IRC log of html-wg on 2007-04-11

Timestamps are in UTC.

00:00:28 [sbuluf]

and as you well pointed, we can't quite keep adding more and more specs, indiscriminately, without thinking on the most precius resource in the whole web: grandma's mind.

00:00:51 [sbuluf]

html? css? xml? rdf? xslt?...how many?

00:01:33 [sbuluf]

*who* really ensures that this does not get out of control in grandma's head?

00:01:59 [Dashiva]

I think most people consider grandma to be the business of the editor creators

00:02:16 [anne]

blogs and such make it trivially for people to start contributing content

00:03:04 [sbuluf]

dashiva, your very sentence worries me. no ofense, no problems, but...see? who is really more important here?

00:03:50 [sbuluf]

anne, partially true. not trivial. and incomplete. but are you going to call our conversatrion "silly" after a while? if so...should i keep talking?

00:04:29 [anne]

sbuluf, heh

00:05:06 [Dashiva]

sbuluf: I'm not sure I can agree. This WG deals with markup, the very markup grandma is not supposed to see. We are completely abstracted away from her world.

00:05:11 [sbuluf]

again, no ofenses intended. really. i hope you can see it.

00:05:40 [anne]

I don't really have an opinion on end users. It's hard for to imagine at what level they'd interact with HTML. I tend to think it would be completely abstracted away from them.

00:05:52 [anne]

for me*

00:06:15 [anne]

*on end users in the context of the HTML language

00:06:32 [anne]

so what Dashiva says I think

00:07:02 [sbuluf]

dashiva, partially true, i'd say. on one hand, we could allow standard makers to just say )as usual) "is someone else's problem. comfy, isn't it? but the end result is this "as of 12007, grandma is in trouble" So should we not start questioning the wisdom of all this? again, who ensures gloabl coherence?

00:07:09 [gavin_]

what does "end users" mean in the context of the HTML specification?

00:07:30 [Dashiva]

sbuluf: But how are we supposed to help grandma? W

00:07:37 [gavin_]

people browsing the web?

00:07:51 [anne]

yeah

00:08:03 [sbuluf]

dashiva, on the other hand...does it not affect the spec? are you sure we could do wymiwyg editors, with today's spec? have you considered the problem? particularly, a UI?

00:08:31 [Dashiva]

As much as I've considered the problem, I consider it practically impossible regardless of the spec

00:08:39 [anne]

whatever editor there will be, it has to export to HTML at some point

00:09:54 [sbuluf]

dashiva, a fair enough answer. but...have we devoted concerted time to explore possibnilities? shouldn't we, at least? i might add i mihgt have some ideas about it. i'm no expert, however, can't do it alone.

00:10:20 [sbuluf]

anne...something not too far away from xopus, might perhaps work. i can expad too.

00:10:36 [sbuluf]

*expand

00:10:38 [Dashiva]

I think a wymiwyg editor is a close cousin of the semantic web

00:10:51 [Dashiva]

They might be possible, but the users form an apparently unsurmountable obstacle

00:11:13 [sbuluf]

dashiva...yep, somehow. do you think that is bad, though?

00:11:29 [Dashiva]

I have no faith in the semantic web, myself

00:11:33 [sbuluf]

noty trivial, dashiva, right.

00:12:31 [sbuluf]

dashiva, is very questioned, and qwuestionable, right. we could discuss it, anytime. i think at least some ideas are sound, or at least desirable, myself.

00:13:16 [Dashiva]

I'll leave the discussion to the pros. Until they make some breakthroughs, I'll remain a grumpy young naysayer :)

00:13:37 [sbuluf]

i'm no expert either =P

00:14:04 [kingryan]

kingryan has left #html-wg

00:14:59 [anne]

yeah, until someone has made a non-wysiwyg editor that's actually widely adopted by "typical" end users there's not much point in debating it I think

00:15:39 [anne]

and afaik people have tried to do that and failed

00:16:22 [sbuluf]

the problem might be impossible with today's scpecs. they were not thought with that idea as a design goal, precisely.

00:16:34 [sbuluf]

but i talked too much here already, so i'll shut up. one last thought: unless wymiwyg edtors, beware of including words about caring about end users.

00:16:43 [Dashiva]

Then the semanticists need to get on the soapbox and tell us what we need to change

00:17:12 [anne]

end users care about wysiwyg typically

00:17:28 [anne]

identical results everywhere and such

00:17:33 [anne]

(visually)

00:17:46 [sbuluf]

(i could answer...i just do not want to abuse)

00:18:33 [Dashiva]

I think we all agree it's a noble cause, we just have different degrees of hope

00:19:35 [anne]

(I also don't think the problem is with the HTML specification. Editors not contrained by HTML also haven't really solved this problem.)

00:20:49 [sbuluf]

anne, we could discuss it, if you wanted. something not to far away from xopus *might* work

00:21:20 [gavin]

gavin has joined #html-wg

00:21:40 [anne]

feel free to e-mail public-html, www-archive or some other list I follow with persuasive arguments

00:23:06 [sbuluf]

good night (i tried that, unfortunately, last time tim berners lee himself ignored me)

00:31:31 [sbuluf]

http://lists.w3.org/Archives/Public/www-tag/2006Nov/0077.html <--for those curious, this is me asking timbl for w3c to make a browser (wymiwyg, or similar, perfect code generation, semantic web ready, etc)

00:32:26 [DanC_lap]

DanC_lap has joined #html-wg

00:32:51 [sbuluf]

hope you do well, danc. we heard.

00:33:02 [DanC_lap]

hi. thanks.

00:34:10 [myakura]

myakura has joined #html-wg

00:34:59 [DanC_lap]

I thought the influx of new WG members would die off a bit after hixie's blog post, but it's been weeks and they continue to come in at about the same rate

00:35:03 [karl]

yes. on 3 discussions on 3 channels

00:35:26 [DanC_lap]

have you sent anything to www-html@w3.org, karl? I don't think I ever did

00:36:17 [karl]

hmmm... you mean recently?

00:36:23 [DanC_lap]

since March 7

00:36:43 [karl]

hmm I guess one or two mails.

00:36:43 [DanC_lap]

RRSAgent, pointer?

00:36:43 [RRSAgent]

See http://www.w3.org/2007/04/11-html-wg-irc#T00-36-43-1

00:36:57 [DanC_lap]

was either of them an announcement about the new HTML WG?

00:37:07 [karl]

I don't think

00:37:10 [karl]

let me check

00:37:16 [DanC_lap]

RRSAgent, make logs world-access

00:38:13 [karl]

http://www.w3.org/Search/Mail/Public/advanced_search?keywords=&hdr-1-name=subject&hdr-1-query=&hdr-2-name=from&hdr-2-query=karl%40w3.org&hdr-3-name=message-id&hdr-3-query=&index-grp=Public__FULL&index-type=t&type-index=www-html&resultsperpage=20&sortby=date

00:38:35 [karl]

3 messages and not announcements about HTML WG

00:39:52 [karl]

the last message was a message about someone asking about HTML 4.01

00:40:01 [karl]

so I said to contact the public-html.

00:40:05 [karl]

but that's all

00:40:22 [karl]

do you want me to announce the HTML WG on www-html?

00:40:32 [DanC_lap]

umm... yes, I'd like that.

00:40:56 [karl]

ok. on my todo pile for today

00:41:59 [karl]

* 305 group participants,

00:41:59 [karl]

* 305 in good standing,

00:41:59 [karl]

* 44 participants from 16 organizations

00:41:59 [karl]

* 261 Invited Experts

00:42:10 [karl]

I have some people queued for today too.

