Options:
- # Session Start: Sat Apr 21 00:00:00 2007
- # Session Ident: #html-wg
- # [00:04] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [00:09] * Joins: gavin_ (gavin@74.103.208.221)
- # [00:18] * Joins: claudio (claudioc@89.97.35.74)
- # [00:29] * Quits: h3h (bfults@66.162.32.234) (Quit: |)
- # [00:35] * Quits: jgraham (jgraham@81.178.225.78) (Ping timeout)
- # [00:38] * Joins: xover (xover@193.157.66.5)
- # [00:54] * Parts: billmason (billmason@69.30.57.156)
- # [01:21] * Joins: sbuluf (inlh@200.49.140.179)
- # [01:44] * Quits: loic (loic@90.29.123.106) (Quit: hoopa rules)
- # [01:48] * Quits: mjs (mjs@17.255.105.140) (Quit: mjs)
- # [01:52] * Parts: hasather (hasather@81.235.209.174)
- # [01:52] * Joins: mjs (mjs@17.255.105.140)
- # [02:11] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [02:11] * Quits: hober (ted@69.45.6.105) (Quit: ERC Version 5.2 (IRC client for Emacs))
- # [02:16] * Joins: gavin_ (gavin@74.103.208.221)
- # [02:19] * Joins: mjs_ (mjs@17.203.14.247)
- # [02:22] * Quits: mjs (mjs@17.255.105.140) (Ping timeout)
- # [02:39] * Quits: mjs_ (mjs@17.203.14.247) (Connection reset by peer)
- # [02:39] * Joins: mjs (mjs@17.203.14.247)
- # [02:44] <mjs> I nominated a co-editor
- # [02:45] <mjs> looking forward to hearing what our chairs think
- # [02:46] * Quits: kingryan (rking3@66.92.187.33) (Quit: kingryan)
- # [02:46] <Lachy> mjs, is Dave willing to be a co-editor?
- # [02:47] <Philip`> Aha, promoting the Google/Apple conspiracy
- # [02:47] <zcorpan> Lachy: did you read the mail? :)
- # [02:47] <Lachy> ah, I missed the second last paragraph
- # [02:49] <mjs> Philip`: well, Chris Wilson said he doesn't care that much about whether the editor(s) work for a Microsoft competitor, just to have more than once voice
- # [03:00] <Hixie> "inventor of xbl" is more relevant than some may think, given that i then took that invention and specced v2 of it
- # [03:22] * Joins: DanC_lap (connolly@128.30.52.30)
- # [03:40] * Quits: mjs (mjs@17.203.14.247) (Quit: mjs)
- # [04:18] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [04:23] * Joins: gavin_ (gavin@74.103.208.221)
- # [04:26] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
- # [04:52] * Joins: Zeros (Zeros-Elip@69.140.48.129)
- # [05:11] * Joins: mjs (mjs@64.81.48.145)
- # [05:38] * Parts: zcorpan (zcorpan@84.216.43.44)
- # [05:45] <Hixie> man, chris isn't very responsive
- # [05:59] * Quits: Zeros (Zeros-Elip@69.140.48.129) (Ping timeout)
- # [06:00] * Joins: Zeros (Zeros-Elip@67.154.87.254)
- # [06:25] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [06:30] * Joins: gavin_ (gavin@74.103.208.221)
- # [06:37] <heycam> is it correct that all conforming xhtml 1.0 (strict/transitional/whatever) documents are conforming html5 documents?
- # [06:41] <heycam> anyone who has access to edit http://www.w3.org/MarkUp/ may want to fix a typo "XHTML asic" on that page
- # [06:42] <Hixie> no
- # [06:42] <heycam> Hixie, what things would need to be avoided?
- # [06:42] <Hixie> though it somewhat depends on what you mean by "conforming xhtml 1.0 document"
- # [06:42] <Hixie> well, <table border> is no longer conforming, e.g.
- # [06:43] <heycam> with a specific value, ym?
- # [06:44] <Hixie> i mean the attribute is no longer conforming
- # [06:44] <Hixie> <table border="1"> is not conforming HTML5
- # [06:44] <Hixie> but it is conforming xhtml 1.0 strict
- # [06:44] <heycam> oh ok
- # [06:44] <Hixie> (though maybe not xhtml1.1? not sure)
- # [06:44] <Hixie> various other things are changed, though i couldn't tell you off the top of my head what
- # [06:45] <Hixie> there's a wiki entry on it in wiki.whatwg.org somewhere
- # [06:45] <heycam> [also http://www.w3.org/MarkUp/ is served with the wrong encoding]
- # [06:45] <heycam> ok i'll have a look
- # [06:45] <Hixie> MarkUp/ is XHTML2 wg's -- see #xhtml2
- # [06:49] * heycam is wondering what flavour of (x)html to use for xbl2 tests
- # [06:50] <sbuluf> <Hixie> there's a wiki entry on it in wiki.whatwg.org somewhere <--perhaps good time to have these kind of url's handy
- # [06:50] <Hixie> heycam: i'd use HTML4
- # [06:50] <sbuluf> as htese are question likely to pop up often
- # [06:50] <heycam> yeah, but then there are also some tests which will require xhtml (because of xbl subtrees included in them)
- # [06:51] <Hixie> true
- # [06:51] <Hixie> but for those it doesn't really matter
- # [06:52] <heycam> is it likely that an html but not xhtml UA will implement xbl2?
- # [06:52] <Hixie> IE, maybe
- # [06:52] <Hixie> who knows
- # [06:52] <heycam> mm
- # [06:53] <heycam> might start with xhtml, to avoid having some html some xhtml
- # [06:54] <heycam> shouldn't be too hard to convert them html later if needed
- # [06:54] <heycam> (easier than the other way around, probably)
- # [06:56] <sbuluf> Getting IE to accept application/xhtml+xml <--apparently xhtml2 people found a hack to get IE, to accept it
- # [06:57] <sbuluf> http://lists.w3.org/Archives/Public/public-xhtml2/2007Apr/0015.html
- # [07:37] * Quits: myakura (myakura@60.239.122.32) (Quit: Leaving...)
- # [07:46] <Lachy> I don't understand Matthew Raymond's bandwidth argument against the forced bug mode switch
- # [07:46] <Lachy> is he trying to say that including the switch increases the size of the page? I hardly think a few bytes matters that much to be worth mentioning
- # [07:47] <Hixie> it starts mattering a lot when you're sending out a billion pages a day
- # [07:47] <Hixie> (google probably wouldn't add a version switch)
- # [07:48] <Lachy> google uses quirks mode anyway
- # [07:48] <Hixie> exactly
- # [07:48] <Lachy> you could save more by stripping whitespace, omitting end tags and and quites around attribute values
- # [07:49] <Lachy> or even using compression
- # [07:49] <Hixie> google uses compression when the client accepts it
- # [07:49] <Hixie> and yes, there is work that can be done to make it even smaller
- # [07:51] <Lachy> anyway, my point is that bandwidth isn't a significant problem with a bug mode switch, the fact that it's IE's way of introducing forced versioning into HTML is
- # [07:52] <Hixie> certainly the monopoly-sustainment aspect is far, far more dangerous
- # [07:52] <Hixie> but the bandwidth issue is non-trivial with large sites
- # [07:59] <sbuluf> a culture whose most famous game is called "monopoly"...
- # [08:31] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Quit: Leaving)
- # [08:32] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [08:37] * Joins: gavin_ (gavin@74.103.208.221)
- # [09:09] * Quits: DanC_lap (connolly@128.30.52.30) (Ping timeout)
- # [09:27] <hsivonen> heycam: compared to XHTML 1.0, %Flow content models have, in general, become bimorphic for the purposes of document conformance
- # [09:29] <Hixie> ah yes, that's another example
- # [09:30] <Hixie> <div> Hello <p> World </p> </div> is no longer valid
- # [09:33] <heycam> aha
- # [09:46] <Zoffix> nice
- # [09:47] <Zoffix> What does make it invalid though? Having text nodes as children of the <div>?
- # [09:47] <Lachy> mixing inline and block level content with the same parent.
- # [09:48] <Zoffix> Hm.
- # [09:52] <Zoffix> Wouldn't it create a problem with, for example, creating of drop-down menus. I see them done like this: http://paste.css-standards.org/10035 . Should the "Section1" <a> be put into a <p>? I wouldn't say that it is a paragraph though. In case of enclosing it in a <div>, seems like an overkill.
