/irc-logs / w3c / #html-wg / 2007-04-19 / end

Options:

# Session Start: Thu Apr 19 00:00:00 2007
# Session Ident: #html-wg
# [00:16] * Quits: olli- (olli@80.203.95.229) (Ping timeout)
# [00:29] * Joins: Zeros (Zeros-Elip@67.154.87.254)
# [00:55] * Quits: Sander (svl@80.60.87.115) (Quit: And back he spurred like a madman, shrieking a curse to the sky.)
# [00:58] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [00:58] * Joins: karl (karlcow@128.30.52.30)
# [01:03] * Joins: gavin_ (gavin@74.103.208.221)
# [01:06] * Parts: hasather (hasather@81.235.209.174)
# [01:08] * Joins: sbuluf (rits@200.49.140.243)
# [01:12] * Quits: billmason (billmason@69.30.57.156) (Quit: .)
# [01:25] * Quits: kingryan (rking3@66.92.187.33) (Quit: kingryan)
# [01:38] * Quits: Zeros (Zeros-Elip@67.154.87.254) (Ping timeout)
# [01:56] * Joins: kazuhito (kazuhito@210.232.34.13)
# [02:06] * Joins: anne5 (annevk@131.181.85.131)
# [02:13] * Joins: Lachy (chatzilla@131.181.47.44)
# [02:14] * Joins: heycam (cam@131.181.85.133)
# [02:15] * Joins: Zeros (Zeros-Elip@69.140.48.129)
# [02:16] <Zeros> okat
# [02:16] <Zeros> +++111
# [02:16] <Zeros> this is just getting silly
# [02:17] <Zeros> That's not even valid mathematically or programatically.
# [02:17] <Zeros> :/
# [02:17] <Dashiva> It is
# [02:18] <sbuluf> why exactly isn't there a sort of maling list markup widespread?
# [02:18] * zcorpan_ wonders why public-html is full of +1s and -1s, while the whatwg list basically doesn't have them at all
# [02:18] <Lachy> because whatwg is much more mature
# [02:18] <Dashiva> Because the w3c is based on consensus rather than merit? :)
# [02:18] <Lachy> public-html is filled with a lot of newbies
# [02:18] <Philip`> Zeros: It's syntactically valid Perl
# [02:19] <Zeros> Philip`, what do three +'s mean?
# [02:19] <Lachy> Philip`: what does it mean in perl?
# [02:19] <Zeros> Now, If there were four...
# [02:19] <Philip`> I'm unsure whether it's parsed as ++(+111) or +(++111) - in both cases it fails because you can't pre-increment a constant
# [02:20] <Philip`> (The unary + operator does precisely nothing)
# [02:20] <Dashiva> It casts to number in JS :)
# [02:20] <Zeros> looks like ruby is okay with it
# [02:20] <Zeros> But its parsed as +(+(+(111)))
# [02:22] <Zeros> anyway, that's not the point. People need to stop with the ±whatever (dis)agreement emails.
# [02:22] <zcorpan_> +1
# [02:22] * zcorpan_ hides
# [02:25] * Quits: Voluminous (Voluminous@66.195.32.2) (Quit: Leaving)
# [02:28] * Joins: Shunsuke (kuruma@133.27.61.17)
# [02:34] <Hixie> hm, ten replies so far, about half of the people who replied said they'll attend http://www.w3.org/2002/09/wbs/40318/tel26Apr/results
# [02:35] * Joins: marcos (chatzilla@131.181.99.92)
# [02:36] <zcorpan_> i don't understand why the possibility of the editor stops being able to be edit the spec means we must have more than one editor now. if the situation arises then we can find a new editor then, no?
# [02:41] <Dashiva> It might be informative to split attendees into IRC and phone participants
# [02:42] <h3h> indeed.
# [02:43] <Hixie> Dashiva: i believe this form is intended to only be answered "yes" by phone participants, but you raise an interesting point
# [02:43] * Quits: Shunsuke (kuruma@133.27.61.17) (Ping timeout)
# [02:43] <Hixie> you should point this out to our chairs :-)
# [02:43] <zcorpan_> i answered "no" but commented that i might attend on irc
# [02:45] <h3h> same
# [02:45] <Dashiva> "we'll make an effort to accomodate IRC-only participation." and then the question asks "We'd like to know how many participants to plan for."
# [02:46] <h3h> yeah, that is confusing
# [02:46] * Quits: h3h (bfults@66.162.32.234) (Quit: |)
# [02:49] <Philip`> zcorpan_: A possible problem is that if there is a single editor, they may keep significant amounts of private information (plans, tasks, works-in-progress, etc), which would be lost if that person stopped
# [02:49] <Hixie> yeah, editors should imho keep all the information in the spec as i do
# [02:50] <Philip`> (whereas if there are always multiple editors, they would be expected to share all the editorial information (in a format that they can all understand), so the group would not lose access to it)
# [02:54] <Philip`> Perhaps there's also a problem that if there's a single editor, who at some point turns rogue and has to be forcibly removed by the <whoever it is that has control>, that removal would be very disruptive (since there's nobody to take over the role immediately) and those whoevers would be in an unpleasant situation, so they would prefer to avoid that situation entirely
# [02:55] <mjs> having one rogue editor with a co-editor doesn't seem obviously better
# [02:55] <Philip`> (whereas if there are multiple ones, you can throw one out and it won't cause as much visible disruption)
# [02:55] <mjs> anyway, I can see there might be some benefits, but there are also costs
# [02:55] <mjs> so it depends on who the specific candidate is
# [02:56] <Zeros> Is the CSS3 WG open to anyone?
# [02:57] <Lachy> Zeros: no
# [02:57] <Hixie> anyone with $$
# [02:57] <mjs> you need $$ or an invitation
# [02:58] <anne5> I'd like to attend I think
# [02:58] <Zeros> thanks, I didn't think it was
# [02:58] <anne5> For fun, mostly
# [02:58] <mjs> I'm kind of curious what will happen
# [02:59] <anne5> but i've no idea if i'm able to attend
# [02:59] <Hixie> i'd like to attend to see what happens, but i don't really have 90 minutes to spare
# [03:01] * Joins: olivier (ot@128.30.52.30)
# [03:02] * Joins: Zoffix (Zoffix@74.111.228.242)
# [03:05] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [03:10] * Joins: gavin_ (gavin@74.103.208.221)
# [03:13] * Quits: zcorpan_ (zcorpan@84.216.41.52) (Ping timeout)
# [03:33] <Lachy> I can't believe there was something that I agree with Chris Wilson about!
# [03:33] <Lachy> that some bugmode/version attribute must be a proprietary and non-conforming extension.
# [03:33] <Lachy> it seems the only thing we don't yet agree on is the default being always-standards mode or always-IE7
# [03:35] <Zeros> Seems always standards mode would be the best default. Then Intranet and "must work forever" apps can specify the attribute to force old rendering behavior in newer versions
# [03:38] <Philip`> Isn't that assuming the people who are developing intranet and "must work forever" apps are competent and understand what they're doing, when they copy-and-paste the document header from whatever source they happen to be looking at?
# [03:38] <Lachy> Zeros: that's basically what I proposed in my compromise email. Though, I suspect Chris may have misunderstood it and I'll try and get him to agree to always-standards mode in some otherway
# [03:39] <Philip`> If it's possible to accidentally select always-standards mode, people will do it accidentally, and everything will work exactly the same for the next few years, and then IE will be stuck because if they change their "always-standards mode" they'll broke all those people, and those people will be unhappy
# [03:40] <Zeros> That's also true
# [03:40] <mjs> I think the problem w/ always-standards being the default mode is that the people who want bug lock-in are often less skilled and knowledgeable than those who want constantly updating standards support
# [03:41] <Lachy> Philip`: you're making the assumption that most pages will rely on IE[n] bugs that will be removed when IE[n+1] is released
# [03:41] <Lachy> that assumption is simply false
# [03:41] <Zeros> They also may introduce new bugs that are completely unrelated, which is a problem
# [03:42] <Zeros> Always standards mode, might be "broken code in X edge case" and with IE's release schedule that could be a real problem to throw new rendering behavior at people unexpectedly
# [03:44] <Philip`> If there is this versioning, will IE-for-HTML5 try hard to fix all the fundamental problems (like non-tree DOM and everything) in the first release, and actually conform to HTML5 and be of similar quality to all the other browsers? If they do all that breakage in this one transition, hopefully that would greatly reduce the number of IE[n] bugs that people will rely on in the future
# [03:45] <Philip`> (If IE-for-HTML5 is just IE7 with new features hacked on and some bugs worked around, that won't actually fix the problems and they will need new version flags in the future, I guess)
# [03:45] <Zeros> Philip`, possibly. They are interviewing people in the standards community for advisory roles on the IE team. HTML5 is some years off too which gives them time to roll out IE8 with fixed support for common HTML features to allow HTML5 support in the future
# [03:46] <Lachy> there will always be bugs in browsers, but the correct way to move forward is to actively fix bugs where possible and, if it can be shown with real evidence that some bugs are relied on, then those bugs should be standardised
# [03:47] <Lachy> Chris Wilson's approach of "we don't need research, we can just assume all bugs are relied upon" is backwards and unworkable
# [03:48] <Zeros> Lachy, true, but I do see where MS has a particularly fragile position. Gecko can roll out a new version and expect huge adoption because a lot of users who use it know what a browser actually is, and because their web market share is a whole lot smaller
# [03:48] <Zeros> IE *is* the internet to a lot of people, and with a very large portion of the market big changes have nasty results
# [03:49] <Zeros> I can see where he's coming from
# [03:49] <Philip`> Zeros: If they release IE8 with HTML5-versioning, I suppose they'd only have a couple of years (i.e. until IE9) to actually get it right, else the number of people accidentally using the "IE8 bugs" profile of HTML5 will be too much to break. So we need to give them lots of test cases and encouragement to get it right fast enough :-)
# [03:50] <Lachy> although IE may be specifically targeted by a lot more intranet apps that other browsers, but as far as the internet is concerned, Mozilla, IE and all other browsers should be compatible with the same set of pages
# [03:51] <Zeros> Lachy, its also used in a great number of applications that aren't related to the web at all
# [03:52] <Zeros> I imagine that Webkit will suffer a similar fate in the future. Lots of desktop apps are using it now, and drastic changes mean broken apps
# [03:53] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
# [03:57] <Lachy> AFAICT, the worst case scenario with some sort of bugmode attribute is that we'll be forced to use it, but it can be worked around using bugmode="IE9999" (or some other large, random number)
# [03:59] <Lachy> that is assuming IE will make opt-ins for IE[n+1] also opt-in to the latest mode in IE[n]. That's really the only way it could work, AFAICS
# [04:00] <Philip`> They could avoid that by naming the versions "IE Bon Echo", "IE Gran Paradiso", etc, so nobody can predict the future values
# [04:01] <Lachy> Then IE[n] would have to accept any random string as the latest mode, since you don't want IE[n] to fallback to IE7-mode when you opt-in to IE[n+1]