00:42:26 [karl]

I wish we had more Authoring tools developers, and CMS as well

00:42:39 [DanC_lap]

yeah...

00:42:56 [karl]

just to improve the ecosystem, having more diversity and variety of inputs.

00:43:05 [DanC_lap]

any volunteers to write rousing blog articles to encourage CMS folks to join the WG?

00:43:35 [karl]

I could do that on QA Weblog or maybe I wonder if WASP could help on this.

00:43:43 [DanC_lap]

perhaps a message to public-evangelist@w3.org would help?

00:44:12 [karl]

wonder what is the best strategy

00:44:43 [DanC_lap]

I lean toward the shotgun approach: use all of the above

00:44:44 [karl]

maybe QA weblog and asking the community to promote the information and to ask people to bug their favourite tool developers to join

00:45:24 [DanC_lap]

better if we can give reasons why joining will help the CMS vendors. nobody likes to be bugged

00:47:21 [DanC_lap]

I have vague impressions that there are accessibility concerns around <canvas>. Anybody have pointers to coherent arguments?

00:48:58 [karl]

I wonder if we could get people from Pages, iWeb, Mail.app (Apple), FrontPage (Microsoft), Dreamweaver, Golive (Adobe), Textmate, BBedit, etc.

00:50:04 [heycam]

heycam has joined #html-wg

00:50:21 [DanC_lap]

any idea if Mail.app folks will be at the html-email workshop, karl?

00:50:52 [karl]

no idea. I asked maciej if he could give me a contact in Mail.app Team, but no luck :)

00:50:59 [DanC_lap]

the evolution email compose thingy is pretty good. I wonder if they're tuned in

00:51:44 [karl]

and I wonder who is doing the ThunderBird authoring part.

00:51:45 [DanC_lap]

Dreamweaver... is there a good connection between them at WASP these days?

00:52:57 [karl]

hmmm I can ping them

00:54:11 [Philip]

DanC: there are some comments about <canvas> accessibility a little way down http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2006-May/006394.html - I don't think I've seen any other relevant points, though I've not been watching at all comprehensively

00:57:19 [DanC_lap]

thanks Philip

00:58:05 [karl]

maybe Charles McCathieNeville (Opera, ex WAI-W3C) has done an evaluation of accessibility concerns with canvas.

00:58:21 [DanC_lap]

ah; he's an obvious person to ask, in any case

00:58:26 [DanC_lap]

I should do that

00:59:20 [DanC_lap]

I took the versioning thing to www-tag; the answers are mostly "it depends". I guess I should ask a more specific question

01:00:41 [DanC_lap]

going over the differences between http://esw.w3.org/topic/HTML/ProposedDesignPrinciples and /TR/webarch is slow going. I hope I can get more people involved so that it doesn't stop when I take one in the eye

01:00:54 [karl]

QA hat - I have the feeling that versioning is important for documents, in the same way that signing it or dating it is important. Now what certain class of products should do with the version number is a different matter.

01:01:50 [DanC_lap]

simplicity argues for "no; documents should not bear version numbers". I think there are arguments to the contrary, but no strong ones have been articulated yet.

01:02:12 [DanC_lap]

i.e not bothering with version numbers is simpler than bothering with them.

01:02:45 [karl]

one of the reasons in terms of forward compatibility. I have the belief that our design choice of today might not be respected in future generations

01:02:54 [DanC_lap]

as you said, karl, there's an implicit version # in every document: the date it was written.

01:03:53 [DanC_lap]

yes, it's true that design choices of today might not be respected in the future; but the cost of guarding against that risk is considerable, and it's not clear how much effort is worthwhile.

01:05:25 [DanC_lap]

a big part of me says "HTML is that language that everybody that does web stuff agrees to. When that agreement breaks down, HTML becomes worthless. so be it."

01:07:53 [karl]

yep. pretty much the same. I hear different opinions here and there on what is HTML, or should be HTML. My only concerns are when "macho style discussions". Some people will not say interesting things because they are afraid of answers or reputations.

01:08:13 [karl]

my main struggled is how to ensure that everyone has the possibility to express himself/herself

01:08:22 [karl]

s/struggled/struggle/

01:08:22 [DanC_lap]

on a totally different topic: I wonder about a survey about focus areas: test suite development, testing browser X/Y/Z, reviewing the spec as a teaching tool, etc. In fact, I wonder about a separate mailing list for new features. I don't want to hear about them until a proposal has reached some maturity.

01:10:00 [DanC_lap]

or maybe a separate mailing list for test suite development, and I'd focus on that one.

01:11:41 [DanC_lap]

maybe I'll start a wiki topic on tasks for HTML WG members

01:11:50 [mjs]

karl: I think you should assume Apple's reps represent all of our various interests in HTML, and will be in close touch with other teams as appropriate

01:11:56 [DanC_lap]

and after some wiki-brainstorming, turn it into a WBS survey

01:12:06 [mjs]

karl: for Mail in particular, the WebKit team does more work on their HTML composition than they do

01:13:37 [karl]

mjs: could you talk then about the HTML in mail workshop, that would be cool. Daniel Glazman is the chair.

01:13:39 [mjs]

hi DanC_lap, hope you're feeling better

01:13:42 [DanC_lap]

do KDE/KHTML and WebKit share new code these days? I wonder if I should recruit KHTML participation

01:14:07 [mjs]

DanC_lap: there is a Qt port of WebKit, but classic KHTML still exists for now

01:14:08 [DanC_lap]

somewhat better, thanks. sort of a stupid way to get hurt. very painful and scary all the same.

01:14:19 [mjs]

I've asked people from both constituencies and none of them think they have time

01:15:23 [DanC_lap]

you asked fairly recently?

01:15:34 [DanC_lap]

if so, you saved me the trouble ;-)

01:15:45 [mjs]

yes, fairly recently

01:16:39 [DanC_lap]

I should probably ask them anyway... but I should do a lot of things. Since you asked, I think I'll give that pretty low priority.

01:17:18 [mjs]

karl: I'm not sure what W3C workshops actually do, so I'm not sure what I would be getting them into

01:17:43 [mjs]

I do want to make sure that any HTML WG specs for HTML are usable for mail

01:18:07 [DanC_lap]

on a good day, W3C workshops get some exposure to good ideas that deserve more exposure, and get groups who are doing similar work but didn't know it in touch with each other.

01:19:02 [DanC_lap]

they also collect data on quesitons like "how many people care enough about HTML and email to devote a couple days and a plane ticket on it"

01:20:26 [DanC_lap]

they're somewhat random by nature. they're intended to introduce a bit of brownian motion into things.

01:22:23 [DanC_lap]

karl, anything you want/need before I wander off?

01:23:13 [DanC_lap]

oh, and it's been said, but it bears repeating: mjs, thanks for collecting all the bits and pieces about HTML5 and putting it in one proposal

01:23:13 [karl]

nope I think I have already a huge pile of things to iron before the next laundry ;)

01:24:31 [mjs]

DanC_lap: really I was just the secretary for the people whose names are at the bottom

01:25:17 [mjs]

DanC_lap: Apple rarely sends engineers to travel to things so generally it would have to be high value

01:25:23 [DanC_lap]

well, secretarial work is too often thankless work. so thanks.