- # [09:55] <Hixie> that one is still valid
- # [09:56] <Hixie> <ol> counts as inline as well as block in the whatwg html5 proposal
- # [09:56] <Zoffix> I should spare some non-existent free time to read it in whole.
- # [10:11] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
- # [10:11] * Joins: hasather (hasather@81.235.209.174)
- # [10:16] * Joins: mjs (mjs@64.81.48.145)
- # [10:18] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
- # [10:21] * Joins: mjs (mjs@64.81.48.145)
- # [10:22] <mjs> hey everyone
- # [10:22] * Joins: ROBOd (robod@86.34.246.154)
- # [10:33] * Joins: loic (loic@90.29.123.106)
- # [10:37] * Joins: olli- (olli@80.203.95.229)
- # [10:39] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [10:44] * Joins: gavin_ (gavin@74.103.208.221)
- # [11:19] <sbuluf> question about versioning: what the necessary and sufficient conditions so that content in a given format can always be machine converted to a new version of the same language ina lossless, unambiguous, meaningful way? or in another words, how can we guarantee that we can always get rid of older versions of a language withour ever loosing our content, and being able to convert it just by pointing a routine to it?
- # [11:33] * Parts: asbjornu (asbjorn@84.48.116.134)
- # [11:35] <Hixie> easiest way is to make the language compatible all the way back
- # [11:35] <Hixie> so we don't have to get rid of anything
- # [11:40] <sbuluf> you can't foresee all possible future ideas. at some point you can conceive a better idea, adn need to throw the old one. the trick would be to ensure you do not loose your content in the change.
- # [11:41] <Lachy> sbuluf, any new idea needs to be refined so that it can be implemented without breaking compat
- # [11:43] <Lachy> you don't need versioning for that, you just need to think harder
- # [11:43] <sbuluf> lachy, one of my design goals, is *methodic*. *systematic* breaking of back compat. so as not to pile garbage along the way. the needed trick is to be able to keep the content
- # [11:44] <Lachy> that might work in closed environments, but would never work well on the web with HTML
- # [11:46] <Hixie> sbuluf: an idea can't be "better" unless it is backwards compatible, by definition.
- # [11:47] <sbuluf> i guess i need to re-read definitions of back and forwards compat
- # [11:49] <Hixie> http://esw.w3.org/topic/HTML/ProposedDesignPrinciples basically everything in the first section
- # [11:50] <sbuluf> ah, links to the tag doc, thanks, hixie
- # [11:50] <Hixie> i meant the bits from Don't Break The Web to Evolution Not Revolution
- # [11:52] <sbuluf> hixie, these are thought for keeping back compat
- # [11:52] * Joins: asbjornu (asbjorn@84.48.116.134)
- # [11:52] <Hixie> yeah
- # [11:52] <sbuluf> what i'm trying to think is the set of conditions that would allow us exactly the opposite
- # [11:53] <Hixie> i can't really think of any set of events that would allow us to do the opposite
- # [11:53] <sbuluf> ito get rid of all obsolete stuff at each change
- # [11:53] <Hixie> except maybe the destruction of the entire internet and all related data networks
- # [11:53] <Hixie> but if that happened i'd have bigger fish to fry than redesigning html
- # [11:54] <Hixie> having said that, of course, the xhtml2 working group is working on a version of html that does indeed do away with backwards compatibility
- # [11:54] <Hixie> they're an open working group as well, you might want to join their work
- # [11:55] <sbuluf> hixie, yes, i'm aware. my question, however, is...is there something ensuring that the same thing qwill not happen again?
- # [11:55] <Hixie> ensuring that what will not happen again?
- # [11:55] <sbuluf> let's say we forget the past, start all from scratch again. can we ensure we will never run into piling complexity/obsolescence again?
- # [11:56] <sbuluf> can we insure that at each major change, we can keep out content along the way, upgradede it mechanically to the latest version?
- # [11:56] <sbuluf> s/out/our/
- # [11:59] <Hixie> wouldn't that just move the complexity from the renderer to an ever increasing set of convertors?
- # [11:59] <sbuluf> it's a possibility, yes.
- # [11:59] <Hixie> you'd end up having to convert each html2 document to html3, then html3 to html4, then html4 to html5, them html6 to html7, etc, up to html38 or whatever you're using
- # [12:00] <Hixie> it seems better to simply have html38 support everything in html2.
- # [12:00] <Hixie> then you don't need to do any converting
- # [12:01] <sbuluf> (i think not, because that way, you would need all the complexity in the UA's, while in the other way, the unnecessary code represented by the convertors, is dettachable
- # [12:02] <sbuluf> hence entry point for new UA's is lower, less room for code complexity in UA's less room for bugs, etc
- # [12:02] <sbuluf> (does that make sense?)
- # [12:03] <sbuluf> in short, complexity due to legacy would be dettachable
- # [12:06] <Hixie> the complexity isn't detachable though
- # [12:07] <Hixie> you'd still want your browser to render html2 docs
- # [12:07] <Hixie> so it needs to have that convertor
- # [12:07] <sbuluf> why so?
- # [12:07] <Hixie> because cnn.com, say, might use html2
- # [12:07] <Hixie> and people want to read cnn.com
- # [12:07] <sbuluf> what would stop cnn from uograding all the time, to live always in the latest version?
- # [12:08] <sbuluf> why would anyone not live in currentversionland?
- # [12:08] <sbuluf> (assuming convertors are always free and available for all, of course)
- # [12:08] <olli-> you mean like myspace that looks like the horrow web of the 90'2 :-p
- # [12:08] <Hixie> they're not going to convert their old content, why would they?
- # [12:09] <Hixie> it's a fact of life that there are millions of web pages out there that people need to read occsasionally that have nobody maintaining them
- # [12:09] <Hixie> they will never be converted to be newer
- # [12:09] <sbuluf> hixie, to access whatever advantages your new version has/allows? new functionality?
- # [12:09] <Hixie> the files aren't changing
- # [12:10] <Hixie> i'm talking about ancient static files
- # [12:10] <Hixie> e.g. a bunch of lyrics that someone put up in 1997
- # [12:10] <sbuluf> ahh, but we could even do automatic conversion
- # [12:10] <Hixie> or academic papers
- # [12:10] <Hixie> who would do it?
- # [12:11] <Hixie> i'm not going to go and convert all my academic papers every time a new html version comes out
- # [12:11] <sbuluf> you put your site up, then die. but you could simply authorize that each time a new version comes out, the convertor be applied, automatically
- # [12:11] <Hixie> what if the convertor has a bug, and it breaks the pages?
- # [12:11] <sbuluf> hixie...were we not assuming that automatic upgrading was possible?
- # [12:11] <Hixie> i don't think automatic upgrading is possible
- # [12:12] <Hixie> in fact, i believe it has been proved to be impossible to generally perform automatic conversion when you have scripts involved
- # [12:12] <Hixie> like javascript
- # [12:12] <Hixie> due to an effect that is commonly known as the "halting problem"
- # [12:12] <Hixie> but which basically means that a program can't analyse another program and work out exactly what it does
- # [12:13] <sbuluf> i see, yes, and scripting worries me, in fact
- # [12:13] <Hixie> (it's not just that it is hard to do -- it's mathematically impossible to do)
- # [12:13] <Hixie> (in the general case)
- # [12:14] <sbuluf> if you saw my idea of total content/presentation/structure/meaning/behaviour separation....well this was my worry about behaviour
- # [12:14] <Hixie> basically all these complications mean that it's just easier to simply have a language that is always compatible with all earlier content
- # [12:14] <Hixie> imho
- # [12:15] * Joins: zcorpan (zcorpan@84.216.40.107)
- # [12:15] <sbuluf> (i'm against indiscriminate use of behaviours, however, and for quite a few reasons. perhaps a much more limited set of commonlu used behaviours could be achieved and used instead
- # [12:17] <Hixie> gotta go, bed time
- # [12:17] <Hixie> nn
- # [12:17] <sbuluf> night, thanks for answers
- # [12:24] <zcorpan> oh, so the charter is the law, eh?
- # [12:24] * zcorpan better starts to ask questions about our charter
- # [12:33] * Quits: olli- (olli@80.203.95.229) (Ping timeout)
- # [12:43] * Joins: kazuhito (kazuhito@222.151.153.229)
- # [12:44] * Quits: sbuluf (inlh@200.49.140.179) (Ping timeout)
- # [12:46] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [12:51] * Joins: gavin_ (gavin@74.103.208.221)
- # [12:55] * Quits: kazuhito (kazuhito@222.151.153.229) (Quit: Quitting!)