# [04:01] * Quits: karl (karlcow@128.30.52.30) (Quit: This computer has gone to sleep)
# [04:02] <Lachy> so there really is no way for IE to force web developers to only opt-in to the current version.
# [04:04] <mjs> Zeros: if we had to add compat hacks solely for desktop apps, we'd probably limit scope to just that app and try to make it temporary
# [04:04] <Lachy> The only other way would be to require that authors explicitly opt-in to all IE versions since IE8, as in bugmode="IE8 IE9 IE10 IE11 ... IE100"
# [04:04] <mjs> (no promises on what we'd actually do though)
# [04:04] <Philip`> IE[n] could store a public-key-encrypted hash of every version name (which they prepare in advance for the next few decades, but keep totally secret), so it can verify IE[n+1]'s name without anyone being able to discover that name until Microsoft tells them. Not sure if that'd be useful in practice, though...
# [04:05] <Lachy> Philip`: that would be an evil way for IE to ensure that they get their own way
# [04:05] <Lachy> and Chris Wilson seems to want to avoid the perception that he is evil
# [04:06] <mjs> is he succeeding?
# [04:06] <mjs> (at avoiding that perception)
# [04:06] <Lachy> mjs, not really
# [04:08] * Joins: Shunsuke (kuruma@133.27.61.17)
# [04:08] <Lachy> although I'd like to give him the benefit of the doubt, he seems intent on doing evil things in spite of all the objections.
# [04:09] <sbuluf> i pinted to three questions here a couple of days back
# [04:09] <sbuluf> 1)is there a compromise possible?
# [04:09] <sbuluf> 2)if so...is it worth it?
# [04:09] <Zeros> Lachy, optin like that would never work
# [04:10] <Zeros> IE versions step on each other, opting into IE7 and IE8 would surely make your page broken in both
# [04:10] <Lachy> Zeros: do you mean listing each version explicitly?
# [04:10] <Zeros> yes
# [04:10] <sbuluf> 3) if the worth is really marginal...is it worth anough to let microsoft appear somehoow validated, with a "seal of approval�"?
# [04:10] <Lachy> Zeros: I know, I was just coming up with a solution that didn't have a particularly easy work around
# [04:10] <sbuluf> what PR sentences will MS be able to pull afterwards?
# [04:10] <Lachy> I'm in no way suggesting that they should use that
# [04:15] <Philip`> Rather than considering Chris to be evil, it's probably friendlier (and hence (maybe) conducive of productive discussions rather than arguments) to blame anonymous bosses at Microsoft, for creating the IE6 situation and for giving a "you must not do another IE7 since that'll break important paying customers" decision that can't be argued against
# [04:15] <mjs> I don't think Chris is personally evil
# [04:15] <mjs> though I think he is misguided on some points
# [04:16] <mjs> I do think Microsoft has incentives that don't align with those of others
# [04:17] <Zeros> The same is, of course, true of Apple and the Mozilla foundation, and Opera
# [04:18] <Zeros> Not to say MS hasn't done some profoundly "evil" things in the past
# [04:20] <Philip`> I would assume Microsoft's [the corporation's, not the individuals'] desires are to make money for Microsoft, while Apple's are to make money for Apple and Opera's are to make money for Opera and Mozilla's are to, uh, 'promote choice and innovation on the internet'
# [04:21] <Philip`> so it's just fortunate that improving HTML will (hopefully) help all of those people achieve those desires
# [04:21] <Zeros> hopefully
# [04:22] <Zeros> the Mozilla Corp. is for profit so they're not quite as completely unbiased anymore
# [04:24] * Quits: Zeros (Zeros-Elip@69.140.48.129) (Quit: Leaving)
# [04:25] * Quits: Shunsuke (kuruma@133.27.61.17) (Ping timeout)
# [04:27] <mjs> I think Mozilla, Apple and Opera have somewhat more aligned incentives
# [04:28] <anne5> I think you can remove "somewhat"
# [04:30] <mjs> since for all of us, interoperability with IE is a bigger issue than compatibility with old versions of IE
# [04:30] <mjs> er, compatibility with old versions of our own browsers
# [04:30] <mjs> plus, since none of us has 80% market share, none of us benefit much from content being browser-specific
# [04:39] <sbuluf> " Mozilla, Apple and Opera, with W3C blessing, just made a deal with Microsoft that perpetuates, in the Web's lingua franca, not just all errors past, but all vendor-specific past errors, AND all vendor-spacific FUTURE errors."
# [04:39] <sbuluf> sentences like that might pop up
# [04:41] <anne5> seems inaccurate
# [04:42] <anne5> but I'm not sure what it means
# [04:42] <sbuluf> might be, yes, i was wondering how accurate might be, in fact.
# [04:43] * Joins: Shunsuke (kuruma@133.27.61.17)
# [04:43] <sbuluf> it means all design errors in html get frozen, plus all IE past bugs, plus IE future bugs
# [04:43] <mjs> so far, I haven't seen any Apple, Opera or Mozilla reps agree with Microsoft's versioning ideas
# [04:43] <sbuluf> it was an hipotetical example, mjs
# [04:44] <anne5> i think you should be more specific in what you're saying
# [04:45] <sbuluf> perhaps, anne, yes.
# [04:57] * Quits: kazuhito (kazuhito@210.232.34.13) (Quit: Quitting!)
# [05:01] * Quits: Shunsuke (kuruma@133.27.61.17) (Ping timeout)
# [05:02] * Joins: Shunsuke (kuruma@133.27.61.17)
# [05:13] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [05:15] * Quits: dbaron (dbaron@63.245.220.242) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [05:16] * Quits: jmb (jmb@81.179.74.126) (Ping timeout)
# [05:18] * Joins: gavin_ (gavin@74.103.208.221)
# [05:21] * Quits: Shunsuke (kuruma@133.27.61.17) (Ping timeout)
# [05:27] * Joins: Shunsuke (kuruma@133.27.61.17)
# [05:28] * Quits: Shunsuke (kuruma@133.27.61.17) (Connection reset by peer)
# [05:29] * Joins: Shunsuke (kuruma@133.27.61.17)
# [05:32] * Quits: Shunsuke (kuruma@133.27.61.17) (Connection reset by peer)
# [05:45] * Joins: karl (karlcow@128.30.52.30)
# [06:24] * Quits: myakura (myakura@60.239.122.32) (Ping timeout)
# [06:46] * Joins: marcos_ (chatzilla@131.181.99.92)
# [06:48] * Quits: marcos (chatzilla@131.181.99.92) (Ping timeout)
# [06:48] * marcos_ is now known as marcos
# [06:48] * Quits: Lachy (chatzilla@131.181.47.44) (Ping timeout)
# [06:54] * Joins: Lachy (chatzilla@131.181.47.44)
# [07:14] * Joins: myakura (myakura@60.239.122.32)
# [07:18] <mjs> is it just me that finds Dan's messages about editors a little weird?
# [07:19] <anne5> i sort of stopped caring for the moment
# [07:19] <anne5> two editors seems incompatible with the WHATWG
# [07:20] <anne5> so I suppose it's either Hixie and WHATWG and HTML WG will stay in sync or it's someone else (unless Hixie is ok with editing together with someone else on the WHATWG doc) and the WHATWG doc will be a superset
# [07:21] <mjs> or it will be one of the few people in the universe that Hixie could actually work with
# [07:21] <mjs> (how likely does that seem?)
# [07:22] <anne5> It seems hard to work together with someone else on such a job
# [07:22] <anne5> but maybe he figures something out
# [07:23] * Joins: Voluminous (voluminous@24.16.148.66)
# [07:25] <anne5> http://simon.html5.org/temp/valid-html5.png is awesome
# [07:32] <anne5> http://krijnhoetmer.nl/irc-logs/xhtml/20070418#l-107
# [07:58] * Joins: Shunsuke (kuruma@133.27.175.115)
# [08:03] <mjs> <Steven> Mark: We should just say, if you are not well-formed, it is undefined what a UA does
# [08:03] <mjs> <Steven> Mark: We shouldn't say that it aborts, just that it is undefined
# [08:03] <mjs> wow, they want to turn XHTML into tag soup!
# [08:05] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [08:10] * Joins: gavin_ (gavin@74.103.208.221)
# [08:10] * Quits: Voluminous (voluminous@24.16.148.66) (Quit: Leaving)
# [08:12] <anne5> me too
# [08:12] <anne5> I think that's a goal we have in common
# [08:17] <Hixie> they want to turn it into _undefined_ tag soup
# [08:18] <anne5> fair enough
# [08:22] * Joins: jmb (jmb@81.179.74.126)
# [08:26] * Joins: loic (loic@90.29.116.168)
# [08:27] <mjs> seriously, it sounded like they wanted to remove the requirement to abort on content that is not well-formed even
# [08:27] <mjs> doesn't that defeat the whole purpose of using XML?
# [08:27] <hsivonen> mjs: it depends on whether the purpose is techical or whether it is a marketing purpose
# [08:33] <anne5> I want to remove namespace well-formed from XML
# [08:33] <anne5> but also define how to handle the errors
# [08:33] * Joins: Zeros (Zeros-Elip@69.140.48.129)
# [08:33] <karl> I wonder if Lachy has only geek friends. :) "but most authors still know HTML to some degree "
# [08:34] <karl> :))) most people I know do not even know what is HTML
# [08:34] <anne5> guess it depends on your def of authors
# [08:34] <karl> but they still produce content for the Web
# [08:34] <Lachy> karl, HTML is the most used language on the web
# [08:34] <karl> anne5: exactly what I was saying
# [08:34] <Lachy> Do you think authors are writing it without a clue what on earth it is?
# [08:34] <karl> yes Lachy
# [08:34] <anne5> if they're using a tool they'd count as users imo
# [08:34] <hsivonen> Is this multiple editor thing about increasing out truck number or a replay of what happened with the editorship of the XML spec?