01:26:04 [DanC_lap]

yeah, travel decisions are tough

02:09:52 [Shunsuke]

Shunsuke has joined #html-wg

02:14:10 [Shunsuke_]

Shunsuke_ has joined #html-wg

02:19:43 [htmlr]

htmlr has joined #html-wg

02:28:06 [gavin]

gavin has joined #html-wg

02:44:22 [Zeros]

Zeros has joined #html-wg

02:46:47 [adele]

adele has joined #html-wg

02:46:50 [adele]

adele has left #html-wg

03:03:24 [foca]

foca has joined #html-wg

03:41:54 [Shunsuke]

Shunsuke has joined #html-wg

03:45:12 [myakura]

myakura has joined #html-wg

03:54:16 [sbuluf]

sbuluf has joined #html-wg

03:58:04 [dbaron]

dbaron has joined #html-wg

04:53:05 [foca]

foca has joined #html-wg

04:54:01 [foca]

foca has left #html-wg

05:35:30 [gavin]

gavin has joined #html-wg

05:45:56 [karl]

http://esw.w3.org/topic/HTMLAsSheAreSpoke

05:51:29 [Zeros]

karl, I'm pretty sure Webkit has a lot of test cases for the HTML parser

05:51:46 [Zeros]

I'm sure Mozilla does too

05:52:37 [Zeros]

So much of its ad hoc when a page that's broken is found too...

05:53:55 [Zeros]

karl, there should be a HTML parser committee of the browser vendors where people can submit inconsistencies or pages that are broken and then everyone can come to a solution.

05:54:35 [karl]

Zeros: yes that's cool. But I'm greedy. I would like to have more than browser vendors.

05:54:44 [mjs]

mjs has joined #html-wg

05:54:54 [karl]

I would like to have people who are *producing* HTML and not consuming *HTML*

05:55:32 [karl]

They have their own constraints, difficulties, and I think their input would be good for the ecosystem of the group.

05:55:46 [Zeros]

karl, ah okay. I'm just for consistency.

05:55:58 [sbuluf]

karl, would contacting news organizations be of use for you?

05:55:58 [mjs]

FWFI, my team at Apple is both a producer and consumer of HTML

05:56:03 [Zeros]

Right now everyone implements their own fixes for how to handle broken content.

05:56:25 [mjs]

WebKit is not only a browser engine but also an HTML editing engine, used among other things to create content in Mail and Aperture

05:56:50 [Zeros]

For example people fix the DOM on bad markup in different ways. HTML needs a big "bug tracker" where failure conditions and solutions can be agreed upon to be implemented.

05:56:52 [karl]

mjs: that is very good to hear.

05:57:37 [karl]

When I edit an HTML mail in Mail.app, does it produce valid HTML or XHTML code?

05:57:51 [karl]

I haven't checked this, I should.

05:57:53 [mjs]

karl: it produces what I hope is valid HTML

05:58:08 [mjs]

although our markup output is a bit wonky at times

05:58:48 [karl]

hehe gavin: invalid English acronym ;)

05:59:59 [Zeros]

karl, no doctype. some other things are a little weird. Does look like its respecting block vs. inline elements.

06:00:19 [Zeros]

Doesn't look like it creates <p> elements, makes a lot of divs, spans and brs

06:00:33 [Zeros]

Interesting it uses -khtml-nbsp-mode: space; and -khtml-line-break: after-white-space; too

06:01:34 [mjs]

we use weird techniques to get whitespace behavior to be what users expect in a text editor

06:02:20 [Zeros]

Oh, that prevents white-space collapsing?

06:02:25 [karl]

mjs, as in more than using CSS?

06:03:46 [Zeros]

karl, the email I have from outlook (maybe exchange server) is worse. <HTML><BODY><PRE></PRE></HTML><!DOCTYPE ... then the document with lots and lots of font tags.

06:04:31 [Zeros]

At least Mail.app is a head of that.

06:09:47 [karl]

hmmm we should send an email on the list, with a kind of simple mail template. with an image and one or two colours. and ask people to test vers HTML implementation in their email clients. Then we could send these files to the list

06:12:05 [mjs]

karl: alternating space and non-breaking space (which all HTML mail composers do afaik) plus a few nonstandard CSS properties to make it work even more nicely in Mail.app

06:14:54 [myakura]

Tb2 generates <!DOCTYPE...><html><head></head><body bgcolor...> and message w/ b, i, u and etc.

06:16:04 [myakura]

breaks in <br>, i see no <font>

06:16:21 [Zeros]

I wonder they don't put the subject line in the <title> element

06:18:27 [Lachy_]

Zeros: probably because mail clients don't pay any attention to the title element anyway

06:18:31 [myakura]

thinking that it's enough w/ Subject:?

06:19:35 [Zeros]

myakura, HTML requires a <title> anyway

06:19:46 [Zeros]

I was just thinking of ways to use it

06:20:35 [myakura]

ah, yeah. totally fogotten about that..

06:34:31 [Zeros]

mjs, is there a way to get Mail.app to groups threads on the subject line?

06:35:40 [Zeros]

People are forking the discussion and changing the subject and mail still groups them, so I have 63+ email long threads that weave in and out of topics.

06:36:10 [mjs]

Zeros: it does have thread grouping, but I don't believe there is a way to customize what rule it uses to put things in a thread

06:38:03 [Zeros]

ah okay

06:38:46 [heycam]

heycam has joined #html-wg

06:46:59 [karl]

Zeros: the thread grouping is made on subject line and not on the References: and In-Reply-To: headers

06:47:02 [karl]

unfortunately

06:47:24 [karl]

which means if a few people are sending mails which are completely unrelated with the same subject:

06:47:28 [Zeros]

karl, In Mail.app its not grouping on the subject line for me

06:47:31 [karl]

Mail.app groups them

06:47:41 [Zeros]

karl, its grouping on something else

06:48:02 [karl]

huh

06:48:42 [Zeros]

karl, http://services.ath.cx/inbox.png is the kind of thing I'm getting

06:48:45 [mjs]

Zeros: I usually see a new thread when someone changes the Subject line

06:48:53 [Zeros]

its pretty weird

06:49:25 [mjs]

I totally don't see all that stuff in the innerHTML thread

06:49:26 [karl]

oh no then it does in-reply-to AND subject

06:50:09 [Zeros]

I wonder why mjs's is different

06:50:12 [karl]

I see why for simple users how it could be useful.

06:50:29 [mjs]

mine seems to be strictly Subject

06:50:29 [karl]

I wish it does only in-reply-to

06:51:25 [karl]

I also wish I could configure my Smart Mailbox with more options (rule A and rule B) and NOT rule C

06:51:43 [karl]

or (rule A or rule B) and NOT rule C

06:52:06 [karl]

or more filtering on headers

06:52:14 [karl]

anyway it is kind of off-topics ;)

06:52:15 [mjs]

you can make more than one rule that has the same result

06:52:27 [mjs]

though that's admittedly a pain sometimes

06:52:45 [Zeros]

Rules can't move into a smart mailbox I don't think

06:53:01 [karl]

rules and Smart Mailbox have differents UI

06:53:20 [Zeros]

I wonder if its something gmail is sending that is causing the grouping this way mjs

06:53:39 [loic]

loic has joined #html-wg

06:53:53 [karl]

basically my mail is in a imap dated space, and all my *topic* mailboxes are smart mailboxes

06:53:56 [tylerr]

tylerr has joined #html-wg

06:54:50 [marcos]

marcos has joined #html-wg

06:55:55 [Voluminous]

Voluminous has joined #html-wg

07:15:56 [karl]

<HTML>

07:15:56 [karl]

<BODY>

07:15:56 [karl]

<HTML><BODY style="word-wrap: break-word; -khtml-nbsp-mode: space; -khtml-line-break: after-white-space; "><DIV>Hi,</DIV><DIV><BR class="khtml-block-placeholder"></DIV><DIV style="font-size: 13px;

07:16:00 [karl]

etc.

07:16:10 [Zeros]

yep

07:16:15 [Zeros]

not valid

07:16:33 [mjs]

what's invalid about that?