- # [13:04] * Parts: asbjornu (asbjorn@84.48.116.134)
- # [13:07] * Joins: kazuhito (kazuhito@222.151.153.229)
- # [13:08] * Joins: Grauw (ask@202.71.92.74)
- # [13:08] <Grauw> "I believe you're saying the definition of what constitutes a corner case is in the hands of the WG. We (Microsoft) have to be in control of our own destiny there. Unless you're suggesting that the WG would shoulder the financial burden when we (Microsoft) are sued because we broke compatibility and caused some company's multi-million-dollar intranet app to break.
- # [13:08] <Grauw> would that be... France? :)
- # [13:08] <Grauw> er, sorry, wrong quote
- # [13:08] <Grauw> "A single MySpace page? Hmm, probably no big deal. A single government who locks us out of their market because we broke their intranet app (even if they were ua-switching and giving us bad content, and it was "clearly their fault")? Probably a very big deal."
- # [13:12] <Grauw> would explain why Chris mentioned that he wasn't too sure about job at some point :)
- # [13:29] * Quits: marcos (chatzilla@203.206.31.102) (Ping timeout)
- # [13:35] * Parts: hasather (hasather@81.235.209.174)
- # [13:36] * Joins: marcos_ (chatzilla@203.206.31.102)
- # [13:36] * marcos_ is now known as marcos
- # [13:39] * Joins: hasather (hasather@81.235.209.174)
- # [13:43] <Grauw> quote Hixie: "Currently the overwhelming body of argument is in favour of
- # [13:43] <Grauw> not having any explicit versioning in the spec (even if you only consider
- # [13:43] <Grauw> feedback from browser vendors, but also if you consider feedback from the
- # [13:43] <Grauw> Web authors who have spoken).
- # [13:44] <Grauw> does he want me to say "+1" or something?
- # [13:44] <Grauw> I thought numbers didn't count, only arguments did
- # [13:45] <xover> Well, sooner or later it boils down to consensus; or majority decision in the absence of substantive objections.
- # [13:45] <Grauw> it's like that on the HTML WG, but supposedly not on the WHATWG list, which Hixie was referring to
- # [13:46] <xover> Oh. Well, my impression was that WHATWG was Hixie's “benevolent” dictatorship. :-)
- # [13:46] <Grauw> well yes, that’s what it boils down to.
- # [13:50] <xover> Hmm. I wonder if we have any ISO liaisons, and what they think of HTML5 redfining conformance criteria and interpretation of ISO-HTML.
- # [13:50] <Grauw> fwiw I understand the need for a version number, given that nobody can tell exactly what content is out there, not even Google’s survey
- # [13:51] <Grauw> it is not there to prevent the spec to reflect the current pages on the web, or at least most of them
- # [13:52] <xover> I must admit I have much greater trouble understanding how it's possible to believe that lack of versioning is in any way responsible or sustainable.
- # [13:52] * Quits: gsnedders (gsnedders@86.139.123.225) (Quit: Don't touch /dev/null…)
- # [13:52] <Grauw> but it would be there to allow browser-specific authored pages to remain working in that specific browser, even though on the web there is functionality that is incompatibly implemented accross browsers and choices must be made
- # [13:52] <Grauw> e.g. <button>’s default type
- # [13:53] <krijnhoetmer> Or button's submitted value :/
- # [13:53] <Grauw> if you go one way, it breaks pages, if you go the other way, it also breaks pages.
- # [13:53] <Grauw> krijn, heh :). I didn’t know that.
- # [13:54] <krijnhoetmer> Grauw: innerHTML in IE you mean?
- # [13:54] <Grauw> innerHTML, eh?
- # [13:54] <krijnhoetmer> That IE submits button.innerHTML, not button.value
- # [13:54] <Grauw> oh, yes, I didn’t know that.
- # [13:55] <krijnhoetmer> http://krijnhoetmer.nl/stuff/html/bug-internet-explorer-submit-button/
- # [13:55] <Grauw> I entirely believe you though :)
- # [13:55] <krijnhoetmer> Watch out with what you say ;)
- # [13:55] <krijnhoetmer> I generaly suck at testing
- # [13:55] <Grauw> eheh
- # [13:56] <zcorpan> http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-April/010909.html
- # [13:56] <Grauw> I suppose backends for a <button name="submit"> currently just do some isset($_GET['submit']) thing to check whether the form is submitted
- # [13:56] <krijnhoetmer> zcorpan: Grauw is a bit behind with the mailing list I think
- # [13:57] <krijnhoetmer> Grauw: Yeah, they probably do, but then you can only use one <button>
- # [13:57] <krijnhoetmer> You can't check which button was pressed
- # [13:57] <zcorpan> krijnhoetmer: that's understandable
- # [13:57] <Grauw> I don’t read the WHATWG’s list, but yes I’m behind on the HTML WG’s :)
- # [13:57] <Grauw> krijn, yeah it sucks
- # [13:58] <krijnhoetmer> Grauw: I figured, since you were copy pasting old stuff :)
- # [13:58] <krijnhoetmer> And my irc client doesn't like utf-8 :/
- # [13:58] <Grauw> but using innerText isn’t really appealing either
- # [13:58] <zcorpan> ☺
- # [13:58] <Grauw> krijn, what client is that?
- # [13:59] <krijnhoetmer> *cough*
- # [13:59] <Grauw> even mIRC ‘gets’ UTF-8 now (I mean, wow!)
- # [13:59] * Joins: Ashe`` (Ashe@213.47.199.86)
- # [13:59] * Quits: Ashe (Ashe@213.47.199.86) (Connection reset by peer)
- # [13:59] <krijnhoetmer> It does?
- # [13:59] <Grauw> as of version 6.2 or something, yes
- # [13:59] * Joins: DanC_lap (connolly@128.30.52.30)
- # [13:59] <krijnhoetmer> Ah
- # [14:00] <krijnhoetmer> I don't like versioning :>
- # [14:00] * krijnhoetmer is now known as krijnh
- # [14:01] <Grauw> I don’t like it either, but I understand the need @ microsoft, and I agree with Chris that it’s impossible to not break anything
- # [14:01] <Grauw> but, I thoroughly dislike using !DOCTYPE as a method for versioning
- # [14:01] <krijnh> Same here
- # [14:01] <Grauw> and I also prefer a scheme that includes a date, like tested-on="2006-08-23"
- # [14:02] <Grauw> because that makes more sense, even though it’ll be a little hard to explain to authors perhaps
- # [14:03] <krijnh> ie-version-at-time-of-writing="8" ?
- # [14:03] <Grauw> But then again, just abusing minor version-numbers in e.g. version="5.01" to actually mean ‘the next version of IE’, I am not particularly bothered about that
- # [14:03] <zcorpan> i think ms will be forced to invent new versioning *mechanisms* each time, because whichever they use can be gamed (e.g. tested-on="9999-01-01")
- # [14:03] <Grauw> problem if you name a browser is that you’re going to have to add a list of those attributes for all browsers
- # [14:03] <Grauw> zcorpan: mwa
- # [14:03] <Grauw> that’s asking for trouble :)
- # [14:04] <zcorpan> Grauw: an using an xhtml1.0 strict doctype with text/html isn't?
- # [14:05] <krijnh> Nope, it isn't
- # [14:05] <krijnh> That's just going to work forever
- # [14:05] <Grauw> <html ie="8.0" firefox="1.0.2.3.5.87.22" icab="4"> that list grows endlessly with the number of browsers
- # [14:05] <Grauw> yup
- # [14:06] <zcorpan> Grauw: you only "need" an attribute for ie
- # [14:06] <Grauw> a date will be sufficient to hook on to, I think, and there’ll probably be web pages documenting exactly what dates you should use if you depend on certain buggy behaviour on browsers X and Y :)
- # [14:06] <krijnh> How about we introduce the new hype; machine tags!
- # [14:07] <Grauw> zcorpan: I’m not so sure about that, Mozilla has held off mimicking certain bugs in IE to get standards compliance
- # [14:07] <xover> I wonder if we should try to review this “Versioning” issue against the Architecture of the World Wide Web document from the TAG.