# [08:35] <anne5> and the tool would be the author/editor
# [08:35] <hsivonen> s/out/our/
# [08:35] <Lachy> I explicitly excluded those who use wysiwyg editors without ever looking at the source
# [08:35] <mjs> what happened with the editorship of the XML spec?
# [08:35] <Lachy> but there are many who have a basic understanding of HTML, even if they use stuff like <font> and tables for layout
# [08:35] <karl> most people using weblog publishing platform don't have any knowledge of HTML
# [08:36] <hsivonen> mjs: an editor was added to balance the editor affiliations
# [08:37] <mjs> hsivonen: I suspect that may be the issue for some
# [08:37] <mjs> personally, having seen Ian's work under three different affiliations, I am not especially concerned about his current one
# [08:37] <Zeros> karl, I don't think that's quite true. If they're adding links and basic styles to the blog entries they tend to know /of/ HTML
# [08:38] <Zeros> Just like someone can know how to export to HTML, and thus know what HTML is, but not how to author it by hand.
# [08:38] <karl> Zeros :) make a poll around you with people who are not geeks.
# [08:38] <Zeros> karl, how do people add links to their wordpress entries then?
# [08:38] <anne5> WYSIWYG editors
# [08:38] <anne5> WordPress supports that
# [08:39] <karl> zeros: not many do I think, and They have wysiwyg UI in browser forms
# [08:39] <mjs> I think HTML is one of the more widely recognized technical terms out there
# [08:39] <Lachy> karl: regardless of the actual percentage of authors that do and do not know HTML, there are millions of them that do and they shouldn't be ignored
# [08:39] <Lachy> yet, that is exactly what Chris Wilson is doing
# [08:39] <Zeros> karl, I don't know. It seems like we're underestimating how common the term HTML is
# [08:39] <karl> millions: yoohooo, stats source?
# [08:39] <mjs> Also, authors that don't know what HTML is are almost by definition authoring via editing tools that generate the markup for them
# [08:40] <hsivonen> Lachy: btw, what authors want and what they say they want may be two different things
# [08:40] <mjs> in which case, the spec needs to give consideration to such editing tools, since users of them will not be exposed to or directly affected by details of the markup
# [08:40] <Lachy> there are billions of web sites, are you going to claim that they were written by only a few thousand authors that know HTML and the rest using WYSIWYG editors without a clue what they're doing?
# [08:40] <mjs> I think a lot of people sincerely believe they want standards, but still get mad if you break them
# [08:40] <Lachy> hsivonen: what I want and what I say are want are the same thing
# [08:40] <karl> I'm always surprised by how much we, geeks, think that the world is revolving around our own knowledge.
# [08:40] <hsivonen> Lachy: to find out what they really want, you have to try alternative spec stimuli and see how they react
# [08:41] <hsivonen> Lachy: or rather, not spec stimuli but browser environment stimuli
# [08:41] <Lachy> hsivonen: I have no idea what you mean
# [08:42] <Lachy> karl: I don't think the world revolves around geeks, just that geeks aren't some minority that can be ignored
# [08:42] <Zeros> I hardly think you need to be a geek to know what HTML is
# [08:42] <hsivonen> Lachy: I mean that authors may say they want and always standards mode but still scream murder when a new browser version breaks their stuff
# [08:42] <karl> Lachy: I didn't say they should be ignored. but it is a minority of people or concerns. (important in terms of actions, sure)
# [08:43] <hsivonen> Lachy: if that happens, what they want and what they say they want is different
# [08:43] <karl> but I always try to think out of my own box.
# [08:43] <Zeros> deviantART for instance doesn't support bbcode or any type of special markup. You *have* to use HTML to add special content.
# [08:43] <Zeros> And the majority of their users are quite technically challenged
# [08:43] <karl> hsivonen: agreed with henri on this
# [08:44] <hsivonen> Lachy: so to find out what authors really want, you have to draw on the data about the behavior of authors in response to actual stimuli
# [08:44] <Lachy> hsivonen: ok
# [08:44] <mjs> FWIW, Safari releases have occasionally broken things by fixing standards compliance bugs, and authors don't scream bloody murder too much, because we also fix a lot of things, and we release often enough and are already compliant enough that breakage is gradual, not wholesale
# [08:44] <karl> plus the fact that there is NOT only ONE UNIQUE community of Web developers
# [08:44] <Lachy> karl: I know that
# [08:45] <Zeros> mjs, you also release nightlies so its easier to catch changes as they happen and test incrementally. By the time the IE beta comes out there's not that much in the way of changes that get done. Lots of CSS bugs were reported and never fixed for IE7.
# [08:45] <karl> which reminds a web dev in 2001 in the web design agency who was thinking that it "world wild web". He was speaking seriously. I had to show him articles.
# [08:46] <karl> s/reminds/reminds me/
# [08:46] <beowulf> :)
# [08:46] <Zeros> karl, as in, he thought www = 'world wild web'?
# [08:46] <Lachy> the major problem with MS is their proposal takes away the author's right to choose what mode they want (beyond the limited choices offered), and it assumes that the majority of web sites rely on bugs that will break with the next release
# [08:46] <beowulf> probably more accurate :)
# [08:46] <karl> Zeros: yes
# [08:46] <Zeros> scary
# [08:47] <Zeros> Lachy, for IE? I think that's easy to prove to be true
# [08:47] <anne5> seems accurate
# [08:47] <Lachy> That kind of backwards thinking is extremely detrimental to the web, and the problem is that MS is in such a position to inflict it on the world
# [08:47] <Zeros> Lachy, Most websites that are attempting standards compliance and use CSS end up using IE hacks or conditional comments, else they're probably rendering in quirks mode or didn't test outside IE very much from what I've seen
# [08:48] <Zeros> That's where new versions of IE cause huge breakage
# [08:48] <Lachy> Zeros: hacks are different issue
# [08:48] <Lachy> good and well document CSS hacks are designed so that they will not break with more standrads compliant behaviours.
# [08:49] <Lachy> Unfortunately, not all hacks are benign and some rely on the presence of 2 (or more) bugs that may not be fixed in the same release
# [08:50] <Zeros> I'm not sure I'd call any hack on a non-frozen code base benign
# [08:50] <Lachy> and unfortunately, that occured with * html in IE7, but that was just a bad design choice by MS that they shouldn't have made
# [08:50] <Lachy> I call a hack like this benign: div { float: left; margin: 1em; display: inline; }
# [08:51] <Zeros> They should have fixed the * html bug, but they also should have fixed the layout bugs that are related to it
# [08:51] <Lachy> The display: inline; fixes the double margin float bug, but doesn't cause any problem in standards compliant browsers at all
# [08:51] <Zeros> That's the problem. They fixed parser bugs, but not the reason why people used the parser bugs.
# [08:51] <Lachy> Zeros: yes!
# [08:52] <Zeros> Lachy, btw, *+html {} works in IE7
# [08:52] <Zeros> kind of funny how they fixed * html, and added +... and then look where we ended up
# [08:52] <anne5> I think that works in Opera too
# [08:52] <Lachy> that was their mistake, and instead of learning from it and asking how we can make changes without doing that again, they're saying we're not going to make any changes to the handling of existing pages at all, under any cirumstances
# [08:52] <anne5> well, worked
# [08:53] <anne5> (if you include some <?xml?> at the top)
# [08:53] <Lachy> they want to perpetuate all bugs, for all time, and that solution is simply unworkable, even for MS
# [08:55] <Lachy> gotta go, cya
# [08:55] <Zeros> entropy will catch up with them eventually, sure
# [08:55] * Quits: Lachy (chatzilla@131.181.47.44) (Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.3/2007030919])
# [08:55] <Zeros> night
# [08:56] <anne5> he's not going to bed, fwiw
# [08:56] <anne5> it's 17:00 here
# [08:56] <Zeros> oh okay, I wasn't sure of his timezone
# [08:57] <anne5> +10GMT atm
# [08:57] * marcos is now known as Lachy
# [08:57] <karl> http://esw.w3.org/topic/PeopleLocation
# [08:57] * anne5 is +10GMT too atm
# [08:58] <Zeros> karl, thanks
# [08:58] <anne5> that page is wrong for Marcos btw
# [08:58] <anne5> Brisbane is +10GMT throughout the year
# [08:59] <Lachy> I'm back, from another computer
# [08:59] * Joins: olli- (olli@80.203.95.229)
# [08:59] <anne5> heh, olli-!
# [08:59] <anne5> olli-, planning to join?
# [09:00] <karl> Japan doesn't have change of time which is quite cool.
# [09:00] <anne5> Not having a change is quite annoying imo, a lot less sun in the evening
# [09:03] <hsivonen> anne5: oh, you still don't believe that DST is evil?
# [09:03] <olli-> anne5: hei :-)
# [09:03] <olli-> anne5: just hanging around
# [09:04] <anne5> hsivonen, quite the opposite
# [09:04] <karl> anne5: it depends on your position on the solar time
# [09:04] <karl> solar time != clock time often
# [09:04] <anne5> sure, it's annoying here in Brisbane
# [09:05] * karl being an early bird it is quite cool too ;)
# [09:05] <karl> woke up this morning around 5:30am
# [09:05] <hsivonen> off-topic, but my position on solar time is that the world should move to atomic time and decouple the calendar year from the solar year (i.e. kill the leap seconds)
# [09:05] <karl> and went to the office at 6am
# [09:05] <karl> with day light
# [09:06] <Lachy> hsivonen: that would be awesome
# [09:07] <karl> hsivonen: as long as the poet can smell the roses in the garden... no issue.
# [09:07] <mjs> hsivonen: we should also put up giant reflector dishes to blot out the sun, so nonconformists can't cheat and still be on solar time
# [09:07] <Lachy> lol
# [09:08] <hsivonen> karl: it would take a rediculously long time for the solar year to drift a day's worth of leap seconds from a leap secondless atomic time Gregorian year, so no problem for poets in the next couple of thousand years
# [09:08] <karl> mjs: there was a USSR project in the 90s for giant orbital reflectors to give light to the mines in Siberia in winter.