07:16:35 [myakura]

no <title> either

07:16:40 [Zeros]

mjs, missing the title

07:17:00 [Zeros]

html, head and body are all optional, but title is required

07:17:33 [mjs]

a <title> is of dubious value for a Mail message, but I suppose it could put an empty one or use the subject line

07:18:25 [anne]

anne has joined #html-wg

07:18:30 [karl]

or it could put the subject of the message itself maybe

07:18:55 [anne]

morning

07:19:45 [Zeros]

karl, have you looked at the css limitations for outlook 2007?

07:20:24 [Zeros]

http://msdn2.microsoft.com/en-us/library/aa338201.aspx

07:21:04 [Zeros]

raises a lot of questions for how useful HTML is going to be; Mail.app in leopard adds html templates, but Outlook 2007 supports rather limited subset of css

07:21:05 [karl]

if I tried to validate forcing the doctype to be HTML 4.01 Transitional

07:21:09 [karl]

i have 12 errors

07:22:08 [karl]

no alt on images, and the missing doctype and title create problems

07:22:48 [mjs]

The doctype and title would be easy to fix (if it does not make other mail clients barf) but adding a made-up alt value would probably do more harm than good

07:23:11 [mjs]

(as would forcing the user to type something whenever they paste an image)

07:23:12 [karl]

mjs: yes for the title it requires more UI modification.

07:23:25 [mjs]

so probably the alt requirement is not appropriate for HTML email

07:23:32 [karl]

it's why I'm always interested by authoring tools requirements

07:23:59 [karl]

s/title/alt/

07:24:04 [mjs]

since the value of making my email to my friends theoretically more accessible is surely outweighted by the cost of typing alt values

07:24:58 [karl]

mjs it depends on the type of friends you have. But it would be better to ask someone who has real needs for accessibility and using mail app

07:25:24 [sbuluf]

does any mail app generate not just html, but also easily separatable css? (to an external stylesheet, or to easily editable UI) if not, shouldn't they?

07:25:25 [mjs]

karl: if I were sending personal email to a blind friend I probably would not include photos

07:25:55 [karl]

sbuluf: the css is put on each element not in a style element

07:26:01 [Zeros]

Or describe the photo yourself below it mjs

07:26:12 [mjs]

and if I had both visually impaired and normally visioned friends that I wanted to Cc on the same email, I'd include the photo and a description in the text

07:26:29 [mjs]

alt seems unlikely to be good for email unless you are mailing a pre-existing web page

07:26:29 [karl]

yes it seems a fair approach

07:27:10 [karl]

mjs: which seems to make the point that there is different class of products with different requirements even for renderers

07:27:49 [mjs]

karl: this seems more like a content generation issue than a rendering issue

07:28:23 [karl]

the HTML document would not be valid in the end if no alt.

07:28:27 [mjs]

you have to either have multiple conformance classes for documents (maybe private vs. public?) or accept that HTML email won't be conforming HTML

07:28:46 [mjs]

or you give up on alt being mandatory

07:28:46 [karl]

I vote for 1.

07:28:52 [Zeros]

mjs, adding a doctype, title and alt="" would make that valid

07:29:02 [karl]

different conformance requirements depending on the products

07:29:13 [Zeros]

Outlook already adds an empty title and a doctype

07:29:18 [mjs]

Zeros: well, adding alt="" would be, in my opinion, more harmful than helpful

07:29:29 [Zeros]

mjs, how does that differ than no alt at all?

07:29:29 [mjs]

while it satisfies the letter of the rule, it surely does not satisfy the spirit

07:29:33 [karl]

conformance requirements are not things to enforce but more a guide on what you need to implement depending on the circumstances.

07:29:43 [mjs]

it's like cargo cult conformance

07:30:03 [mjs]

having an empty-valued alt attribute gives you none of the benefits

07:30:12 [Zeros]

mjs, it prevents inconsistencies. If we don't think alt should be required then it shouldn't be required.

07:30:36 [Zeros]

Adding special markup rules for every type of renderer just makes implementing a HTML UA harder

07:30:45 [karl]

Zeros: I think alt="" would be harmful in this case as mjs says

07:31:03 [Zeros]

I still don't see where its harmful

07:31:12 [karl]

specifically if the mail goes from mail user agent to Web pages.

07:31:15 [Zeros]

no alt and alt="" give the same meaning

07:31:34 [anne]

no it doesn't

07:31:35 [karl]

alt="" means "ignore completely the image"

07:31:40 [edas]

edas has joined #html-wg

07:31:41 [anne]

alt="" means no content

07:31:52 [anne]

no alt means there's no alternative content for the image

07:31:52 [karl]

usually used for spacer and things of this type

07:32:13 [karl]

ooops I have to catch a train.

07:32:15 [Zeros]

anne, how does no content and no alternative content differ?

07:32:17 [mjs]

hmmm

07:32:26 [Zeros]

you just added the word alternative

07:32:34 [mjs]

if empty alt is allowed, then no alt should probably be allowed as well

07:32:53 [mjs]

the point of requiring alt is supposed to be to encourage accessible authoring, but adding alt="" doesn't satisfy that goal at all

07:33:17 [mjs]

anyway, would be good for someone to capture these thoughts on the mailing list

07:33:23 [Zeros]

The validator could still warn, and the spec could encourage its use without making it invalid to not use it.

07:33:31 [mjs]

is <title> the only mandatory element?

07:33:44 [anne]

Zeros, it means that the user agent can tell the user there's an image but no alternate content was provided for it

07:33:55 [anne]

Zeros, for instance, it could read out the URL used to retrieve the image

07:34:03 [Zeros]

anne, it could do the same for alt=""

07:34:12 [anne]

Zeros, that would be non-conforming

07:34:41 [Zeros]

anne, the spec says alt is required. Not having at all is non-conforming.

07:35:13 [mjs]

Web Apps 1.0 allows alt="" but not missing alt, and says missing alt should be treated same as empty-valued alt

07:35:30 [Zeros]

:)

07:35:56 [anne]

That makes the distinction pretty useless

07:36:13 [mjs]

it says empty alt should mean: "In such cases, the image could be omitted without affecting the meaning of the document."

07:36:33 [mjs]

I don't think that will be true for end-user-authored HTML emails that contain images

07:36:46 [mjs]

so adding empty-valued alt would technically be nonconforming

07:37:02 [mjs]

perhaps the alt requirement should be omitted for content generated by a WYSIWYG editor

07:37:14 [Zeros]

mjs, how does that apply to dreamweaver then?

07:38:02 [Zeros]

I'm not sure HTML's rules for conformance should be dependent on who's generating it.

07:38:03 [mjs]

I don't think dreamweaver is the sort of thing intended to be covered by the wysiwyg exception for <font> tags

07:38:54 [Zeros]

mjs, what's the issue with style attributes for fonts?

07:39:22 [Zeros]

in a context where a style block isn't possible I guess, and the font needs to be encoded right there

07:39:35 [mjs]

Zeros: I actually think style should be retained as a global attribute but I haven't heard a lot about the arguments for removing it

07:39:53 [mjs]

the <font> element in HTML5 is the only thing allowed to have a style attribute

07:40:06 [mjs]

and is meant to be used for presentational inline styling done by WYSIWYG editor

07:40:06 [mjs]

s

07:40:18 [Zeros]

mjs, that invalidates a whole lot of existing web content that uses style

07:40:54 [mjs]

yes, it does

07:41:05 [mjs]

(though presumably the rendering section will say UAs must support it on everything)

07:41:17 [Zeros]

I can't say I agree with undeprecating <font>

07:41:50 [gorme]

gorme has joined #html-wg

07:42:23 [anne]

It's undeprecated for WYSIWYG editors as it appears they need it

07:42:42 [mjs]

well, having an official way to do presentational markup seems like the best way to handle the classes of content generators that need to make presentational markup

07:42:43 [anne]

However, that whole section is in its very early stages at this point

07:43:07 [anne]

marcos, contenteditable / designMode geneates it

07:43:10 [anne]

generates*

07:43:22 [marcos]

k, that makes sense then

07:43:38 [Zeros]

contenteditable in safari generates spans IIRC

07:44:24 [Zeros]

mjs, what's to prevent all the content on the web from ending up full of font tags again if its created with in browser WYSIWYG editors?