- # [14:07] <krijnh> <html versioning="buggy:IE=6 buggy:IE=7 working:Fx=2">
- # [14:07] <Grauw> and they have introduced various buggy behaviours under the Quirks mode switch
- # [14:07] <Grauw> hehe
- # [14:07] <zcorpan> Grauw: they said they didn't want to introduce more versioning
- # [14:08] <zcorpan> they might be forced to do so anyway though
- # [14:08] <Grauw> zcorpan, it didn't stop them in the past
- # [14:08] <Grauw> indeed.
- # [14:08] <zcorpan> doesn't make it a good idea
- # [14:08] <Grauw> or, the availability of a version mechanism is too tempting for them
- # [14:08] <krijnh> They don't have to mess around with the intranet stuff MS is worried about
- # [14:08] <Grauw> so krijnh, would that make IE prompt that you have to use Firefox? ;p
- # [14:09] <Grauw> intranet applications are generally created by Java programmers who just got out of high school and have no clue about web design at all
- # [14:09] <zcorpan> krijnh: <html versioning="buggy:IE=6 buggy:IE=7 working:Fx=2"> that's implied isn't it? :)
- # [14:09] <krijnh> zcorpan: ;)
- # [14:10] <xover> Intranet Applications seem to be in general poorly represented on the WG.
- # [14:10] <krijnh> xover: You mean their developers?
- # [14:10] <krijnh> They don't come out in public
- # [14:10] <xover> That user bases' needs.
- # [14:10] <krijnh> I think cwilso is speaking for them
- # [14:11] <Grauw> from the sound of it, he got a significant beating after releasing IE7
- # [14:11] <xover> Put it this way, I've bit the head off vendors for changing the rug I'm standing on even if the change was arguably “correct”.
- # [14:11] <Grauw> even though the web tech community was in general positive, I think
- # [14:11] <Philip`> With tested-on="yyyy-mm-dd", how would browsers decide which cutoff point to use? If I test my site in IE8 RC3 and it works, I'd expect it to work in IE8 Final (whenever that's going to be released) and would want to opt in to its new improved behaviour before it's been released to users, but I wouldn't know what date to use
- # [14:11] <krijnh> On the other hand, why would they introduce <!DoCtYpE hTmL> to their app?
- # [14:11] <krijnh> They won't, so IE can still switch on this doctype
- # [14:12] <Grauw> Philip, you’re right there I suppose
- # [14:12] <Grauw> krijnh, if need be they can, yes
- # [14:12] <Philip`> (and then I might test my site in IE8 Beta or IE8 Alpha or post-IE7 CVS Trunk, and it'd have to cut off at some point which doesn't seem to be clearly definable)
- # [14:12] <Grauw> but what if they want to release IE 8.5 before HTML6
- # [14:12] <Grauw> Philip, but, they can document the dates
- # [14:13] <krijnh> Grauw: IE 8.5 would improve their new standards mode
- # [14:13] <Grauw> maybe tested-on is bad after all, should it be version="yyyy-mm-dd"
- # [14:13] <Grauw> krijnh, well I’d hope they fix bugs in it, yes :)
- # [14:14] <Grauw> can we call it ‘really really standards mode, this time’ instead btw? :)
- # [14:14] <Grauw> oh!
- # [14:14] <Grauw> <html mode="standards">
- # [14:14] <Grauw> <html mode="really standards">
- # [14:14] <Grauw> <html mode="really really standards, this time">
- # [14:15] <Grauw> that would work
- # [14:15] <Philip`> I still prefer <html mode="IE5 bugs">, <html mode="IE7 bugs">, ... because it scales better and it puts the blame in the right place
- # [14:15] <Dashiva> I wonder why Chris insists on arguing for spec versioning at the same time as IE modes. It only makes it easier to disagree with him
- # [14:15] * Joins: gsnedders (gsnedders@86.139.123.225)
- # [14:15] <Philip`> (I don't really mean that syntax - just that way of naming the modes)
- # [14:15] <Grauw> :)
- # [14:16] <Philip`> ((Er, "naming" in terms of talking about it in English, not naming it inside HTML documents))
- # [14:16] <Grauw> I think in the end probably version="5" version="5.01" version="5.02" is the best
- # [14:16] <Grauw> where the browser vendors just increment the minor version number whenever they feel it’s appropriate
- # [14:16] <krijnh> <html standards-for="css2.1 dom3 html5">
- # [14:16] <Grauw> that should sort itself out well enough, I think
- # [14:16] <Grauw> it did with quirks mode
- # [14:17] <zcorpan> people can continue to use CCs to fix bugs in current versions of ie... that way they can fix bugs in ie.next without breaking pages that rely on bugs (and making [if IE] == [if lte IE 7] would help them with current content)
- # [14:17] <Grauw> krijnh: heh, CSS and DOM are implemented very very incrementally
- # [14:17] <krijnh> Grauw: So is HTML
- # [14:17] <Grauw> I suppose it is
- # [14:17] <krijnh> That's why version="5" doesn't make sense
- # [14:17] <Grauw> so we’re back at the date? :)
- # [14:18] <krijnh> Or your mode="standards" :)
- # [14:18] <Grauw> in what I suggested, the version wouldn’t really represent the actual version of the spec that’s used
- # [14:18] <Grauw> just the major version number follows the spec, and the minor the browsers
- # [14:18] <Grauw> or... well anyway, anything will work in the end I think.
- # [14:19] <Dashiva> Then you're basically doing version="5.IE8"
- # [14:20] <Grauw> dashiva, hm, maybe... though something with numbers is easier to compare
- # [14:20] <krijnh> zcorpan: that makes the most sense to me as well
- # [14:20] <Dashiva> And harder to remember, and document. mode=IE8 really can't be misunderstood or forgotten
- # [14:21] <Grauw> the advantage of having a version attribute (as opposed to having none) is that you will only end op with a list that goes like this: ‘check if version attribute, use version, else check against DOCTYPES, if this then quirks, else standards'
- # [14:21] <Grauw> that's true
- # [14:21] <krijnh> Web developers using standards are used to CCs anyway
- # [14:21] <Grauw> so, spec version number dot browser version signature? :)
- # [14:22] <Grauw> what if you want to add multiple browsers?
- # [14:22] <Grauw> space separated list? kinda weird syntax
- # [14:22] <Philip`> (<html mode="really standards"> couldn't work because an old UA that only understands mode="standards" wouldn't know that it should try its best (i.e. "standards" mode rather than "incredibly broken" mode) when it encounters a mode="really standards" page, unless those modes are constructed in some regular pattern like /(really )+standards(, this time)?( I actually mean it now)?( really)*/)
- # [14:22] <Grauw> version="5.IE8 FX1.5"?
- # [14:22] <Grauw> or version="5" browsers="IE8 FX1.5"
- # [14:23] <Grauw> ah yes :)
- # [14:23] <Grauw> I think version="5.number" is still better... a little harder to remember, yes, maybe, but I’m sure it’ll be documented as well as the DOCTYPE switch
- # [14:24] <Dashiva> Why is it better?
- # [14:24] <Grauw> it’s easier to compare
- # [14:24] <Philip`> "documented as well as" = "nobody will understand it and they'll just copy-and-paste from whatever example they happen to have closest to them"? :-)
- # [14:24] <Grauw> so browsers can ‘try their best’
- # [14:24] <Grauw> Philip, right, true :)
- # [14:25] <Dashiva> One simple comparison made easier OR a correct version, easier for authors, more intuitive...
- # [14:25] <Grauw> I suppose for me I’m so used to knowing about quirks and standards mode to realise that it’s all very confusing to new authors :)
- # [14:26] <Grauw> so, version="5.IE8 FX1.5" or version="5" browsers="IE8 FX1.5" then?
- # [14:26] <Grauw> people don’t test in one browser, and different browsers can have different ‘browser-version-tags’
- # [14:26] <Dashiva> Don't need version, don't need non-IE browsers
- # [14:27] <Philip`> I believe Chris said 50% of the top 200 sites use 'standards mode', which could be interpreted as meaning that the choice of quirks vs standards is utterly random and you might as well just throw dice to decide which browser mode you're going to use
- # [14:27] <Grauw> the top 200 sites are created by professional (‘’) web designers
- # [14:27] <Grauw> they probably consciously made the decision that yes, they want standards mode, because quirks mode sucks eggs
- # [14:28] <Grauw> however, even though they want standards mode, they did not realise they also opted in for ‘eternal maintenance’
- # [14:28] * Quits: kazuhito (kazuhito@222.151.153.229) (Quit: Quitting!)