# [09:09] <hsivonen> FWIW, POSIX, GPS and Web Forms 2.0 already eschew leap seconds
# [09:11] <Hixie> leap seconds are important... but not for system clocks, which are probably accurate to +/- minutes rather than seconds most of the time
# [09:13] <hsivonen> Hixie: you are joking, right? leap seconds are a problematic fix to a non-problem. for IAT and UTC to diverge by 24 hours would take about forty-thousand year, if I'm not mistaken
# [09:13] * Joins: kazuhito (kazuhito@210.232.34.13)
# [09:13] <Hixie> they don't have to diverge by 24 hours for it to be a problem
# [09:14] <Hixie> we have daylight savings time to correct for just one hour of difference to sunlight hours
# [09:15] * heycam notes to karl that http://www.w3.org/MarkUp/Activity is sent with the wrong encoding
# [09:15] <Hixie> hsivonen: doing multi-minute leaps is far more complicated and problematic than the occasional second leap can ever be
# [09:16] <Hixie> hsivonen: it doesn't really matter to most people whether something happened at 23:59:59 or 23:59:60, but it _does_ matter to a lot of people whether something happens at 23:59:00 or 23:60:00
# [09:16] <Hixie> (assuming just a one-minute leap)
# [09:17] <hsivonen> Hixie: but you can't magically make algoritms change when someone makes a decision in France
# [09:17] <Hixie> having leap seconds lets us do correction without having to worry about the general populace knowing
# [09:17] <Hixie> hsivonen: you don't need to. You can ignore leap seconds for almost all purposes
# [09:17] <Hixie> hsivonen: the correction of UTC has to be arbitrary because it's correcting for the difference in earth's orbit, which isn't current predictable
# [09:18] <Hixie> it's affected by all kinds of stuff like rocket lauches, etc
# [09:18] <mjs> karl: a step in the wrong direction!
# [09:18] <Hixie> the "algorithm" for handling leap seconds is just NTP
# [09:18] <Hixie> the algorithm for handling leap minutes (which would be just as unpredictable) is far, far more difficult.
# [09:19] <Hixie> e.g. it would affect things like VCRs trying to record TV shows
# [09:19] <Lachy> I don't understand exactly what the problem is that leap seconds solve
# [09:19] <Lachy> if everyone just used UT, then there would be no need for corrections
# [09:20] <Hixie> Lachy: UTC is ever so slightly (and unpredictably) out of alignment with the earth's rotation
# [09:20] <hsivonen> Lachy: they solve the "problem" of not having the solar year drift compared to the cesium year
# [09:20] <karl> heycam: Thanks for the notice! and fixed.
# [09:20] <Hixie> Lachy: if we didn't correct UTC every now and then, "noon" would end up being at night, and vice versa
# [09:20] <Lachy> But that would take thousands of years to happen
# [09:20] <mjs> couldn't leap seconds be pre-announced and always on the extra day of a leap year (say)?
# [09:20] <hsivonen> Lachy: the basic assumption is that the concept of a "year" and the concept or a revolution around the sun have to be the same
# [09:21] <Lachy> right, but why?
# [09:21] <heycam> old chinese calendars used to use whole leap months
# [09:21] <Hixie> mjs: leap seconds are pre-announced and always at xxxx-12-31 23:59:60.
# [09:21] <hsivonen> Lachy: because some people are attached to the idea of solar time keeping instead of atomic time keeping
# [09:21] <heycam> http://en.wikipedia.org/wiki/Intercalation
# [09:21] <Zeros> heycam, lunar calendars still add a new month
# [09:22] <hsivonen> we've already upgraded from lunar to solar. why not upgrade from solar to atomic?
# [09:22] <Lachy> right, so leap seconds exist simply because people don't use UT, and that problem would go away if they did
# [09:22] * Quits: olli- (olli@80.203.95.229) (Ping timeout)
# [09:22] <Hixie> for what it's worth, i'm perfectly happy with the idea of dropping astronomical time altogether and time zones and having the whole planet use UTC instead of UT1
# [09:22] <Hixie> but most people would not
# [09:22] <heycam> apparently intercalation (insertion of leap seconds/weeks/whatever) is disallowed by islam
# [09:22] <hsivonen> Hixie: isn't UTC the thing with leap seconds and IAT the thing without?
# [09:22] <Lachy> Hixie, yeah, I agree that changing from UTC would probably be like adopting XHTML2 in the real world.
# [09:23] <Lachy> it's just not going to happen
# [09:23] <Hixie> hsivonen: UTC is TAI corrected to match UT1
# [09:23] <Zeros> What does IAT stand for?
# [09:23] <Lachy> international astronomical time, I think
# [09:23] <Hixie> TAI is Temps Atomique International
# [09:24] <hsivonen> Zeros: Internation Atomic Time, but I should have tried French
# [09:24] <Hixie> no, atomic, not astronomical
# [09:24] <Lachy> ah, yes, I was won
# [09:24] <Lachy> wrong
# [09:24] * Lachy hates typing on laptop keyboards
# [09:25] <Hixie> (UTC changes at the rate of TAI, but has leap seconds added to match UT1)
# [09:25] * Quits: Shunsuke (kuruma@133.27.175.115) (Ping timeout)
# [09:26] <Zeros> If we went with Atomic time and segmented the atomic year into months and days, wouldn't time start to drift with respect to the seasons since it wouldn't be coupled with the solar calendar anymore?
# [09:26] <karl> http://www.bipm.org/en/home/
# [09:26] <Hixie> yes
# [09:26] <Hixie> zeros: that's exactly why atomic seconds are needed
# [09:26] <Hixie> er
# [09:26] <Hixie> leap seconds
# [09:27] <hsivonen> Zeros: drifting by a day in 40000 years is not a big deal considering the cultural notion of seasons
# [09:27] <Lachy> so eventually, there could be several hours worth of leap seconds
# [09:27] <karl> ok time to go home.
# [09:27] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [09:29] <Hixie> hsivonen: like i said, the problem is drifting minutes, not days
# [09:33] <Hixie> also, dropping leap seconds would violate our design principles
# [09:33] <Hixie> as it would break compatibility with legacy user agents like sundials
# [09:34] * Joins: Shunsuke (kuruma@133.27.53.98)
# [09:35] <Hixie> hm, this page describes the various aspects of this topic quite well http://www.cl.cam.ac.uk/~mgk25/time/leap/
# [09:36] <mjs> lol
# [09:37] <Zeros> hah
# [09:42] <xover> Heh. An excellent analogy for “Set in Stone”. :-)
# [09:43] * hsivonen notes that as far as versioning flags for timekeeping systems go, the legacy version identifier (GMT) is reused for the new thing (UTC) for backwards compat
# [09:43] * Quits: heycam (cam@131.181.85.133) (Quit: bye)
# [09:43] * Quits: Lachy (chatzilla@131.181.99.92) (Quit: ...and I'm gone.)
# [09:44] <hsivonen> In addition, the name of UT1 suggests that whoever defined it, prematurely added a version number
# [09:47] * Quits: Zeros (Zeros-Elip@69.140.48.129) (Quit: sleep)
# [09:47] <Hixie> UT1 is part of a whole set of time scales
# [09:48] <Hixie> specifically, UT1 is basically UT0 corrected for polar motion
# [09:48] <Hixie> iirc
# [09:48] <Hixie> and UT2 is an averaged UT1
# [09:49] <hsivonen> ok
# [09:49] <mjs> maybe we should use that naming convention
# [09:49] <mjs> instead of HTML5, HTMN
# [09:49] <mjs> to be followed by HTMM
# [09:50] <mjs> the advantage is that the versioning is in the name of the root element
# [09:52] <Hixie> nah the naming scheme for universal time is UT<version>
# [09:52] * Quits: anne5 (annevk@131.181.85.131) (Ping timeout)
# [09:53] <Hixie> UT0, UT1, UT2, UTC, UT1R, UT2R, UTC-SLS, UTS, etc
# [09:56] <xover> It's interesting to see the browser oriented crowd struggle with the same issues the (very much SGML-oriented) Validator has.
# [09:59] <hsivonen> xover: what do you mean? sniffing what some thing is?
# [09:59] <xover> yes
# [10:00] <Hixie> are we still struggling?
# [10:00] <Hixie> i thought we'd resolved it a while back
# [10:00] <xover> Well, the debate seems to be raging on.
# [10:03] <hsivonen> for some reason, the debate keeps going on after issues have been solved and some people in the HTML WG seem to think that having pondered stuff profoundly is somehow a liability
# [10:04] * Quits: Shunsuke (kuruma@133.27.53.98) (Ping timeout)
# [10:04] <hsivonen> I'm referring to the opinions about being worried about Hixie being so involved with HTML5 and about people who have participated in the WHATWG explaining stuff by pointing at the design principles or past discussions
# [10:05] <mjs> that's the sort of argument made by people who care more about being involved than what the resul tis
# [10:06] <hsivonen> as if the people who have been involved in the WHATWG were saying random stuff just to defend "their position" instead of the opinions actually being based on thinking and research
# [10:06] <hsivonen> mjs: right.
# [10:06] <xover> Well, Hixie may have two heads, but most people find it difficult to juggle two hats; particularly when the difference in goals may be subtle.
# [10:06] <mjs> personally, I'm happier if something has a good outcome and I don't need to be involved
# [10:07] * Joins: edas (edaspet@88.191.34.123)
# [10:07] <Hixie> i think i've shown quite obviously that i'm not biased by my employer, since i've had the exact same biases and opinions through three different employers and a year of unemployment
# [10:07] <hsivonen> mjs: yeah, we should care about getting a good spec instead of letting people feel good about making their mark
# [10:07] <Hixie> but i don't really mind people thinking i'm biased
# [10:08] <Hixie> i don't _have_ to be the editor for the html wg, if anything i'd probably be happier not being the editor :-)
# [10:08] * Joins: Shunsuke (kuruma@133.27.53.98)
# [10:08] <xover> hsivonen: That sounds like you're dismissing the dissent on the same basis you complain of them objecting to, say, Hixie or Design Principles.
