginning of the Last Call review period. To this end, the co-chairs also published an announcement of a LC decision poll (poll results). I'm expecting all active members of the HTML WG to participate in this important poll.
The poll covers six different documents: HTML5, HTML5+RDFa, HTML Microdata, HTML Canvas 2D Context, HTML/XHTML Compatibility Guide, and the HTML5 alternative text guide.
The co-chairs have also asked for consensus on publishing three other working drafts at the same time as the LC publication of the previously listed six documents. These three drafts cover the HTML: The Markup Language, HTML5 diffs from HTML4, and HTML to Platform Accessibility APIs Implementation Guide documents.
That's a lot of documents. For all of the contentiousness in the group, it has been quite prolific.
Speaking of which, the HTML WG co-chairs also published a revised WG Decision Policy procedure this morning.
The WHATWG email list seeming rather quiet this last week.
José Lucas Teixeira de Oliveira brought up the idea of an active menu—associating a nav element with an article in some way.
Ian Hickson caught up with correspondence from last December and January. One of the items concerned the selectionDirection property and especially what the value of 'none' means.
Ian was also busy making HTML5 edits, several of which I wanted to share.
One edit is applying an indeterminate pseudo-class to an indeterminate progress bar, and another clarifies how to make a determinate and indeterminate progress bar. Ian also provided additional adaptive renderings for progress bars, including vertical progress bars, and even a square progress bar.
Another significant edit, the removal of hgroup, triggered a considerable amount of discussion in the W3C HTML WG.
This one went back and forth for a bit. The best summation I found of the entire event was a Twitter posting, referencing a post by HTML5 Doctor:
They put their hgroup in.
They pull their hgroup out.
They put their hgroup in
and shake it all about...
Fun stuff. The long and short of it is that the edit to remove hgroup was reverted, and the issue will be discussed more fully during Last Call.
Also of note in the W3C HTML WG last week was the poll regarding HTML5 licensing. Maciej Stachowiak provided a summary of the results. Interesting reading.
I haven't been covering other W3C groups as much as I should have. I was reminded by another Tweet about a proposed SpellCheck API to check out the Web Application WG email list. In the email list, I noted a very long discussion on the status of WebSQL or whether IndexedDB has replaced it.
In my wanderings about web application land, the status of WebSQL seems to cause the most confusion. It is my understanding at this time that the WebSQL specification is no longer being maintained or pursued.
If you're a web application developer, you'll definitely want to follow the W3C Web Apps email archives. It does have an RSS feed you can subscribe to.
The W3C also kicked off a new working group, the Web Real-Time Communications Working Group, to define client-side APIs to enable Real-Time Communications in Web browsers.
For more of what's happening in the W3C, do check out Karl Dubost's weekly reviews in the W3C Blog.
I'll be ending my stint at the Weekly Review on May 23rd. Hopefully in that time, we'll see a lot of activity.
Simon Heckmann provided a proposal for an application descriptor that triggered a good discussion in the WHATWG email list. Check out the HTML version of his proposal.
Related to the app descriptor proposal, Dave Raggett mentioned efforts by Mozilla and others regarding web apps asking for extra privileges and mentioned that a new W3C WG on this topic will be starting within a few months.
Discussion began anew on a content-disposition property for <a> tags.
A long discussion in the W3C HTML WG was triggered by a patch proposed by Ian Hickson related to a decided HTML WG issue, in this case, I believe it was related to Issue 131.
Point of fact, the discussion on Issue 131 more or less dominated the HTML WG email list, spawning multiple threads, including one on applying the Issue 131 decision and another on a proposed alternative.
However, there was still room for a discussion on another decision related to the meta-generator exception for missing alt text (including an objection to the decision), and two others on longdesc: one on thinking of longdesc outside the box, and another on the expectations related to aria-describedby pointing to a link element.
HTML WG co-chair Paul Cotton re-published the LC schedule for HTML5 and also instructed the HTML5 editor to finish implementations on any remaining decided issues with the exception of Issue 152, and to publish a new working draft incorporating the decisions. The new draft incorporates previous decisions except those for Issue 131, which, as noted earlier, is still under discussion.
The schedule shows that the HTML5 LC WD goes to survey the week of May 15-22, and is scheduled to be published on May 24.
The W3C HTML5 WG co-chairs published a questionnaire on preferred HTML5 license types. As of this morning,
3341 out of 388421 members have responded. The poll continues to May 5th.
The W3C issued several news announcements related to spec drafts and publications:
Moving outside the W3C and WHATWG, CSS3.info introduced us to the new CSS3 Grid Layout module, as well as the release candidate for CSS3 Multicolumn Layout and an updated working draft for CSS3 Text.
Silvia Pfieffer summarized, clarified, and provided demonstrations related the discussions about HTML5 multi-track audio or video that has threaded both the HTML WG and WHATWG email lists the past few months.
Dimitri Glazkov asked a question about pseudo classes for state changes with the audio and video elements so that designers/developers could style the devices without having to listen for events.
Discussion continued on a question that Justin Karneges had about microformats, microdata, and custom data attributes. A point was made in the discussion about not using custom data attributes (data-*) for document interchange, as these are for in-page development, only.
The HTML5 editor, Ian Hickson, made a change to the documentation for video and audio, noting that both elements will accept audio and video content. He's also done more tweaking related to the Media Controller.
In the W3C, RDFa API and RDFa 1.1 Primer Drafts were updated and Web Applications WG published four drafts: WebSocket API, Indexed Database API, File API: Writer, and File API: Directories and System.
The CSS WG published a Call for Review for the CSS 2.1 Proposed Recommendation. The group also published an update of the CSS3 Speech draft.
Dominique Hazael-Massieux posted a note in the HTML WG that the Device APIs and Policy WG would like feedback on a new approach for HTML Media Capture.
The W3C co-chairs posted a possible survey for HTML5 licenses, seeking feedback from the HTML WG.
The W3C co-chairs also published three decisions related to two issues: Issue 31 and Issue 80. Both of these issues have to do with alternative text.
The first decision was related to the requirements section for alternative image text. The co-chairs decided on the change proposal to keep the requirements section in the HTML5 document.
The second decision was about the text used in the img element section of the HTML spec. The co-chairs decided on the following text:
An img element represents an image.
The image given by the src attribute is the embedded content; the
value of the alt attribute provides equivalent content for those who
cannot process images or who have image loading disabled.
The last decision had to do with alternative text validation. The co-chairs decided in favor of a morphing between a couple of change proposals, resulting in the following validation criteria:
- The presence of aria-labelledby does not make missing alt conforming.
- The presence of role=presentation does not make missing alt conforming.
- The presence of <meta name=generator> makes missing alt conforming.
- Use of private communications does not, in itself, make missing alt conforming.
- The presence of title makes missing alt conforming.
- The presence of figcaption makes missing alt conforming.
I'm going to go out on a limb here and offer my personal opinion that it's simpler just to remember to provide alt text.
In the news this week were discussions about Apple and Google auto-logging location history on the iPhone and Android. Appropriately enough, Philippe Le Hégaret posted his position paper for an upcoming presentation at Web Tracking & User Privacy.