07:45:06 [mjs]

Zeros: I doubt the front page of cnn.com will ever be made that way

07:45:25 [Zeros]

mjs, blogs probably would be, news sites? wikis?

07:46:11 [sbuluf]

what about spans with only predefined css classes? would that be an improvement?

07:46:18 [mjs]

Zeros: well, you can't get WYSIWYG editors to make purely semantic markup, since the very concept of such a thing is presentational

07:46:24 [anne]

<span class=red>

07:46:36 [anne]

maybe in amount of bytes...

07:46:44 [mjs]

<span class="-user-style-1">

07:47:04 [anne]

however, <font color=red> interoperates better

07:47:48 [Zeros]

And makes restyling the page from a content editor's perspective a nightmare

07:48:05 [Zeros]

anyone saving the source of that document is going to get older styles if they override it too

07:48:17 [Zeros]

font[color=red] { color: blue; }

07:48:22 [mjs]

html mail clients have special requirements, since they have to generate markup that can be handled by a wide variety of really bad layout engines

07:48:33 [gorme]

\o Whats the rationale behind the header and footer element ?

07:48:43 [mjs]

the html engines in many popular mail readers are far worse than popular browsers

07:48:58 [mjs]

gorme: a lot of web documents have a header and a footer

07:49:25 [Zeros]

mjs, How does that come into play with Mail 3.0's HTML templates? Do they cater to those old clients?

07:49:40 [Zeros]

Seems like Apple is pushing HTML mail and MS is fighting it

07:49:41 [mjs]

for instance on http://amazon.com/ you can probably pick it out

07:50:19 [mjs]

Zeros: I can't talk about Mac OS X features that haven't shipped yet

07:50:54 [mjs]

Apple is pushing the possibility of rich text email, because people like to communicate with more than just plaintext

07:51:24 [Zeros]

mjs, Well from your personal perspective. If you were writing a mail client with html templates for users, what audience of UAs would you be targeting and where would you draw the line on support?

07:51:24 [mjs]

I think Outlook switching to the Word engine to compose and display HTML email isn't necessarily a vote for or against HTML email

07:51:33 [Zeros]

Like, would you use <center>?

07:51:33 [mjs]

thogh it is certainly a vote against standards compliance

07:51:58 [mjs]

Zeros: I can't really talk about it at that level of detail, sorry

07:52:08 [Zeros]

alright

07:52:35 [Zeros]

I can see the point of needing some kind of control, but at the same time HTML5 lets you put <style> elements in the body of the document

07:52:53 [Zeros]

So why can't a Visual editor generate that instead?

07:53:08 [sbuluf]

<span class="foreing term> would allow to define another font, another colour, size, and what not. but wouldn't the difference with a font tag be that all those rules can be moved from the middle of the text to an style element, or even to an external stylesheet, hence keeping presentational data separated from content?

07:53:08 [Zeros]

or line styles I guess

07:53:17 [Zeros]

inline*

07:53:41 [anne]

sbuluf, you don't know what you're marking up

07:54:08 [anne]

Zeros, <style scoped> is a pretty new feature

07:54:10 [mjs]

sbuluf: but users don't pick "foreign term" from the Font menu

07:54:16 [mjs]

they pick Italic

07:55:05 [Zeros]

anne, even if it wasn't scoped. Provided the generated content is bound by an id you could target that specific chunk of the document

07:55:05 [sbuluf]

msj, what if you allow users to define classes (a bunch of style rules), and let them name them with semantic names, and then pick those names from a drop down?

07:55:13 [Hixie]

someone should mail the img/alt thing to the whatwg list so i don't forget about it

07:55:29 [mjs]

Hixie: I was gonna

07:55:31 [Hixie]

it does make sense to me to have different rules for e-mails than web content

07:55:52 [mjs]

does email need different rules than WYSIWYG-generated content in general?

07:55:55 [mjs]

(not sure a priori)

07:56:17 [mjs]

so is <title></title> the shortest valid HTML document?

07:56:25 [sbuluf]

hixie, mjs, as a side note...wouldn't setting that info in some wiki be better then just the mailing list? it could later be more orderly transformed into rationale material

07:56:25 [mjs]

or do you need a doctype in there too (for HTML4)

07:56:28 [Hixie]

the distinction is in the receive, not hte producer, imho

07:57:16 [mjs]

well, it's not very useful for a WYSIWYG blog post editor to add alt="" when you past in an image either

07:57:27 [Hixie]

shortest HTML4 document that is conforming is <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"><title>A</title><p>A

07:57:48 [Hixie]

a blog post would have to cater for blind users (e.g. search engines, if nothing else)

07:57:52 [Zeros]

mjs, the body can't be empty

07:58:23 [mjs]

well, it could prompt you at paste time to type a description

07:58:24 [Hixie]

so would need alt text that represented the image (which might be alt="" if the image adds nothing that isn't in the prose, but merely repeats it in a different way)

07:58:30 [Hixie]

alt= is not a description

07:58:32 [mjs]

but it's hard to imagine users prefering such an editor

07:58:43 [Hixie]

the HTML5 spec defines <img> very carefully

07:58:54 [Hixie]

<img> represents _text_ with an _alternative image representation_

07:59:19 [Zeros]

nice wording

07:59:20 [mjs]

that's a weird way to define it

07:59:31 [Zeros]

makes sense to me

07:59:33 [mjs]

but I guess that would make <img> unsuitable for WYSIWYG editors

07:59:56 [mjs]

since when the user drags in an image from their desktop, they definitely do not intend the semantic of "text with an alternative image representation"

08:01:48 [Hixie]

most images on most sites, especially graphics-heavy sites made with WYSIWYG packages, should be in the CSS layer.

08:01:50 [Zeros]

mjs, "If this image doesn't display in the receiver's client, enter text to display instead: " would work for WYSIWYG editors.

08:02:21 [Hixie]

mjs: anyway i'd love to discuss this over lunch sometime, but i'm going afk for now

08:02:25 [Hixie]

ttyl

08:02:26 [Zeros]

"If your mail recipient can't view this image..."

08:02:28 [mjs]

later

08:09:08 [icaaq]

icaaq has joined #html-wg

08:11:46 [ROBOd]

ROBOd has joined #html-wg

08:50:36 [MikeSmith]

MikeSmith has joined #html-wg

09:18:07 [anne]

http://www.digital-web.com/articles/html5_xhtml2_and_the_future_of_the_web/

09:18:14 [anne]

"HTML5 (also sometimes referred to as Web Applications 1.0)" heh

09:24:33 [MikeSmith]

anne - clearly a polemic, though a well written one

09:25:10 [mjs]

I just saw that on digg

09:25:31 [anne]

I think it's accurate though. People more often call it HTML5

09:27:33 [anne]

The comment always say you cannot use this for business stuff etc. but then Y! Pipes is using <canvas> already.