# [10:09] <xover> Hixie: My point exactly.
# [10:09] <mjs> hsivonen: apparently those who want to make their mark do not want to do so enough to actually be an editor
# [10:09] <mjs> xover: if people dissent for a substantial reason, I'm fine with that
# [10:09] <Hixie> understandably, being editor of this spec is a fulltime job
# [10:10] <Hixie> literally
# [10:10] <hsivonen> xover: what I think I'm saying is that profoundly pondered opinions based on research and reality-based constraints and off-the-cuff remarks don't have equal merit on the face of it
# [10:10] <mjs> xover: if they dissent only because they weren't involved in the original discussion, then I think that is invalid
# [10:11] <xover> mjs: I agree.
# [10:12] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [10:12] <hsivonen> xover: anyway, I found your objection to Hixie's editorship exceedingly weird and illogical
# [10:12] <xover> EXPN?
# [10:13] <hsivonen> xover: as if you wanted the editor to be ignorant to the point of not having opinions. and as if Hixie's expertise was a liability rather than exactly the reason why he should be the editor
# [10:13] <xover> Hmm. I must not have expressed myself clearly then.
# [10:13] <Hixie> yeah i didn't really understand what you mean either
# [10:14] <Hixie> have i not shown an ability to treat my personal opinion as no more important than anyone else's?
# [10:14] <xover> The Editor must work to represent the entire breadth of the group (in many situations).
# [10:14] <mjs> besides just knowledge of web technology, an important requirement for editorship is ability to come up with and clearly write good conformance requirements
# [10:14] <mjs> this is a skill that not many people have, even many web experts
# [10:14] <Hixie> xover: right, isn't that what i've done?
# [10:14] <xover> This means the Editor cannot at the same time, at least not effectively, act as an advocate of a point of view.
# [10:14] <mjs> Hixie is really good at it, and a lot of people aren't
# [10:15] <Hixie> the <img/> and <html xmlns=""> things being probably the most obvious examples of my taking into account input that were blatently opposite to my own opinion
# [10:15] * Joins: marcos_ (chatzilla@203.206.31.102)
# [10:15] * marcos_ is now known as marcos
# [10:15] <mjs> xover: we seem to be doing ok with a Co-Chair who advocates a point of view pretty much full time
# [10:15] <mjs> even though a Chair has more responsibility to appear neutral than an Editor
# [10:15] <xover> Since Hixie has a very strong and well developed point of view, leaving him free to act as an advocate for that point of view would be more valuable than sticking him with the Editor role.
# [10:16] <mjs> well, that depends on whether you think his advocacy of points of view is more valuable than his writing skills
# [10:16] <mjs> I don't think that is the case
# [10:16] <Hixie> yeah, i'd have to say that i care less about my own personal opinions than i care about having a quality spec, and i believe i can deliver hte latter
# [10:16] <mjs> there are lots of people who can advocate useful points of view just as well
# [10:16] <hsivonen> xover: do you mean that people whose point of view is not well-developed should be allowed to feel good about their ill-developed stuff being treated equally to the well-developed stuff?
# [10:17] <Hixie> (but thank you for saying i have good opinions :-) )
# [10:17] <mjs> it's a lot harder to find a good spec writer than a person with lots of opinions
# [10:17] <xover> Hixie: So long as we're clear that I don't necessarily _agree_ with those opinions, you're welcome! :-)
# [10:18] * Joins: gavin_ (gavin@74.103.208.221)
# [10:18] <Hixie> mjs: hah, hear hear
# [10:19] <hsivonen> xover: actually, I'll revise that. sure, well and ill-developed stuff should be considered, but it should follow that the well-developed stuff is upheld unequally often
# [10:19] <mjs> but anyway, it's pretty hard to compare Hixie to Unnamed Person X, or even to Hixie + Unnamed Person X
# [10:19] <xover> hsivonen: The revised version I'll stipulate to. :-)
# [10:20] <mjs> so I can't fairly evaluate a proposal for a different editor or editors without knowing who they would be
# [10:20] <mjs> and we don't seem to be teeming with nominees or volunteers
# [10:20] <Hixie> you're not teeming with me?
# [10:21] <xover> mjs: Note I did _not_ object to Hixie as Editor. I suggested he would be more valuable to the group if he was free of that responsibility.
# [10:21] <mjs> there is no "i" in "teem"
# [10:21] <Hixie> hah
# [10:21] * Quits: marcos (chatzilla@203.206.31.102) (Ping timeout)
# [10:21] <mjs> xover: ok, in that case I think you're quite likely wrong
# [10:22] <xover> Always a possibility.
# [10:22] <mjs> like I said, I think his mechanical skills at spec-writing are harder to replicate than his advocacy of particular technical decisions
# [10:23] <mjs> but I also think it is ok for an editor to have and express opinions, as long as he still considers input from others
# [10:23] <xover> That is a very hard balancing act.
# [10:24] <xover> Which is one reason why I haven't, and won't, volunteer for Editing anything.
# [10:24] <hsivonen> xover: Hixie has an actual track record in that and it is excellent
# [10:25] <hsivonen> we don't know the track record of Mystery Person X
# [10:25] <hsivonen> and we don't know if (s)he would take over self-contained sections (to the extent sections can be self-contained)
# [10:25] <xover> Nobody else in the whole wide world can do as good a job as Hixie at editing th spec?
# [10:25] <hsivonen> and we don't know what kind of political token game is going on if at all
# [10:27] <hsivonen> xover: it's possible for someone else to do as good a job, but you'd have to find the potential people first
# [10:27] <Hixie> xover: i'm sure there are people -- e.g. the xforms spec if pretty good technically -- but the problem is finding one who is good and wants to do it and has the time to do it
# [10:27] <Hixie> xover: i have the advantage of being paid fulltime to do it
# [10:27] <Hixie> :-)
# [10:27] <xover> Yes, and the Chairs seem to be working on finding such people.
# [10:27] <xover> Hixie: Where're you at these days?
# [10:27] <Hixie> google
# [10:27] <Hixie> (specifically google's open source program office)
# [10:28] <mjs> the chairs are working on finding such people for reasons they are unwilling to state publicly
# [10:28] <xover> Hmm. Interesting. Didn't know they cared enough to be involved in this.
# [10:29] <xover> Dan did admit he'd been less than clear about it.
# [10:29] <mjs> and refused to be more clear
# [10:30] <sbuluf> what's google interest in this, btw?
# [10:30] <mjs> so anyway, w/o knowing there reasons, it's hard to tell how long it is reasonable to wait for them to find another good candidate
# [10:30] <sbuluf> something to do with search technology? other?
# [10:32] <citoyen> Google is going for world dominance, as we know, so clearly they have to make sure MS doesn't get it ;)
# [10:32] <Hixie> sbuluf: google's open source program office is basically google's "contribute back to tech society" arm, we do the summer of code, help web standards, that kind of stuff
# [10:33] <xover> Well, being late to the game my patience is still generous on this issue. Should it drag on I may find myself agreeing with you.
# [10:33] * Quits: Shunsuke (kuruma@133.27.53.98) (Ping timeout)
# [10:34] <xover> Hixie: No direct interest in HTML standards for, say, the search engine, then?
# [10:34] <xover> (that's not the "google should rank higher for sites that validate" question rehashed, btw)
# [10:35] <xover> I imagine Google would have a fiarly unique view of metadata issues, for one thing.
# [10:36] <mjs> I would expect Google to have at least long-term interest in improving HTML
# [10:36] <mjs> but as a company, they seem to care about as much as Microsoft cares about improving the x86 instruction set
# [10:37] <mjs> (which is to say, not much)
# [10:37] <hsivonen> xover: to me, HTML5 seem more relevants to Google's browser-based apps than search
# [10:37] <sbuluf> hixie, that contribution might be indiscriminate, or might select areas where the effect helps google back as well. no self-interest in this case?
# [10:39] <xover> hsivonen: Sure. I meant the search engine (their core cusiness, including the web apps) as opposed to just the feelgoody "giving back to the community" Hixie described above.
# [10:39] <sbuluf> hsivonen, probably correct, yes.
# [10:39] * Quits: Ashe`` (Ashe@213.47.199.86) (Connection reset by peer)
# [10:39] * Joins: Ashe`` (Ashe@213.47.199.86)
# [10:40] <Hixie> xover: google as a company doesn't really have any strong opinions, though there are features we would like to see (e.g. cross-site XMLHttpRequest)
# [10:41] <Hixie> xover: i'm basically just paid to work on this as a way of google giving back to the community, since the entire company is basically based on html
# [10:41] <sbuluf> hixie, thanks.
# [10:42] <xover> Hmm. Without belitteling their contribution then, I might wish they'd more involve themselves in this effort.
# [10:42] <Hixie> sbuluf: well don't get me wrong, i mean, i take input back from the company -- e.g. i had a lot of input from the google video and youtube guys for the <video> element
# [10:43] <sbuluf> hixie, oh, so there are preferences
# [10:43] <Hixie> sbuluf: sure
# [10:43] <xover> Right now I'm feeling somewhat steamrollered by the essentially browser-oriented crowd. I would love to see more varied points of view.
# [10:43] <Hixie> sbuluf: though ironically in the case of <video> i think apple got their way more than google :-)
# [10:44] <sbuluf> hixie =P
# [10:44] <Hixie> xover: you don't think google is involved enough? most people seem to think it's some big google-sponsored conspiracy ;-)
# [10:44] <mjs> god forbid anyone should consider Apple's opinion on multimedia, what would we ever know about that :-p
# [10:44] <xover> Hixie: *Everything* is a Google conspiracy. :-)
# [10:45] <sbuluf> xover, some of us have opinions *so* ouit of line, that you would *not* hear them in the mailing list, for example.