- # [14:28] <Grauw> because you have to jump in and repair whatever’s gone broken everytime a major browser decides to fix its broken stuff real good
- # [14:28] <Philip`> But that means half of the professional web designers consciously made the decision to suck eggs
- # [14:28] <Grauw> Philip, yes, they do.
- # [14:29] <Grauw> or the sites predate the quirks mode, of course
- # [14:29] <Philip`> which seems kind of depressing :-)
- # [14:29] <krijnh> Jup, they do :)
- # [14:29] <Grauw> I can tell you the exact reason
- # [14:29] <Grauw> IE5
- # [14:29] <Philip`> If they're the top 200 web sites, I assume they're still actively maintained
- # [14:29] <Grauw> They want to support IE5, and if you want to do that with one stylesheet, it’s easiest to go just for quirks mode.
- # [14:30] <krijnh> It is?
- # [14:30] <Grauw> because then all browsers including IE5 render it the same.
- # [14:30] <krijnh> Ow, with one stylesheet
- # [14:30] <Philip`> Ah, that seems sensible
- # [14:30] <Grauw> however I think with IE7 the time now approaches that IE5 can be ignored, somewhat
- # [14:30] <Grauw> there’s tons of sites out there now that don’t render well in IE5
- # [14:30] <krijnh> Yeah, because IE5 users upgrade to IE7..
- # [14:30] <Grauw> no, because their machines died :)
- # [14:31] <krijnh> That has nothing to do with IE7
- # [14:31] <Grauw> Backbase has dropped support for IE5 officially as of their next version, although they still support IE5.5
- # [14:31] <krijnh> Probably with Windows running on their machines
- # [14:31] <Philip`> Presumably the most popular sites care more than others about supporting IE5, since the cost is constant (you just put up with the cries of pain from your web designer) but the value is proportional to the number of visitors you have
- # [14:31] <krijnh> Grauw: ah, sweet
- # [14:31] <Grauw> well, it has more with Windows XP (including IE6), true
- # [14:31] <Grauw> philip, yes
- # [14:31] <krijnh> Grauw: Now go bake in that support for Opera ;>
- # [14:32] <Grauw> Backbase 4 will hav eit
- # [14:32] <Grauw> unofficial, but it’s actually better than the support for Safari (which is official)
- # [14:32] <Grauw> or so I heard
- # [14:33] <Grauw> because Safari sucks more than Opera :) (with no quantitive statement made on how much or little Opera actually sucks)
- # [14:33] <Grauw> (it doesn’t that much, Opera 9 really fixed a lot)
- # [14:33] <krijnh> Or Backbase just sucks ;)
- # [14:34] <Grauw> >.<
- # [14:34] <krijnh> There goes my probability of ever working there ;P
- # [14:34] <Grauw> meh :)
- # [14:34] <Grauw> Backbase 3 sometimes sucked, Backbase 4 is much better :)
- # [14:35] <Dashiva> There's also a difference between supporting and looking pretty doing so :)
- # [14:35] * Grauw wonders if his home page ‘supports’ IE then
- # [14:36] <Grauw> I don’t care about IE at all, except that its users can read the text
- # [14:36] <Grauw> does that mean I support it (it looks kinda bad, even in IE7), or it doesn’t?
- # [14:36] <Dashiva> Should at least be able to read the text, click the buttons, and have nothing overlap.
- # [14:36] <krijnh> It's working in IE7 Grauw
- # [14:37] <Grauw> krijnh: look at the last post’s ‘reply’ button
- # [14:37] <Grauw> IE box weirdness there
- # [14:37] <Grauw> in IE6 it’s worse
- # [14:38] * Quits: ROBOd (robod@86.34.246.154) (Client exited)
- # [14:39] <krijnh> Ow well
- # [14:39] <krijnh> What's the conclusion for today?
- # [14:40] <Grauw> conclusion: 383 messages left to read, too much mail :)
- # [14:40] <krijnh> Only from public-html? :|
- # [14:41] <Grauw> yes, I started at 400-something
- # [14:41] <krijnh> Pff
- # [14:41] <krijnh> Quit your fulltime job and become a freelancer then ;)
- # [14:41] <Grauw> I think from now on I'm going to skip the ones that are not from people I know :)
- # [14:41] <Grauw> I’ve never worked full-time (well, except during summer holidays)!
- # [14:42] <krijnh> Ow, hehe
- # [14:42] <Grauw> Still studying. Somewhat...
- # [14:42] <krijnh> Aren't we all?
- # [14:43] <Grauw> :)
- # [14:46] * Joins: ROBOd (robod@86.34.246.154)
- # [14:47] <krijnh> Right, I'm gone, have fun catching up Grauw ;]
- # [14:48] <krijnh> (Wasn't there some kind of blog which wrote down the daily highlights?)
- # [14:48] <Grauw> oh, is there?
- # [14:48] <Grauw> and thanks :)
- # [14:48] <Grauw> (I guess)
- # [14:48] * Parts: nickshanks (nicholas@195.137.85.17)
- # [14:48] <krijnh> http://i-yudai.blogspot.com/
- # [14:49] <krijnh> Probably means no highlights the last days?
- # [14:50] <Grauw> I wouldn’t want to be the person keeping that up :)
- # [14:54] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [14:56] * Parts: hasather (hasather@81.235.209.174)
- # [14:58] * Joins: hasather (hasather@81.235.209.174)
- # [14:59] * Parts: hasather (hasather@81.235.209.174)
- # [14:59] * Joins: gavin_ (gavin@74.103.208.221)
- # [15:01] <Grauw> I think Jonas Sicking wrote a very good mail, at http://lists.w3.org/Archives/Public/public-html/2007Apr/1049.html
- # [15:01] * Quits: DanC_lap (connolly@128.30.52.30) (Ping timeout)
- # [15:03] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
- # [15:06] * Joins: myakura (myakura@60.239.122.32)
- # [15:06] * Grauw skips all <indent> vs. <blockquote> discussions because it’s silly
- # [15:07] <Grauw> 300 messages left
- # [15:08] <Dashiva> I don't think they've addressed each other's arguments yet
- # [15:08] * Joins: hasather (hasather@81.235.209.174)
- # [15:08] <Grauw> who, the indent and blockquoters?
- # [15:09] <Dashiva> Yeah
- # [15:10] <Grauw> well, I haven’t read any of it really, but indenting isn’t part of content, belongs in the styling. I personally thought pretty much everyone by now had stopped using blockquote for indenting anyway. But apparantly education needs to continue.
- # [15:11] <Grauw> it’s just a core principle that I don’t think the spec should torn with
- # [15:12] <Dashiva> That sounds eerily like the con argument
- # [15:14] <Grauw> I don’t really know what the discussion is about, but I don’t look forward to re-introducing presentational attributes
- # [15:14] <Grauw> zomg two other huge versioning discussion (this time, a "compromise" and "it should be an attribute, not doctype")
- # [15:30] <Grauw> David Baron complains about ‘chameleon namespaces’ (e.g. what XForms does in XHTML2), actually the new XML Events 2.0 draft also introduces support for chameleon namespaces in XML Events, presumably to integrate better with XForms
- # [15:31] <Grauw> but yes, it is not good. It violates why namespaces were introduced in the first place.
- # [15:32] <Grauw> Example: XML Events 2.0 is not backwards compatible with XML Events 1.0. Why? They renamed the ‘target’ attribute to ‘eventTarget’ or something like that
- # [15:32] <Grauw> presumably because it conflicted with the ‘target’ attribute in a namespace it was supposed to be chameleon-ed into (that is, XForms)
- # [15:34] * Parts: hasather (hasather@81.235.209.174)
- # [15:40] * Joins: hasather (hasather@81.235.209.174)
- # [15:42] * Quits: marcos (chatzilla@203.206.31.102) (Quit: Chatzilla 0.9.77 [Firefox 2.0.0.3/2007030919])
- # [16:06] * Joins: WebSam (nikos@213.7.62.81)
- # [16:06] <WebSam> hello
- # [16:07] <zcorpan> heya
- # [16:07] <WebSam> Can i ask you something ?
- # [16:08] <zcorpan> shoot
- # [16:08] <WebSam> Which one is better : html or xhtml ?
- # [16:09] <Grauw> xhtml!
- # [16:09] <Grauw> or html
- # [16:09] <Philip`> or neither
- # [16:09] <Grauw> it's kind of a complex issue.