09:28:57 [anne]

Or how browser vendors will support XHTML2

09:29:22 [anne]

At this point it should be pretty clear that no browser vendor has much interest in that

09:29:37 [anne]

s/comment/comments/

09:30:38 [hasather]

hasather has joined #html-wg

09:42:57 [htmlr]

htmlr has joined #html-wg

09:49:01 [heycam]

heycam has joined #html-wg

10:00:12 [gsnedders]

woah. despite having been around David (the author of that article) on several forums and mailing lists for several years, I've _never_ seen a photo of him before

10:04:31 [sbuluf]

i fail to see what is polemic about this article. it might be due to my particular pov, though.

10:19:57 [hasather]

hasather has left #html-wg

10:30:06 [anne]

I think that statement was about my quote and not the article

10:32:42 [sbuluf]

oh, i see thanks.

10:42:37 [hasather]

hasather has joined #html-wg

11:14:14 [xover]

xover has joined #html-wg

11:15:44 [edas]

edas has joined #html-wg

11:53:47 [icaaq]

icaaq has left #html-wg

12:16:33 [hasather]

hasather has left #html-wg

12:18:05 [hasather]

hasather has joined #html-wg

12:24:40 [loic]

loic has joined #html-wg

12:31:52 [claudio]

claudio has left #html-wg

12:32:08 [claudio]

claudio has joined #html-wg

12:36:10 [claudio]

claudio has joined #html-wg

12:36:33 [Philip]

Philip has joined #html-wg

12:44:36 [myakura]

myakura has joined #html-wg

12:47:27 [claudio]

claudio has joined #html-wg

13:04:34 [txm]

txm has joined #html-wg

13:30:23 [txm]

txm has left #html-wg

13:42:11 [anne]

Openwave joined the HTML WG with two participants

13:43:47 [anne]

Mobile browser

13:44:15 [Lachy]

is it any good, or just one of the bad ones?

13:45:09 [anne]

Ask MikiSmith

13:45:15 [anne]

Mike even

13:45:31 [anne]

he worked for them before he joined Opera before he joined W3C

13:46:36 [Lachy]

well, it's good that they've joined. Mobile browsers definately need to participate more

14:01:15 [Lachy_]

Lachy_ has joined #html-wg

14:14:38 [DanC_lap]

DanC_lap has joined #html-wg

14:26:23 [marcos_]

marcos_ has joined #html-wg

14:31:50 [Shunsuke]

Shunsuke has joined #html-wg

14:36:15 [beowulf]

beowulf has joined #html-wg

14:56:43 [h3h]

h3h has joined #html-wg

15:02:09 [bkemp]

bkemp has joined #html-wg

15:14:44 [alexf]

alexf has joined #html-wg

15:20:35 [Charl]

Charl has joined #html-wg

15:33:50 [anne]

anne has joined #html-wg

15:36:50 [anne]

DanC_lap, when would someone be considered neutral?

15:39:17 [DanC_lap]

very rarely ;-)

15:39:32 [anne]

right

15:40:44 [anne]

What I meant with my e-mail is that I think the HTML5 proposal so far caters quite well for both authors and implementors

15:43:43 [DanC_lap]

I suggest you say that more directly next time. I read your message as suppressing discussion that you didn't like.

15:44:08 [anne]

RRSAgent, pointer?

15:44:08 [RRSAgent]

See http://www.w3.org/2007/04/11-html-wg-irc#T15-44-08

15:50:03 [anne]

DanC_lap, well, to be honest, I'm not sure how much these type of architectural discussions regarding how the specification should be outlined make sense given that most of the spec is already there

15:50:12 [anne]

Also, I wouldn't expect typical authors to read the spec

15:52:30 [DanC_lap]

well, I ask that you don't suppress discussion just because you disagree with it.

15:53:44 [anne]

Shouldn't I be able to just state what I think of such discussions on the list?

15:53:48 [Philip]

I used to read small bits of the HTML4 spec fairly frequently, via the links on http://www.w3.org/TR/html401/index/elements.html, as a reference for what attributes go on what element and roughly what they mean. So I think that kind of thing is useful to have for authors

15:53:54 [anne]

but fair enough

15:54:39 [anne]

Philip, agreed, I'd expect HTML5 to get similar indexes much like CSS 2.1, XBL2, WF2, have

16:01:10 [anne]

There's certainly some room for improvement in terms of authoring requirements here and there but in general elements are defined very clearly for authors with examples and all

16:44:48 [krijnh]

krijnh has joined #html-wg

16:55:16 [Zeros]

Zeros has joined #html-wg

17:02:56 [kingryan]

kingryan has joined #html-wg

17:11:48 [zcorpan]

zcorpan has joined #html-wg

17:25:49 [h3h]

I think I implicitly assumed that the spec would be as detailed as possible, largely targeted at implementors, and there would be companion documents like "guides", "tutorials", "references" for web authors, etc.

17:26:32 [h3h]

which is essentially what Dan outlined

17:26:46 [h3h]

but with the assumption that the spec is for implementors first

17:27:15 [zcorpan]

i was thinking about having a view of the spec that hides everything that only applies to implementors, probably using the status script for that

17:27:30 [zcorpan]

but it's hard to figure out with script which parts don't apply to authors

17:27:34 [h3h]

sounds like a maintenance nightmare

17:27:52 [zcorpan]

could be

17:27:58 [h3h]

I'd rather there be a separate document specifically tailored to authors, with plenty of cross-links back to the spec

17:28:34 [h3h]

that would be the document that would get the PR

17:28:34 [zcorpan]

yeah, i think there's a spot in the whatwg wiki for such a thing

17:29:10 [zcorpan]

PR?

17:29:18 [h3h]

public relations

17:29:22 [h3h]

(press)

17:29:24 [zcorpan]

ok

17:29:59 [h3h]

I'd expect how-to sites, blog posts, etc. to link to the guide document rather than the spec itself

17:30:39 [h3h]

so now I wonder if the guide document needs an editor of its own

17:30:54 [zcorpan]

why not have it in a wiki?

17:30:57 [h3h]

it would seem prudent, for consistent tone and clarity

17:31:07 [h3h]

a wiki will always be fragmented and non-official

17:31:19 [h3h]

I think it could easily be developed on a wiki

17:31:21 [zcorpan]

such a document couldn't be normative anyway

17:31:36 [h3h]

but when the spec is published, there should be some published guide version that's been edited for tone, content, etc.

17:32:01 [h3h]

it doesn't need to be normative, just descriptive

17:32:07 [h3h]

it would defer to the spec for all normative issues

17:32:12 [h3h]

with links

17:32:49 [h3h]

and I hesitate to say "document" because I think all of these should be split up into several HTML pages with a table of contents and inter-linking

17:33:12 [h3h]

the worry about the WA 1.0 draft being too large is extremely valid, IMO

17:33:21 [h3h]

it's very cumbersome and clumsy

17:34:04 [zcorpan]

in fact i prefer it being a single document

17:34:45 [h3h]

I wouldn't mind too much if the spec stayed a single document

17:34:54 [h3h]

but the guide definitely needs to be divided into chapters

17:35:00 [h3h]

each on a separate page

17:35:04 [zcorpan]

sure

17:35:19 [h3h]

right now the page weight is 447KB

17:35:24 [zcorpan]

knock yourself out: http://wiki.whatwg.org/index.php?title=Tutorials&action=edit

17:35:44 [h3h]

heh :)

17:36:29 [Philip]

I have a script that extracts some sections from the WA1 spec into a smaller file, so it doesn't take forever to load and crash browsers - I might try modifying that to make a split-up copy of the whole spec, so it's easier to read small parts of it

17:39:41 [h3h]

so I guess the summary of what I'd like in an ideal world would be for this WG's shining public achievement to be a comprehensive HTML 5 Guide, geared specifically toward web authors; the spec itself being a more technical deliverable and not nearly as public

17:40:09 [h3h]

"not as public" meaning "not to receive as much press"

17:40:18 [h3h]

of course it would still be a public document

17:40:53 [DanC_lap]

DanC_lap has left #html-wg

17:44:23 [anne]

I think the HTML spec should be aimed at authors as well

17:44:53 [anne]

I don't expect that most authors will look there, but it should state clearly all the requirements on authors etc.