# [10:45] <sbuluf> the conspiracy is endless....
# [10:46] <hsivonen> sbuluf: do you want to adopt XHTML 2.0 as the next HTML? :-)
# [10:46] <sbuluf> hsivonen, no. not strict enough.
# [10:46] <mjs> if it's all a Google conspiracy, I want my cut of the bribe money
# [10:47] <xover> I thought all you Apple folk had those nice backdated options? :-)
# [10:47] <xover> (and a crowd of landsharks decends on mjs's IRC client)
# [10:48] <hsivonen> xover: what non-browser points of view you are looking for? authoring tools?
# [10:49] <xover> Authoring tools, conformance checkers, validators (I'm sure you'll agree they differ), screen scrapers, Google's bazillion different web related code,...
# [10:50] <xover> The stuff we _haven't_ considered.
# [10:50] <xover> There seems to be a uniformity to the discussions that I find worrying.
# [10:50] <Hixie> i speak to authoring tools people regularly
# [10:51] <Hixie> hsivonen, whom you're speaking to right now, is a conformance checker author
# [10:51] <hsivonen> xover: my feeling is that the conformance checking point of view has been well heard by the editor
# [10:51] <Hixie> validators are obsolete imho
# [10:51] <xover> Exactly. ANd I'm not hearing the _dissent_ to that point of view.
# [10:51] <Hixie> google's other HTMLWG rep is blind and thus has strong opinions on screen scraping et al
# [10:51] <hsivonen> xover: the name "validator" should not be used as an excuse not to be a conformance checker, in my opinion
# [10:52] <xover> hsivonen: I think you're over-simplifying the position of the v.w3.org team there.
# [10:53] * xover idly wonders whether his involvement with the validator constitutes a reportable affiliation with W3C...
# [10:53] <mjs> Apple ships authoring tools, web services, screen scrapers, an HTML-supporting mail client, etc
# [10:53] <hsivonen> xover: what's the position
# [10:53] <mjs> we also have one of the most visited and most linked sites on the internet
# [10:53] <mjs> I can assure you I talk to all those people regularly
# [10:53] <mjs> even though my main job is on the browser
# [10:54] <xover> hsivonen: So long as conformance is normatively defined in terms of a DTD, SGML Validation is the main deliverable of the Validator.
# [10:54] <xover> hsivonen: But note that I've been the stroingest advocate for extending the Validator to also do conformance checking.
# [10:55] <xover> hsivonen: To the point that I implemented most of the framework that enabled the first steps in that direction in the just released 0.8 beta.
# [10:55] <Hixie> yeah that's the main reason html5 doesn't have even a trace of a dtd
# [10:55] <hsivonen> xover: ok. that position is moot if HTML5 isn't defined in terms of an SGML DTD.
# [10:55] <xover> hsivonen: yes
# [10:55] <hsivonen> xover: would you like it to be?
# [10:56] <xover> I'm undecided. I would like there to _also_ be a DTD, but I'm inclined to think that would be impractical (not impossible, just impractical) given the needs for extensibility.
# [10:57] * Quits: nickshanks (nicholas@195.137.85.17) (Quit: nickshanks)
# [10:57] <xover> Right now my main concern is that I would like there to some kind of formalism for defining conformance; something machine parseable that quacks like a DTD (but probably doesn't quite walk like one).
# [10:59] * Joins: ROBOd (robod@86.34.246.154)
# [11:00] <sbuluf> what's the difference between a validator and a conformance checker?
# [11:01] <xover> An SGML Validator is a fairly well defined term that is basically tied to SGML DTDs (to the exclusion of all else, in some people's opinion).
# [11:01] <xover> A Conformance Checker has a much more complete scope for all the cases we're interested in here.
# [11:02] <xover> In non-SGML contexts, the term “validator” usualy has a different or less well defined meaning.
# [11:02] <hsivonen> xover: fwiw, I think offering a DTD would be actively harmful
# [11:02] <sbuluf> i'm asking about the concepts of "validation" and "checking conformance", independent of everything else. are they different tasks, conceptually?
# [11:03] <mjs> xover: do you care if the machine-parseable formalism is normative?
# [11:03] <hsivonen> sbuluf: validation means checking if a document instance meets the constraints of a schema
# [11:03] <xover> hsivonen: Hmm. I see your point, but I disagree. Or at least weight the results differently.
# [11:03] <xover> mjs: Yes.
# [11:04] * Joins: marcos_ (chatzilla@203.206.31.102)
# [11:04] * marcos_ is now known as marcos
# [11:04] <hsivonen> sbuluf: conformance checking means checking if a document instance meets (machine-checkable) conformance criteria
# [11:04] <mjs> making a formalism that is normative, correct, and expresses all desired machine-checkable constraints is very hard
# [11:04] <xover> mjs: I don't disagree.
# [11:05] <mjs> you could do it in a turing-complete language, like Python or something, but I don't think a reference implementation of a conformance checker would be better than normative requirements in English
# [11:05] <sbuluf> hsivonen, thanks, i do not see any difference yet, though, to be honest.
# [11:05] <xover> mjs: But note that one is allowed to cheat as much as necessary to achieve it. Constraints that are hard or impossible to implement in a schema language can be delegated to prose, without devaluing the normative schema.
# [11:05] <hsivonen> sbuluf: the difference comes if the schema doesn't capture all the conformance requirements
# [11:06] <mjs> sbuluf: there might be conformance criteria that are not expressed by the schema for a language
# [11:06] <mjs> xover: if it is incomplete, then the value of making it normative is limited
# [11:06] <Hixie> xover: yeah but if you do that you end up with validators that aren't full conformance checkers, exactly like we did with html4
# [11:06] <sbuluf> are there some reasons for this to happen? or just because the schema definition is incomplete?
# [11:06] <mjs> I have even seen pathological behavior where groups left out important conformance criteria that they wanted to have because their schema language of choice could not express them
# [11:07] <mjs> for instance, SVG did not allow to allow shapes like circles inside <text>
# [11:07] <xover> HTML 4 was a bastard example.
# [11:07] <mjs> but they did want to allow links, and since <svg:a> can contain shapes too, it is a back door to letting shapes in text
# [11:07] <mjs> and that doesn't actually work in many UAs, since it was not intended to be allowed
# [11:08] <mjs> but afaik there isn't even a textual requirement added for it
# [11:08] <sbuluf> mjs, hsivonen, then the main resons are just incompleteness? (presumable du to authors lazyness or schema language incapacity)
# [11:08] <mjs> sbuluf: schema languages tend to be less expressive than a full programming language
# [11:09] <hsivonen> sbuluf: yes, but the schema language incompleteness is a real problem
# [11:09] <sbuluf> mjs, yes, i can understanda that. but if such is the case...then the two tasks, seem to be basically the same (logicallly, abstractcly)
# [11:09] <hsivonen> sbuluf: yes
# [11:09] <sbuluf> i see
# [11:09] <sbuluf> thank you.
# [11:11] <xover> Hixie: A normative schema does not ipso facto prevent conformance checkers from being complete.
# [11:12] <xover> You're extrapolating from, say, v.w.o's inability to find the necessary developer resources and navigating the then W3C system.
# [11:12] <Hixie> xover: no, but it does prevent them for some strange psychological reason
# [11:13] <Hixie> xover: v.w.o is by far not the only validator for html4
# [11:13] <xover> So your opposition to a normative schema is psychological in nature? :-)
# [11:13] <Hixie> xover: but few if any did anything more than dtd checking
# [11:13] <Hixie> xover: yet with html5, we already have a better conformance checker
# [11:13] <xover> Emphasis on "a".
# [11:13] <Hixie> 1 > 0, to paraphrase a recent idiom
# [11:14] <hsivonen> xover: I'd like to draw attention to the way Validome and Relaxed were "welcomed" on www-validator
# [11:14] <Hixie> xover: it's an objection based on human nature and experience, yes
# [11:14] * Quits: marcos (chatzilla@203.206.31.102) (Ping timeout)
# [11:14] <xover> hsivonen: Please do.
# [11:15] <hsivonen> xover: the reaction to Relaxed was basically that it isn't an "HTML validator" because it doesn't do DTD-based validation
# [11:15] <hsivonen> (ignoring that Relaxed is better)
# [11:15] <xover> URL?
# [11:15] <hsivonen> xover: the reaction to Validome was that it isn't nice that their test cases expose the incompleteness of v.w.o
# [11:16] <hsivonen> (ignoring that Validome is better)
# [11:17] * Quits: kazuhito (kazuhito@210.232.34.13) (Quit: Quitting!)
# [11:17] <xover> hsivonen: Would you like to restate on Validome vs. v.w.o?
# [11:18] <hsivonen> xover: Validome finds conformance criteria violations that v.w.o does not. It also flag violations with no practical consequence (i.e. spec bugs).
# [11:19] <xover> That's a different topic.
# [11:19] <xover> We were discussing its reception on v.w.o.
# [11:20] <hsivonen> http://lists.w3.org/Archives/Public/www-validator/2005Aug/0096.html
# [11:22] <hsivonen> http://lists.w3.org/Archives/Public/www-validator/2005Aug/0103.html
# [11:23] <xover> It sounds like Lachlan found some objective weaknesses in the tool, and Nick praised it as interesting and in line with his own work.
# [11:24] <xover> That was what you wanted to draw attention to?
# [11:24] <hsivonen> xover: yeah, validator developers say it is interesting but still cling on whether it is an "HTML validator"
# [11:25] <xover> Why in the world would you want your "better" tool to be labelled a mere SGML Validator?
# [11:26] <hsivonen> xover: the same reason why the Feed Validator is called what it is called
# [11:26] * Quits: claudio (claudioc@89.97.35.74) (Ping timeout)
# [11:26] <hsivonen> xover: the same reason why my conformance checker carefully mentions the work "validator" in a way that you find it if you look for a validator
# [11:26] * Joins: claudio (claudioc@89.97.35.74)
# [11:27] <hsivonen> s/work/word/
# [11:27] <hsivonen> but I've made sure I have ISO specs backing up my use of the word
# [11:27] <xover> hsivonen: I must be missing your point.
# [11:27] <hsivonen> which is where the Relaxed guys were less careful
# [11:27] <hsivonen> xover: user look for an HTML validator
# [11:27] <hsivonen> users
# [11:28] <hsivonen> in the minds of users "validator" is the generic term for "conformance checker"
# [11:28] * Joins: mw22_ (chatzilla@84.41.169.151)
# [11:28] * xover suspects users google for a "web page checker", but whatever...
# [11:29] <hsivonen> Re: Validome: http://lists.w3.org/Archives/Public/www-validator/2006Apr/0072.html
# [11:30] * Quits: mw22 (chatzilla@84.41.169.151) (Ping timeout)
# [11:30] * mw22_ is now known as mw22
# [11:31] <xover> hsivonen: I'm still waiting for you to make your point. Surely it's not that the v.w.o team had some well reasoned, objective, and constructive, opinions on "competing" tools?