- # [16:10] <Grauw> but you can add /-slashes to <img/> tags if you want
- # [16:10] <zcorpan> WebSam: neither, although xml has drocanian error handling (if that's a good thing or not is up to you) and ie and google don't support xhtml (again, up to you)
- # [16:10] <zcorpan> WebSam: you can use xhtml-ish syntax in html5
- # [16:10] <Grauw> IE and Google only support XHTML if served as text/html with the guidelines presented in appendix C of the specification
- # [16:11] <zcorpan> Grauw: well, that's html in my book :)
- # [16:11] <Grauw> *shrugs*
- # [16:11] <zcorpan> i mean, by that definition mosaic supports xhtml
- # [16:12] <Grauw> that’s true, but saying it doesn’t support kind of suggests that it wouldn’t work at all and is unusable
- # [16:12] <Grauw> and I do not think that’s the case
- # [16:12] <Grauw> anyway, look at the discussion WebSam ended up in :)
- # [16:12] <Philip`> I tried using XHTML as HTML once, and it worked until I had an empty element like <span style="background:red"/> and the whole page went red :-(
- # [16:12] <WebSam> So i must use html ?
- # [16:12] <WebSam> background-color
- # [16:12] <Grauw> WebSam: using HTML is simplest
- # [16:12] <zcorpan> WebSam: if you want to support ie and google, then effectively yes
- # [16:13] <Grauw> Philip: that’s not written according to the appendix C guidelines :)
- # [16:13] <WebSam> Ok thanks very much
- # [16:13] <Philip`> and then it was just as easy to write an HTML serialiser, rather than an XML-with-compatibility-guidelines-from-appendix-C serialiser :-)
- # [16:13] <Grauw> XSLT 2.0 has an ‘XHTML’ serialiser
- # [16:13] <Grauw> that’s very easy.
- # [16:14] * zcorpan never understood the concept of an "xhtml serializer"
- # [16:14] <Philip`> I was using Python's minidom which didn't have an obvious one of those
- # [16:14] <WebSam> I think i must stop learning web design and start learning programming
- # [16:14] <zcorpan> surely you either serialize as html or xml? what's the point in having a format that can be interpreted as either?
- # [16:14] <Grauw> zcorpan: that’s a serialiser for XHTML compatible with appendix C and browsers like IE
- # [16:15] <zcorpan> Grauw: yeah i know what it is, just don't understand the concept
- # [16:15] <Grauw> zcorpan, because the XML serialisation won’t work everywhere, and HTML serialisation has syntactical differences with the XHTML one and can not be processed by XML tools
- # [16:15] <Grauw> with the XHTML serialiser you get best of both worlds
- # [16:16] <Grauw> I don’t see how that’s a difficult concept to grasp
- # [16:16] <WebSam> but my wonder is why did they create xhtml and not html 5
- # [16:16] <Grauw> WebSam we’re creating it now
- # [16:16] <Grauw> XHTML has the same tags as HTML 4
- # [16:16] <zcorpan> yeah, that should explain it. however with the html5 parsing spec it's pretty redudant, you could just use an html parser instead of an xml parser in front of the xml toolshain
- # [16:16] <Grauw> but a slightly modified syntax to be a well-formed XML language
- # [16:17] <WebSam> make it easy
- # [16:17] <Grauw> some people (like me) think it’s ‘cleaner’ and makes it easier to find errors
- # [16:17] <WebSam> what about xhtml 2?
- # [16:17] <Grauw> zcorpan: for every toolchain with an HTML parser there will be a 100 that only have an XML parser
- # [16:18] <Grauw> WebSam: XHTML 2 is kind of an experiment, and not implemented nor recommended, it’s only a working draft
- # [16:18] * Joins: asbjornu (asbjorn@84.48.116.134)
- # [16:18] <WebSam> hm ok thanks
- # [16:18] <Grauw> how it will proceed in the future is currently unknown, now that HTML5 is being developed
- # [16:18] <Grauw> (and XHTML5 at the same time)
- # [16:18] <zcorpan> Grauw: that's probably how it is now, but i'd expect most systems to have html5 parsers and serializers in due course
- # [16:19] <Grauw> maybe in 10 years :)
- # [16:19] <WebSam> I will continue to use XHTML . I like it and i get used to it
- # [16:19] <Philip`> Grauw: Are those toolchains usually extensible, in that you could just stick some HTML5 components onto the ends? (I don't know what XML toolchains are like in the real world)
- # [16:19] <hsivonen> Grauw: sooner if the cunning plan of sneaking one into libxml2 succeeds
- # [16:19] <zcorpan> WebSam: that's fine
- # [16:20] <WebSam> but w3 validator will validate it ?
- # [16:20] <Philip`> *you could easily just stick ...
- # [16:20] <Grauw> Philip: if they’re in Java they could be (if you’re lucky and they’re open source, or allow you a mechanism to switch the factory)
- # [16:20] <zcorpan> WebSam: don't understand the question
- # [16:20] <Grauw> it will, websam
- # [16:21] <Grauw> both xhtml and html
- # [16:21] <zcorpan> they might fix bug 1500, which means that text/html will be validated as html (or perhaps just bug about appendix c or something)
- # [16:22] <Grauw> philip: but like say, if Prince didn’t support HTML but only XML (note: it’s actually listed as a feature), and I had purchased a license for version 5
- # [16:22] <zcorpan> WebSam: you could use http://hsivonen.iki.fi/validator/
- # [16:23] <Grauw> even if a future version would then implement a HTML parser, I’d still be stuck with my old version (the licenses can be expensive)
- # [16:23] <Grauw> in reality, Prince supports HTML parsing, but that’s not surprising because it’s a very ‘web-oriented’ tool
- # [16:24] <WebSam> but w3schools.com recommends XHTML
- # [16:24] <Grauw> various generic XML tools though will be much less likely
- # [16:24] <Grauw> WebSam: XHTML, HTML, either is ok, they’re essentially the same except for a few syntactic differences
- # [16:24] <Grauw> if served as text/html, which you will likely be doing
- # [16:24] <WebSam> w3schools.com was converted to xhtml too
- # [16:25] <hsivonen> WebSam: w3schools is bogus
- # [16:25] <zcorpan> WebSam: w3schools can be ignored. they have nothing to do with w3c
- # [16:25] <Grauw> they’re not official
- # [16:25] <zcorpan> they have lots of misguided information
- # [16:26] <Philip`> Grauw: Ah, okay - I suppose in some easy cases you could run "cat index.html | html2xhtml | xmlprogram | xhtml2html > index2.html" but it's much harder when programs don't/can't work like that
- # [16:26] <Grauw> philip, ah well, I suppose html2xhtml or xhtml2html would make things easier
- # [16:27] <Grauw> but will they e.g. resolve entities or not
- # [16:27] <Grauw> and the UI of Prince doesn’t support running it through html2xhtml first of course (they also have a commandline version, but still)
- # [16:27] <WebSam> oh w3schools and w3c are not the same ?
- # [16:27] <zcorpan> WebSam: no
- # [16:27] <WebSam> sorry
- # [16:28] <Zoffix> WebSam, no, w3schools site mentions that they are not affiliated with W3C.
- # [16:28] <WebSam> lol Zoffix
- # [16:28] <Philip`> I don't understand the details well, but I'd assume they'll have to resolve - XHTML5 doesn't have a DTD so it couldn't use the non-XML entities, and it'd get lost in the HTML->XHTML->HTML conversion
- # [16:28] <Grauw> WebSam: I use w3schools myself sometimes though, because it’s a somewhat easy read to get started with a certain technology, easier than reading the official specification, and some of their references are OK
- # [16:29] <Philip`> (That's why http://www.whatwg.org/specs/web-apps/current-work/multipage/ has Unicode characters where the single-page version has – and ©)
- # [16:29] <Dashiva> Hixie shuns the 8th bit when writing
- # [16:29] <Grauw> Philip`, yeah because otherwise the XML processor can’þ understand them. But the output wouldn’t contain the entities anymore, so that would mean that suddenly the encoding of the page can matter a lot.
- # [16:30] <hsivonen> Grauw: The Encoding is UTF-8 and that's that
- # [16:30] <WebSam> what are the encoding iso.... and utf... ?
- # [16:30] <Zoffix> Should the backtick (`) character be escaped inside attribute values? I cannot get the same response from people.
- # [16:30] <Grauw> e.g. if the input contains ü for ü, encoding doesn’t really matter to make it display right. However if the tools convert it to ü in UTF-8 encoding...