17:45:07 [h3h]

sure

17:45:18 [anne]

I personally think that tutorials are best left up to the community

17:45:19 [h3h]

but I don't think it should be the primary point of reference for authors

17:45:28 [zcorpan]

why not?

17:45:35 [h3h]

speaking directly to the previously raised concern that the W3C's specs are too technical

17:45:41 [anne]

As everyone has different viewpoints on how such a tutorial should be structured.

17:45:54 [anne]

From the ground up, for authors familiar with HTML4, for authors familiar with XML, etc.

17:45:58 [Hixie]

Hixie has joined #html-wg

17:46:17 [Voluminous]

Voluminous has joined #html-wg

17:46:27 [h3h]

I guess I'm not convinced that one generalized guide wouldn't fulfill all of those viewpoints in a reasonable manner

17:46:43 [h3h]

with the spec as backup

17:46:43 [anne]

Write a guide and contribute it to the WG

17:46:53 [anne]

I suppose there's nothing wrong with having an HTML Primer

17:46:56 [h3h]

if only I were paid to work on that full time, I would :)

17:47:05 [anne]

w3.org/tr/html-primer or something

17:47:27 [h3h]

I wasn't really thinking of a primer

17:47:31 [zcorpan]

you could start at the wiki

17:47:35 [h3h]

primer would be more like "tutorial" for me

17:47:47 [h3h]

and I'm advocating more of a reference-style guidfe

17:48:16 [h3h]

it's all nebulous in my head anyway

17:48:30 [anne]

what's a reference guide?

17:48:39 [h3h]

let me see if I can find an example

17:48:42 [anne]

it describes each feature in simple language + example?

17:50:28 [h3h]

yeah

17:50:32 [h3h]

essentially

17:50:53 [h3h]

in a narrative format, but not in a contiguous linear format like a book

17:51:35 [h3h]

also, "simple language" is emphatically non-technical to a degree

17:51:51 [h3h]

a high school student should be able to read the whole thing and understand everything

17:52:25 [gsnedders]

*cough* there are high school students who are members of the WG, and are in the process of implementing the WHATWG draft */cough*

17:52:39 [gsnedders]

but I guess I'm in the minority :)

17:53:25 [gsnedders]

something like the RSS Profile draft though? <http://www.rssboard.org/rss-profile>

17:53:28 [h3h]

yes, absolutely

17:53:35 [h3h]

er, let me check

17:53:55 [h3h]

eh...they have the TOC + examples down

17:53:58 [h3h]

but there's no narrative

17:54:08 [h3h]

like "when should I use this element/attribute?"

17:54:24 [h3h]

and issues surrounding those types of questions

17:54:39 [gsnedders]

doesn't the "The x element provides the x of the channel" cover that?

17:54:44 [gsnedders]

as to when you should use it?

17:54:50 [h3h]

sure, in spec world

17:55:01 [h3h]

but for a new-to-HTML 5 web author? no

17:55:23 [gsnedders]

how else can you phrase it? "The <title> element provides the title of the feed"?

17:55:39 [h3h]

looking for an example :)

17:56:16 [gsnedders]

or something further like, "The <title> element provides the title of the feed when directly within the <channel>, and the title of an item when directly within an <item>"?

17:58:08 [h3h]

like this... http://www.genevaconventions.org/

17:58:16 [h3h]

but with markup examples, obviously

17:58:27 [h3h]

and yes, more like your last quote

18:00:00 [h3h]

it's like... a glossary mashed up with a TOC, a narrative and examples

18:00:07 [h3h]

that's the best I can come up with right now

18:00:48 [h3h]

it shouldn't be (or look) auto-generated in the least

18:00:56 [h3h]

it should look like it was written as a book

18:01:12 [h3h]

but be devoid of the start-to-finish continuity of a typical book

18:01:51 [Philip]

Is something like http://canvex.lazyilluminati.com/wa1/index.xhtml an approximately useful way of splitting the spec into non-enormous readable chunks?

18:02:29 [Philip]

(Er, I'll change it from XHTML back to HTML when I find an HTML serialiser, to make IE users happy)

18:02:54 [h3h]

Philip: yeah, works for me

18:03:34 [h3h]

Philip: with prev/next/up/home links added

18:04:46 [gsnedders]

Philip: how does it actually split it up?

18:05:10 [anne]

if someone writes a script that does the splitting up and keeps all the references working...

18:05:14 [gsnedders]

x.y sections?

18:05:21 [anne]

we might be getting somewhere

18:05:22 [Philip]

It's splitting on all <h2> and <h3> elements

18:05:39 [Philip]

Navigation links/<link>s would be handy - I'll try add those

18:06:08 [Philip]

anne: I believe that's what I've done :-)

18:06:10 [gsnedders]

it'd probably be better to do it splitting it in other places, but that would require some sort of intelligence

18:06:34 [h3h]

I think that works surprisingly well at first glance

18:06:37 [Philip]

(As far as I can tell, the broken references are the ones that have always been broken, like links to references)

18:06:49 [Philip]

(and the rest get redirected to the right page)

18:08:01 [Philip]

The biggest sections are the parsing and embedded ones - probably the latter could be split up more

18:08:22 [Philip]

(This is from an old copy before the recent <video> changes, so maybe that's got even bigger now)

18:08:47 [zcorpan]

Philip: is this generated by script?

18:09:21 [Philip]

zcorpan: Yes - it's just Python/minidom/html5lib

18:10:34 [zcorpan]

cool

18:11:51 [Philip]

I'll try to fix some bits and upload it later this evening, in case it's helpful

18:12:40 [h3h]

definitely, and reply with a link to the thread that was talking about splitting it up, if you want

18:13:10 [Philip]

Does the list still accept mails from non-members?

18:13:21 [h3h]

no idea

18:13:59 [Philip]

I'll try it and see what happens :-)

18:14:29 [Philip]

*away for a while*

18:14:30 [anne]

public-html should

18:15:25 [anne]

Philip, btw, cool!

18:15:47 [anne]

Philip, maybe you can let Hixie use it so he can generate multiple versions of the spec

18:21:38 [claudio]

claudio has joined #html-wg

18:27:58 [anne]

Philip, should be pretty trivial to get HTML serialization from html5lib

18:28:09 [anne]

Philip, I don't think it's supported by default though

18:30:07 [hober]

hober has joined #html-wg

18:42:18 [jgraham]

Philip / anne: We could certainly add a HTMLSerializer class to each treebuilder imp. without much difficulty (I was kinda planning to rearrange how that stuff works, at least in the case of SimpleTree nyway).