# [11:32] <mjs> I think his point was that having a normative schema leads to a bias to making conformance checking tools that only check what the schema captures
# [11:33] <xover> Well, then he hasn't demonstrated it.
# [11:33] <hsivonen> xover: the Validome comparative marketing was downplayed as "complex cases" failing.
# [11:34] <hsivonen> xover: the reaction was explaining it away
# [11:34] <xover> hsivonen: If I define a test suite where the v.w.o gets a 100% score and your checker only achieves, say, 70%, and then use it to claim that my tool is perfect while yours suck...
# [11:34] <xover> ...what would your reaction be?
# [11:35] <hsivonen> xover: that it is bad faith marketing but that the cases where mine fails and the specs support that diagnosis, *I* have fixing to do
# [11:36] <hsivonen> xover: e.g. Christoph Schneegans published tests that showed a bug in my validator, so I fixed the bug
# [11:37] <mjs> I think his messages show that the w3c validator team does not wish to check conformance violations that don't make the document "invalid"
# [11:38] <mjs> I'm not sure having a state of "invalid" that is a subset of "non-conformant" is all that useful
# [11:39] <xover> May I refer to the logs of this channel where I explicitly state that such was, and remains, the goal of both myself and a significant portion of the other contributors?
# [11:39] <xover> In fact, take a look at the just released beta of 0.8. It's starting to take the first baby steps in that direction.
# [11:39] <mjs> s/does/did/
# [11:40] <mjs> if they have changed their mind, I think that's great
# [11:40] <xover> brb, phone
# [11:46] * Joins: Zoffix_ (Zoffix@99.244.41.243)
# [11:46] <sbuluf> aren't "valid" and "conformant" the same? how could one have a conformance violation that does not make the document invalid?
# [11:46] <sbuluf> )feel free not to answer if i'm too far from the mark. i thought we agreed before that both things were equivalent)
# [11:46] <Hixie> yay, the spec is finally starting to have a sensible browsing contexts section
# [11:47] <Hixie> tomorrow i'll attack the Links section and clean that up
# [11:47] <hsivonen> sbuluf: for HTML 4, valid and conforming are very different
# [11:47] <hsivonen> sbuluf: try <ins datetime='blahblah'>
# [11:48] * Quits: Zoffix (Zoffix@74.111.228.242) (Ping timeout)
# [11:49] <sbuluf> hsivonen, that's a particular case. i'm not talking either about schema language limitations (if they are limited...change them...or skip them altogether). i'm talking about the very idea of validating, or checking for conformance.
# [11:49] <hsivonen> sbuluf: that's a matter of definition
# [11:50] <hsivonen> sbuluf: Hixie could define HTML5 validation to mean HTML5 conformance checking
# [11:50] <hsivonen> sbuluf: expect SGML folks to object vehemently
# [11:50] <Hixie> i'm not going there
# [11:50] <Hixie> HTML5 defines Conformance Checking
# [11:50] <Hixie> and bypasses the entire Validation problem
# [11:57] <xover> mjs (and hsivonen): The referenced messages show, among other things, Olivier bending over backwards to be constructive, objective, and welcoming.
# [11:58] <xover> None of the cited messages support your claim.
# [11:59] <xover> I'm beginning to get the feeling you have some sort of bias against the v.w.o team (for whatever reason).
# [12:00] <hsivonen> xover: me?
# [12:01] <xover> Both of you. (that's not really meant as an accusation; it's a description of my possibly flawed perception right now)
# [12:01] <hsivonen> xover: my points are: 1) the thing people notice about Relaxed is whether "HTML validator" is the right way to call it and 2) in the Validome case, there was an effort explain away the test results (I am not supporting Validome's marketing method here)
# [12:02] <xover> "explain away" is a loaded phrasing.
# [12:02] <hsivonen> point #1 tends to be an issue especially with users of v.w.o not with the team
# [12:03] <hsivonen> that's not a bias about the team but it supports Hixie's psychological point
# [12:03] <Hixie> imho such behaviour is unintentional and is merely a result of the spec having a formal official schema -- purely psychological
# [12:03] <Hixie> not having a formal grammar results in competition in conformance checkers, and thus higher quality tools
# [12:04] <xover> Ok, then your sample is statistically biased; a community of users with a vested interest in a SGML Validation tool will not be representative of the wider web community where it comes to non-SGML based tools.
# [12:04] <Hixie> but for validation, there basically _is_ no wider community, that's the problem
# [12:05] <Hixie> having an official schema is what skews the community, not the sampling method
# [12:05] <Hixie> again, purely imho
# [12:05] <Hixie> anyway i should sleep
# [12:05] <xover> Hixie: I don't entirely disagree with that statement.
# [12:05] <Hixie> nn
# [12:06] <xover> `night
# [12:06] <hsivonen> Hixie: nn
# [12:10] <xover> hsivonen: For reference, the v.w.o team considered the Validome test suite to be a valuable resource and used, and continues to use, it to improve the v.w.o validator.
# [12:10] <hsivonen> xover: yeah, I noticed that later that happened
# [12:10] <xover> I have also often lamented the fact that you, among others, went off and did your own thing instead of helping us improve v.w.o.
# [12:11] <xover> Support for one of the XML-sih Schema systems has been on our wish list for a very long time.
# [12:11] <hsivonen> xover: if I have a bias about the v.w.o team, it is related to that lament
# [12:11] <xover> Oh?
# [12:12] <hsivonen> xover: being considered uncooperative for working on a different code base
# [12:12] <hsivonen> xover: v.w.o is a SGML validator and the front end is written in Perl
# [12:12] <hsivonen> xover: my thing started as a RELAX NG validator for XML
# [12:13] <hsivonen> xover: the tooling was the best for Java
# [12:13] <hsivonen> xover: that is RELAX NG--not DTD
# [12:13] <hsivonen> xover: XML--not SGML
# [12:13] <hsivonen> xover: Java--not Perl
# [12:13] <hsivonen> xover: no reasonable code commonality at all
# [12:14] <hsivonen> xover: moreover, when I started doing the HTML5 thing, the politics at the W3C were against
# [12:14] <xover> I would have opposed any radical changes that were solely triggered by WHATWG requirements, yes.
# [12:14] <hsivonen> xover: I think both technically and politically, it made sense to work on a different codebase instead of coming to the v.w.o team as an outsider and try to pull off big changes
# [12:15] <xover> Your premise is faulty. We did not consider you "uncooperative" for doing your own thing.
# [12:16] <xover> Rather we wished you, and several others, would have found it possible to work within the constraints of v.w.o to achieve your goals because that would have been good for _us_.
# [12:16] * Zoffix_ is now known as Zoffix
# [12:16] <xover> That's the lack of a positive for us, not the existance of a negative in relation to you.
# [12:19] <xover> And if there is any complaint I could make, it would be that developers of "competing" tools tended to come in claiming to be much better (and in some cases, that we suck).
# [12:20] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [12:20] <xover> Compare Liam Quinn (WDG) and Nick Kew (webthing/Valet) who actively participate and support v.w.o (often only by discussing topics or explaining how their own tools differ).
# [12:21] <xover> (and Nick's stuff is _not_ trapped in SGML DTD land; he does much wider conformance and quality testing)
# [12:21] <hsivonen> xover: I think I haven't claimed on www-validator that my tool was better. in fact, I've tried to avoid the topic.
# [12:22] <hsivonen> xover: and having it pointed out to me that v.w.o is open source if I mention a bug implies (in my perception) that I should fix it
# [12:23] <xover> Hopefully, that would be after we've declined to do so — at least immediately — for reasons of lack of developer resources.
# [12:24] <xover> Then again, we're all human so I'm sure we've regularly let our own biases and frustrations get in the way of good relations with our surrounding community.
# [12:25] * Joins: gavin_ (gavin@74.103.208.221)
# [12:25] <xover> And let me note, I've been on the receiving end of more rants about the purity of SGML Validation, and the sacrilege of doing anything but DTD Validation, than I care to count.
# [12:28] <xover> hsivonen: BTW, what's the status of your thesis? I note the "submitted in … fulfillment…” line is overstrike. Is it done? In progress still?