- # [16:30] <Zoffix> With UTF-8 encoding
- # [16:30] <hasather> Zoffix: no
- # [16:30] <Zoffix> Thanks.
- # [16:31] <Philip`> (Hmm, should the multipage spec have <meta charset="UTF-8">? The server sends the right charset now but maybe it'd get confused if people saved local copies...)
- # [16:31] <Grauw> WebSam, Zoffix: this is not a Q&A forum though, this is a channel for specification development and there are better places to ask your questions
- # [16:32] <WebSam> Zoffix told me to come and ask you
- # [16:32] <zcorpan> Grauw: really? where are the better places?
- # [16:32] <Grauw> hsivonen: tell that to the kazillion of pages out there using some ISO-8859-x encoding
- # [16:32] <WebSam> OK thanks very much
- # [16:32] <WebSam> bye
- # [16:32] <WebSam> i am late
- # [16:32] <Grauw> ok :)
- # [16:33] <zcorpan> cya
- # [16:33] * Quits: WebSam (nikos@213.7.62.81) (Quit: WebSam)
- # [16:33] <Grauw> we get a lot of questions about encoding with Backbase, we try to tell people to use UTF-8 at all times, but it’s a big subject to try to explain in a simple forum response
- # [16:34] <Grauw> it also involves things like HTTP headers etc.
- # [16:34] <Zoffix> WebSam, Err... I explained to you why using XHTML is pointless in many situations, you didn't believe me so I've said ask the "people who know" if you don't believe me. Don't go around telling people that I invite you here and there.
- # [16:34] <Philip`> Dashiva: I guess there are slightly different constraints when writing - you have to deal with text editors and SVN and things, which are more likely to break with UTF-8. Hopefully it doesn't matter for auto-generated HTML pages since you're not going to be doing anything other than serving to a web browser (which'll understand the HTTP charset)
- # [16:34] <hasather> Grauw: create a FAQ
- # [16:35] * hsivonen was once about to create a FAQ about form submission encoding where the only answer to everything would have been "Use UTF-8"
- # [16:35] <Zoffix> zcorpan, there are plenty of knowledgeable people on, for example, irc.freenode.net in #web for example.
- # [16:35] <Philip`> One nice advantage of UTF-8 is that it's a much easier name to remember than ISO-8859-1, so you won't accidentally write ISO-8559-1 and get confused by it not working
- # [16:36] <Grauw> I never remember ISO-8850-15 :)
- # [16:36] <Grauw> oops, mistake there too
- # [16:36] <Grauw> text editors and UTF-8, yes, don’t get me started :)
- # [16:37] <hsivonen> one should not use ISO-8859-15. it is a politically motivated anachronism
- # [16:37] <Grauw> if I don’t use ISO-8859-15 I can’t write an € character
- # [16:37] <Grauw> unless I use win-1252 (which everyone uses anyway)
- # [16:37] <hsivonen> Grauw: UTF-8 or Windows-1252
- # [16:37] <zcorpan> iso-8859-1 maps to windows-1252 in browsers anyway
- # [16:38] <Zoffix> Grauw, yes, I am aware that this is not a Q&A forum. The only reason that I've asked is because I've asked plenty of other people and their responses did not agree.
- # [16:38] <Grauw> as for auto-generated pages, the database might store the text in Latin-1 (that I remember :)) so you’d have to tell people how to convert across character sets in their backend
- # [16:38] <Grauw> zoffix, ok
- # [16:39] <Grauw> Q: I'm rendering a list of countries into a combobox, but I get XML Parsing error with this line:
- # [16:39] <Grauw> <b:combo-option b:value="147">Saint Vincent & The Grenadines</b:combo-option>
- # [16:39] <Grauw> A: Use & instead of &..
- # [16:39] <Grauw> Q: Thanks for your reply. But that's not a solution... I'm getting data from the database, and I can't change all that data. I'm getting the same error with "special" characters like accents, ñ, ç ... How can I avoid that error? Backbase is not able to deal with this chars?
- # [16:40] <Grauw> A: long answer including references to PHP function calls
- # [16:40] <Grauw> (in essence, htmlspecialchars at first, and later htmlentities, the latter so that the encoding issue didn’t have to be explained)
- # [16:40] <hsivonen> Grauw: A: Don't use PHP. :-)
- # [16:41] <Grauw> :)
- # [16:41] <Grauw> well for my web server it’s either that, Perl or Python
- # [16:41] <Grauw> people don’t always have a choice
- # [16:41] <Philip`> s/(.)/'&#'.ord($1).';'/seg;
- # [16:42] <Philip`> Not the most efficient method, I guess...
- # [16:42] <Grauw> I’d love to have a Java server, but budgetary considerations unfortunately don’t let me have one :)
- # [16:44] <Grauw> run Cocoon... generate pages with XSLT... ahh... :)
- # [16:44] <Grauw> anyway
- # [16:45] <Grauw> I munged through all 500-ish messages, so I’m going to sleep
- # [16:45] <Philip`> Have fun with the next hundred that will arrive while you're sleeping ;-)
- # [16:46] <Grauw> sniff
- # [16:46] * Quits: loic (loic@90.29.123.106) (Ping timeout)
- # [16:46] <Grauw> nite, it’s been a pleasure discussing :)
- # [16:46] * Quits: Grauw (ask@202.71.92.74) (Quit: +1)
- # [16:48] * Joins: Shunsuke_ (kuruma@219.110.80.235)
- # [16:49] * Shunsuke_ is now known as Shunsuke
- # [16:58] * Joins: Deeder (Deeder@83.204.57.241)
- # [17:01] * Joins: loic (loic@90.29.30.52)
- # [17:01] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [17:03] * Quits: gsnedders (gsnedders@86.139.123.225) (Quit: gsnedders)
- # [17:04] * Quits: Shunsuke (kuruma@219.110.80.235) (Ping timeout)
- # [17:05] * Joins: gsnedders (gsnedders@86.139.123.225)
- # [17:06] * Joins: Shunsuke (kuruma@219.110.80.235)
- # [17:06] * Joins: gavin_ (gavin@74.103.208.221)
- # [17:53] * Joins: briansuda (briansuda@130.208.155.107)
- # [18:04] * Quits: briansuda (briansuda@130.208.155.107) (Quit: briansuda)
- # [18:35] * Joins: ROBOd (robod@86.34.246.154)
- # [18:58] * Quits: Shunsuke (kuruma@219.110.80.235) (Connection reset by peer)
- # [19:01] * Joins: Shunsuke (kuruma@219.110.80.235)
- # [19:06] * Joins: BearShare (BeaR@213.7.62.223)
- # [19:08] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [19:11] * Quits: Shunsuke (kuruma@219.110.80.235) (Quit: Shunsuke)
- # [19:12] * Joins: Shunsuke (kuruma@219.110.80.235)
- # [19:14] * Joins: gavin_ (gavin@74.103.208.221)
- # [19:17] <BearShare> Where can someone study web design ?
- # [19:20] <BearShare> Do u know ?
- # [19:21] <BearShare> brb
- # [19:21] * Quits: BearShare (BeaR@213.7.62.223) (Quit: BearShare)
- # [19:23] * Joins: BearShare (BeaR@87.228.150.156)
- # [19:23] <BearShare> Do u know ?
- # [19:27] <zcorpan> BearShare: i don't understand the question
- # [19:27] <BearShare> I want to study web design . Where can i do that ?
- # [19:29] <zcorpan> dunno. anywhere? probably not the right place to ask
- # [19:30] <BearShare> Do u know ? In which university/college ?
- # [19:30] <BearShare> If there are web design courses
- # [19:30] <zcorpan> you'd have to look that up where you live
- # [19:31] <BearShare> Ooh ok
- # [19:31] <BearShare> sorry for my silly question
- # [19:31] <zcorpan> np
- # [19:33] * Joins: Sander (svl@80.60.87.115)
- # [19:46] * Quits: BearShare (BeaR@87.228.150.156) (Quit: BearShare)
- # [20:00] * Quits: Sander (svl@80.60.87.115) (Connection reset by peer)
- # [20:00] * Joins: Sander (svl@80.60.87.115)
- # [20:07] * Joins: h3h (bfults@66.75.149.197)
- # [20:09] * Quits: zcorpan (zcorpan@84.216.40.107) (Ping timeout)
- # [20:15] * Joins: jgraham (jgraham@81.179.93.10)
- # [20:18] * Joins: zcorpan (zcorpan@84.216.40.107)
- # [20:21] * Quits: h3h (bfults@66.75.149.197) (Quit: h3h)
- # [20:37] * Joins: sp (sp@195.222.86.143)
- # [20:38] <sp> ?