18:45:17 [anne]

if you modify --xml or --hilite you get it

18:45:40 [anne]

oh, you want to redo that, fine

18:47:06 [anne]

in theory you'd just use innerHTML

18:57:22 [Philip]

*away for a while*

18:57:26 [Philip]

Whoops

18:57:30 [Philip]

*back*

18:59:04 [Philip]

anne: I'd be fine with Hixie using it - it sounds like it would probably be worthwhile

19:00:12 [Philip]

I have a HTML5-innerHTML-like serialiser (based on one of the html5lib serialisers) that I did a while ago and which seems to work, so I'll just stick that in for now

19:00:41 [anne]

it will shut down all the people who complain about file size :)

19:00:55 [anne]

what are they thinking reading specs from mobile phones anyway :p

19:01:25 [zcorpan]

oh no! scrollbars! completely inaccessible

19:02:22 [anne]

it's the same person complaining about style sheets

19:02:31 [anne]

iirc

19:02:44 [anne]

maybe he should just get a browser that renders it without scrollbars :)

19:03:17 [anne]

or get a bigger screen

19:06:59 [anne]

anne has joined #html-wg

19:19:25 [anne]

For those interested in XHTML2: http://www.w3.org/mid/op.tqmvnxh3smjzpq@acer3010.lan

19:27:22 [gsnedders]

Philip: the parsing section could also be cut down into input stream/tokeniser/tree builder at least

19:30:09 [Sander]

Sander has joined #html-wg

19:38:22 [Philip]

http://canvex.lazyilluminati.com/wa1/ - now in non-X HTML, plus some links and things

19:39:03 [gsnedders]

re-reading HTML 4.01, it's even more hopeless than I remember

19:39:08 [anne]

maybe drop the top template for subsequent pages?

19:39:25 [gsnedders]

how many elements actually have normative definitions!?

19:39:55 [anne]

see the e-mail from dbaron about HTML4 being a list of desired features

19:40:24 [asbjornu]

asbjornu has left #html-wg

19:40:47 [Philip]

Is it worth keeping the title/logo/copyright/etc on the subsequent pages?

19:40:49 [gsnedders]

yeah, I've seen that already

19:40:58 [anne]

Philip, I think keeping the title and the date might be

19:41:16 [gsnedders]

most W3C specs don't have anything more than previous/next/TOC

19:41:21 [anne]

Philip, for heading consistency and awareness of what you're reading

19:42:18 [anne]

otherwise you'd have to remap the headings as well and fix the style sheets etc.

19:48:32 [Philip]

I've cut down the header now

19:48:52 [gsnedders]

are there actually _any_ implementers involved in XHTML2?

19:50:57 [anne]

someone from Microsoft seems to be on the public-xhtml2 list

19:51:27 [anne]

Access (mobile company) is there

19:52:03 [anne]

IBM and HP (for the XHTML Print Profile I think) too

19:52:17 [anne]

only 13 WG members in total though

19:53:26 [anne]

public-forms has 27

19:53:44 [Lachy_]

the XHTML2WG member list should be made public

19:54:09 [anne]

I don't think that's feasible

19:54:15 [Lachy_]

why?

19:54:30 [anne]

Because people thought their information was shared member only, for one

19:54:39 [anne]

Because some of those people may no longer be around, etc.

19:54:44 [Lachy_]

but now they're a pulic group

19:54:49 [gsnedders]

but if it's been re-chatered…

19:54:59 [anne]

Yes, all information from now on will be public

19:55:00 [gsnedders]

they'd all have to agree to the new charter

19:55:33 [anne]

I expect that w3c-html-wg will become obsolete by the end of April

20:04:10 [gsnedders]

where is a public list of the members?

20:04:35 [anne]

dunno

20:04:35 [Lachy_]

which members?

20:04:39 [gsnedders]

XHTML2

20:04:47 [Lachy_]

it's not public

20:04:51 [Philip]

(Split some sections up now - the biggest remaining is tree-construction at 148KB, which I don't think can be split further, followed by interactive1 (datagrid) at 111KB)

20:05:54 [anne]

Lachy_, oh, you mean the list of members?

20:06:02 [anne]

Lachy_, not w3c-html-wg?

20:06:22 [marcos_]

marcos_ has joined #html-wg

20:06:45 [gsnedders]

but I misread what he said anyway :P

20:07:08 [hsivonen]

interesting that the people who have the most reservations about adopting HTML5 are not (well-known) implementors

20:07:57 [hsivonen]

I wonder if Matthew Ratzloff has specific spec parts in mind that he'd like to reject

20:08:45 [anne]

yeah, what's up with people stating things in generic terms instead of just saying: "I'm opposed to feature X, Y and Z"

20:09:10 [anne]

"... and I don't trust my feedback will be taken care of if we don't do it my way."

20:11:44 [gsnedders]

how many active members will be left when we actually start work?

20:12:08 [gsnedders]

unless the WG completely losses relevance, I'll be staying with it

20:56:36 [gsnedders]

gsnedders has joined #html-wg

21:37:32 [marcos_]

Lachy, you should again point Danc to your presentation on HTML5

21:38:16 [Lachy]

marcos_, why?

21:38:42 [marcos_]

"I'm interested in having someone present HTML5, or the differences

21:38:43 [marcos_]

between HTML4 and HTML5, in a teleconference. "

21:38:56 [marcos_]

I'll get the pointer to the email...

21:39:16 [mjs]

I think a telecon might be about the worst possible medium for a presentation, but might be worth doing if enough people want it

21:39:28 [mjs]

I should hassle my work about hosting an f2f in June or something

21:39:48 [marcos_]

Mjs, lachy has a great MP3 file and powerpoint presentation people can follow

21:40:15 [marcos_]

far better then a teleconf presentation

21:40:26 [Lachy]

http://lachy.id.au/dev/presentation/future-of-html/

21:40:58 [mjs]

oh, cool

21:41:08 [marcos_]

Lachy, maybe add a link to it from the WHATWG wiki page on the differences between HTML4/5

21:43:40 [marcos_]

lachy, this is the email pointer: http://lists.w3.org/Archives/Public/public-html/2007Apr/0529.html

21:44:16 [Lachy]

marcos_, I already sent my reply

21:44:23 [marcos_]

ok cool

21:46:55 [Lachy]

marcos_, the presentation doesn't really belong on the changes from HTML4 page. I'll create an HTML5 Presentations page instead

21:49:09 [marcos_]

sounds good

22:09:51 [zcorpan]

zcorpan has joined #html-wg

22:19:08 [hasather]

hasather has joined #html-wg

22:21:21 [asbjornu]

asbjornu has joined #html-wg

22:25:30 [hsivonen]

the WG could use some kind of "Patent Law for Dummies" as required reading

22:26:15 [anne]

HTML5 as required reading?

22:26:36 [hsivonen]

anne: that too

22:27:13 [anne]

but they're like angle brackets, you can't

22:28:30 [mjs]

it's tragic that as a practicing software technologist, you pretty much *need* to have a basic understanding of patent law

22:29:26 [Dashiva]

It's even more tragic that nothing is being done about it

22:30:04 [Lachy]

Dashiva, what can be done about it?

22:30:29 [kingryan]

improving the laws?

22:30:51 [anne]

dropping them :)

22:30:52 [Lachy]

until the silly politicians wake up and listen to the needs of software developers, nothing will happen

22:31:00 [hsivonen]

Lachy: campaing donations to people who are running for Congrees and who pledge to reform the patent system

22:31:06 [hsivonen]

Congress

22:31:32 [Lachy]

yeah, but I'm not even sure what the patent laws are in Australia

22:31:49 [Philip]

How about a Patent Law 5?

22:31:52 [Dashiva]

Lachy: As I said, tragic :)

22:31:56 [Lachy]

nor how much the US laws affect me, yet that's what all the information I've read is about

22:33:11 [hsivonen]

Lachy: the software market is global. *everyone* needs to understand the U.S. situation. Moreover, once the U.S. makes something a law, they pressure the E.U., Japan, Australia and New Zealand to adopt a similar policy

22:33:39 [hsivonen]

enough politics

22:33:40 [hsivonen]

nn

22:35:55 [Hixie]

wow finally finished my e-mail

22:36:08 [mjs]

finished reading it or writing?

22:36:43 [Hixie]

reading

22:36:59 [claudio]

deadly backlog