# [12:29] <hsivonen> xover: the professor got sick, so it is in the limbo awaiting his comments and subsequent submission for next month's bureacracy cycle
# [12:30] <xover> heh heh. sorry to hear that.
# [12:31] <xover> You anywhere geographically near Jukka Korpela or Ville Skyttä?
# [12:31] <hsivonen> xover: I think yes, but I've never met either f2f.
# [12:31] <xover> ah, ok
# [12:32] <xover> I'd be interesting to be a fly on the wall the day the three of you met over beers. :-)
# [12:32] <mjs> xover: I think the existence of competing validators that compete over how well they report conformance problems is a good thing
# [12:33] <xover> mjs: I agree.
# [12:33] <mjs> and if they market themselves actively to authors, so much the better
# [12:33] <xover> v.w.o has benefitted greatly from such "competition".
# [12:33] <hsivonen> gotta go
# [12:33] <xover> hsivonen: cya. thanks for your time!
# [12:35] <xover> mjs: Have I somehow given the impression I would disagree with that sentiment?
# [12:36] <mjs> xover: doesn't matter; I'm glad that you do agree
# [12:36] <mjs> I understand the feeling of mildly lamenting someone working on a separate project when it seems like working together might be remotely possible
# [12:37] <xover> Good. That means I'm at least making /some/ sense. :-)
# [12:38] <mjs> there's some people who recently started a browser engine that reimplements the WebKit API as pure Objective-C, supposedly to be better for mobile devices
# [12:38] <mjs> (SimpleWebKit)
# [12:38] <mjs> and I think they are kind of silly
# [12:38] <mjs> but in general I think browser competition is cool
# [12:38] <mjs> makes us all work harder
# [12:40] <xover> Hmm. WebKit isn't ObjC already? I should probably pay more attention to the project in my copious spare time. :-(
# [12:42] <xover> Speaking of which, I need to head off to work. Thanks (all) for the information, and the discussions.
# [12:42] <mjs> just the surface layer is Objective-C
# [12:42] <mjs> the core is C++
# [12:46] * Quits: sbuluf (rits@200.49.140.243) (Ping timeout)
# [12:53] * Joins: marcos_ (chatzilla@203.206.31.102)
# [12:53] * marcos_ is now known as marcos
# [13:01] * Joins: polin8 (polin8@64.81.134.176)
# [13:16] * Joins: kazuhito (kazuhito@210.232.34.13)
# [13:19] * Quits: marcos (chatzilla@203.206.31.102) (Ping timeout)
# [13:33] * Joins: Shunsuke (kuruma@219.110.80.235)
# [13:44] * Joins: hasather (hasather@81.235.209.174)
# [13:51] * Joins: zcorpan_ (zcorpan@84.216.43.109)
# [13:58] * Joins: olivier (ot@128.30.52.30)
# [14:04] * Joins: karl (karlcow@128.30.52.30)
# [14:08] <zcorpan_> "I expect the best course of action will become clear to all in due course and it will matter little what formal process was used." == we will decide about the editors and then we will tell the group?
# [14:19] <hsivonen> zcorpan_: looks interesting, doesn't it?
# [14:20] <zcorpan_> hsivonen: yes
# [14:26] * Parts: hasather (hasather@81.235.209.174)
# [14:27] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [14:29] * Joins: hasather (hasather@81.235.209.174)
# [14:32] * Joins: gavin_ (gavin@74.103.208.221)
# [15:01] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
# [15:10] * Joins: Lachy (Lachlan@124.168.27.56)
# [15:36] * Joins: DanC_lap (connolly@128.30.52.30)
# [15:43] * Joins: h3h (bfults@66.162.32.234)
# [15:43] * Quits: DanC_lap (connolly@128.30.52.30) (Ping timeout)
# [15:44] * Joins: alexf (alejandro@85.152.42.1)
# [15:53] * Quits: kazuhito (kazuhito@210.232.34.13) (Quit: Quitting!)
# [16:17] * Quits: edas (edaspet@88.191.34.123) (Quit: http://eric.daspet.name/ et l'édition 2007 de http://www.paris-web.fr/ )
# [16:29] * Joins: billmason (billmason@69.30.57.156)
# [16:30] * Quits: myakura (myakura@60.239.122.32) (Ping timeout)
# [16:31] * Joins: myakura (myakura@60.239.122.32)
# [16:31] * Quits: xover (xover@193.157.66.5) (Quit: Leaving)
# [16:32] * Parts: hasather (hasather@81.235.209.174)
# [16:34] * Joins: hasather_ (hasather@81.235.209.174)
# [16:35] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [16:40] * Joins: gavin_ (gavin@74.103.208.221)
# [16:44] * Quits: loic (loic@90.29.116.168) (Ping timeout)
# [16:58] * Joins: polin8_ (polin8@64.81.134.176)
# [17:00] * Quits: polin8 (polin8@64.81.134.176) (Ping timeout)
# [17:00] * Quits: polin8_ (polin8@64.81.134.176) (Client exited)
# [17:00] * Joins: polin8 (polin8@64.81.134.176)
# [17:01] * Joins: olli- (olli@80.203.95.229)
# [17:02] * Quits: hasather_ (hasather@81.235.209.174) (Ping timeout)
# [17:03] * Joins: loic (loic@90.29.155.209)
# [17:12] * Quits: olli- (olli@80.203.95.229) (Ping timeout)
# [17:14] * Joins: olli- (olli@80.203.95.229)
# [17:36] * Quits: polin8 (polin8@64.81.134.176) (Quit: polin8)
# [17:41] * Joins: polin8 (polin8@64.81.134.176)
# [17:42] * Joins: kazuhito (kazuhito@222.151.148.139)
# [17:48] * Quits: gsnedders (gsnedders@86.139.123.225) (Client exited)
# [17:53] * Joins: Sander (svl@80.60.87.115)
# [17:54] * Joins: gsnedders (gsnedders@86.139.123.225)
# [18:07] * Quits: loic (loic@90.29.155.209) (Ping timeout)
# [18:07] * Quits: zcorpan_ (zcorpan@84.216.43.109) (Ping timeout)
# [18:22] * Parts: alexf (alejandro@85.152.42.1)
# [18:27] * Joins: loic (loic@90.29.155.209)
# [18:31] * Joins: Voluminous (Voluminous@66.195.32.2)
# [18:42] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [18:42] * Quits: Shunsuke (kuruma@219.110.80.235) (Quit: See you...)
# [18:44] * Joins: Zeros (Zeros-Elip@204.97.106.110)
# [18:44] * Quits: Zeros (Zeros-Elip@204.97.106.110) (Quit: Leaving)
# [18:47] * Joins: gavin_ (gavin@74.103.208.221)
# [18:51] * Joins: hasather (hasather@81.235.209.174)
# [18:51] * Parts: hasather (hasather@81.235.209.174)
# [19:01] * Joins: MikeSmith (mike@219.165.62.28)
# [19:06] * Joins: xover (xover@193.157.66.5)
# [19:10] * Joins: hasather_ (hasather@81.235.209.174)
# [19:18] * Joins: edas (edaspet@88.191.34.123)
# [19:22] * Parts: asbjornu (asbjorn@84.48.116.134)
# [19:25] * Joins: hasather (hasather@81.235.209.174)
# [19:25] * Parts: hasather (hasather@81.235.209.174)
# [19:28] * Joins: zcorpan_ (zcorpan@84.216.40.253)
# [19:29] * Joins: kingryan (rking3@66.92.187.33)
# [19:30] * Quits: loic (loic@90.29.155.209) (Ping timeout)
# [19:32] * Quits: MikeSmith (mike@219.165.62.28) (Ping timeout)
# [19:36] * Joins: asbjornu (asbjorn@84.48.116.134)
# [19:46] * Quits: claudio (claudioc@89.97.35.74) (Quit: Leaving)
# [19:48] * Joins: loic (loic@90.29.155.209)
# [20:29] * Quits: hasather_ (hasather@81.235.209.174) (Connection reset by peer)
# [20:30] * Joins: hasather_ (hasather@81.235.209.174)
# [20:34] * Quits: edas (edaspet@88.191.34.123) (Ping timeout)
# [20:45] * Joins: dbaron (dbaron@63.245.220.242)
# [20:47] * Joins: Zeros (Zeros-Elip@67.154.87.254)
# [20:49] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [20:54] * Joins: gavin_ (gavin@74.103.208.221)
# [21:01] * Quits: myakura (myakura@60.239.122.32) (Ping timeout)
# [21:13] * Joins: edas (edaspet@88.191.34.123)
# [21:34] * Quits: mjs (mjs@64.81.48.145) (Quit: mjs)
# [22:04] * Quits: gsnedders (gsnedders@86.139.123.225) (Quit: Don't touch /dev/null…)
# [22:04] * Joins: gsnedders (gsnedders@86.139.123.225)
# [22:16] <Sander> hrm, do w3c mailinglists have a page somewhere where you can tell the server not to send you a copy of the email if you were in to the to or cc fields already?
# [22:16] * Joins: hasather (hasather@81.235.209.174)
# [22:17] * zcorpan_ thinks that is a client issue, and indeed opera doesn't give me dups
# [22:22] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
# [22:22] <xover> Sander: AFAIK, no. Sorry.
# [22:22] <Sander> You know, I've heard the "that is a [other part of the system] issue" line before - in that tabs really should be done by the OS, not by individual products like browsers. But when that isn't happening, and it's really easy to do it in the browser/mailinglist-software, doing it there definitely is the right thing to do after all.
# [22:22] <Sander> xover: :(
# [22:23] * Sander goes figure out how to reliably filter
# [22:25] <xover> List address in To: field, but no List-ID header present; or something like that...
# [22:33] <zcorpan_> Sander: don't disagree (re tabs)
# [22:34] <zcorpan_> but indeed, it's solvable both on the server and on the client, and one might be more convenient than the other
# [22:34] <hasather_> Sander: procmail? (if you're on unix)
# [22:38] <Sander> I run qmail on my mailserver - afaik that doesn't integrate nicely with procmail. And I'm on win2k locally for the moment.
# [22:39] <Sander> But it's okay, filtering in the client will suffice for now.
# [22:56] * Quits: gavin_ (gavin@74.103.208.221) (Ping timeout)
# [22:59] * Parts: hasather (hasather@81.235.209.174)
# [23:01] * Joins: gavin_ (gavin@74.103.208.221)
# [23:03] * Joins: marcos_ (chatzilla@203.206.31.102)
# [23:04] * marcos_ is now known as marcos
# [23:05] * Quits: kazuhito (kazuhito@222.151.148.139) (Quit: Computer goes to sleep!)
# [23:09] * Quits: zcorpan_ (zcorpan@84.216.40.253) (Ping timeout)
# [23:14] * Joins: zcorpan_ (zcorpan@84.216.43.136)
# [23:15] * Joins: kazuhito (kazuhito@222.151.148.139)
# [23:29] * Quits: polin8 (polin8@64.81.134.176) (Quit: polin8)
# [23:35] * Quits: kazuhito (kazuhito@222.151.148.139) (Quit: Quitting!)
# [23:54] * Joins: mjs (mjs@17.255.97.188)
# [23:57] <Dashiva> No doubt someone will put together a DTD. As long as it's clearly unofficial, it shouldn't do too much harm
# [23:57] <Dashiva> Ack
# [23:57] * Dashiva hurts scroll bar
# Session Close: Fri Apr 20 00:00:00 2007

The end :)