- # [20:38] <sp> der ber spielt hokkey
- # [20:38] <sp> ]yes
- # [20:38] <sp> yes anybody here?
- # [20:40] <Zoffix> no
- # [20:40] <sp> ?
- # [20:41] <sp> do you speake russian?
- # [20:43] <sp> fuck you
- # [20:43] <Zoffix> Yeah, but I can't read your encoding
- # [20:43] <Zoffix> Type in Russian with English letters and don't swear.
- # [20:44] <sp> i`m sorry, i think that nobody can read my words
- # [20:44] <sp> *can`t
- # [20:44] <Zoffix> I see garbage. Not Russian letters.
- # [20:44] <sp> OK
- # [20:44] <sp> is that a gaming servers?
- # [20:45] <Zoffix> sp, no. This is a server for developers who make web standards.
- # [20:45] <Zoffix> Would you like me to repeat that in Russian to be more clear?
- # [20:45] <sp> can you tell me an germany or england server to talk with people?
- # [20:45] <sp> irc
- # [20:45] <Zoffix> sp, Try google.
- # [20:46] <sp> it`s very hard to find
- # [20:46] <Zoffix> sp, I bet "irc german server" would reveal plenty of relevant results.
- # [20:46] <sp> i`ll try
- # [20:46] <sp> thx
- # [20:47] <sp> and where are you from?
- # [20:48] <Zoffix> I am from Russia but I live in Canada.
- # [20:49] <sp> mmm
- # [20:49] <sp> i think you understand me, if i`ll write not veryright
- # [20:50] <Zoffix> I would prefer to discontinue this conversation.
- # [20:50] * Quits: sp (sp@195.222.86.143) (Quit: #u1 - ! ;))
- # [20:51] * Joins: sp (sp@195.222.86.143)
- # [20:52] * Parts: sp (sp@195.222.86.143)
- # [20:53] * Quits: Shunsuke (kuruma@219.110.80.235) (Client exited)
- # [21:11] <zcorpan> someone should paste invisible glue on this channel
- # [21:16] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [21:21] * Quits: myakura (myakura@60.239.122.32) (Quit: Leaving...)
- # [21:21] * Joins: gavin_ (gavin@74.103.208.221)
- # [21:26] <zcorpan> ...and w3c is too similar to wc3
- # [21:29] * Joins: foca (foca@190.64.6.4)
- # [21:32] <Zoffix> heh, you should visit #css on EFNet :) every 5th user is asking for "war 2vs2" >:)
- # [21:35] <xover> Klanwar! Klanwar!
- # [21:41] * Joins: sbuluf (lxcpi@200.49.140.83)
- # [21:55] * Quits: jmb (jmb@81.179.74.126) (Quit: Lost terminal)
- # [21:55] * Joins: jmb (jmb@81.179.74.126)
- # [22:16] <Deeder> Zoffix : nice conversation with sp
- # [22:16] * Joins: Roger (roger@213.64.74.230)
- # [22:17] <Deeder> I confess : this made me laugh
- # [22:17] <Deeder> ;)
- # [22:19] <Philip`> It just made me wonder how to get UTF-8 set up properly in my IRC client, which seems unpleasantly non-trivial :-(
- # [22:21] <Deeder> Philip` : /charset command don't work properly ?
- # [22:24] <Philip`> Deeder: Doesn't seem to - I just get question marks when trying to paste Greek letters into here
- # [22:24] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
- # [22:25] <Deeder> ha okay, that's strange
- # [22:25] <Deeder> what's your IRC client ?
- # [22:25] <Philip`> It's irssi in screen in ssh in Konsole - I've tried to set the appropriate flags in each of those, but probably missed something in the middle...
- # [22:27] <Deeder> I can't help you, i never tried irssi.
- # [22:28] <Deeder> The only one IRC console client i tried was xchat-text once my Xorg didn't want to launch my graphical desktop
- # [22:39] * Quits: Roger (roger@213.64.74.230) (Quit: Roger)
- # [22:40] * Joins: myakura (myakura@60.239.122.32)
- # [22:41] * Quits: loic (loic@90.29.30.52) (Quit: hoopa rules)
- # [22:41] * xover suspects Matthew Raymond has stopped actually reading messages and is now only ranting on auto-pilot...
- # [22:43] * gsnedders can't be bothered to contribute further to that thread
- # [22:43] <gsnedders> too many people blindingly sure they're right
- # [22:44] <gsnedders> nobody has truly convinced me one way or the other
- # [22:44] <Lachy> I just cannot believe Matthew Raymond is complaining about 14 bytes being a serious problem!
- # [22:45] <gsnedders> 14 bytes! that's OMFG no!
- # [22:46] <gsnedders> I mean, even I'm older than 14! (Well, since yesterday)
- # [22:46] <gsnedders> or two days ago if you live in the future like Lachy :)
- # [22:46] <Philip`> (Happy birthday yesterday :-) )
- # [22:48] <Philip`> I think arguing that 14 bytes is significant for Google or Yahoo isn't very relevant in a discussion about the web, since they're two sites out of billions and they shouldn't have that much influence
- # [22:48] <Lachy> gsnedders, so your birthday was on April 20?
- # [22:48] <gsnedders> Google doesn't use valid markup in anyway whatsoever anyway
- # [22:48] <gsnedders> Lachy: yes
- # [22:48] <Lachy> right, that is 2 days ago :-)
- # [22:49] <gsnedders> Lachy: depends on the timezone :)
- # [22:51] <Lachy> perhaps I should reply and say he wasted far more bytes sending that pointless argument, than a bugmode switch ever would
- # [22:52] <gsnedders> I just wrote an email longer than most of the essays I've done as coursework. Maybe I should get my priorities right :P
- # [22:55] <Philip`> Lachy: To be fair, he only wasted as many bytes as a quarter of a million visits to web sites with that extra code, while those fourteen bytes will continue to add up over years
- # [22:57] <Philip`> (although, to be less fair, that would only be relevant if his post actually changed the future of HTML and saved the world from those fourteen bytes, which seems unlikely)
- # [22:57] <Lachy> how did you calculate that? DId you count the total number of bytes in the one copy you received, or did you factor in the 350+ subscribers who recieved it also?
- # [22:58] * xover throws in the towel…
- # [22:58] * Quits: asbjornu (asbjorn@84.48.116.134) (Quit: asbjornu)
- # [22:59] <Philip`> That was the size of the message (10KB) multiplied by the 356 subscribers to the mailing list
- # [22:59] <xover> We need the PageRank for the list archive page for that message too, to estimate bandwidth wasted when accessing the archived copy.
- # [22:59] <Philip`> (so I'm ignoring everybody reading the list archive)
- # [22:59] <xover> heh heh
- # [22:59] * Joins: foca_ (foca@190.64.18.137)
- # [23:01] <xover> On a side note, Google seem to be trading off page size for some other requirements.
- # [23:01] * Quits: foca (foca@190.64.6.4) (Ping timeout)
- # [23:01] <Philip`> Compatibility with ancient browsers that nobody else in the world cares about?
- # [23:02] <xover> I've seen people experiment and manage to shave off quite a few bytes compared to what they've currently deployed.
- # [23:03] <Philip`> (I can't think why else they'd e.g. leave newlines in the CSS)
- # [23:04] * xover shudders to think of the amount of bandwidth Google as a whole must be consuming...
- # [23:09] * Joins: h3h (bfults@66.75.149.197)
- # [23:09] <Sander> xover: do you have a website somewhere with your public key? I'm getting annoyed at enigmail wanting to find it on a keyserver for each of your messages, and continuously failing
- # [23:09] * Quits: h3h (bfults@66.75.149.197) (Quit: h3h)
- # [23:13] * Joins: laplink (link@193.157.66.154)
- # [23:15] * Quits: sbuluf (lxcpi@200.49.140.83) (Ping timeout)
- # [23:16] * Joins: sbuluf (vaztp@200.49.140.83)
- # [23:18] * Parts: foca_ (foca@190.64.18.137)
- # [23:24] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
- # [23:29] * Joins: gavin_ (gavin@74.103.208.221)
- # [23:52] * Joins: zdenko (zdenko@84.255.203.169)
- # Session Close: Sun Apr 22 00:00:00 2007
The end :)