/irc-logs / w3c / #html-wg / 2007-08-24 / end

Options:

# Session Start: Fri Aug 24 00:00:00 2007
# Session Ident: #html-wg
# [00:24] * Quits: tH (Rob@87.102.90.11) (Quit: ChatZilla 0.9.78.1-rdmsoft [XULRunner 1.8.0.9/2006120508])
# [00:25] * Quits: zcorpan_ (zcorpan@84.216.40.15) (Ping timeout)
# [00:27] * Quits: hendry (hendry@89.16.172.32) (Quit: sleep&)
# [00:33] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [00:38] * Quits: heycam (cam@203.214.42.247) (Ping timeout)
# [00:39] * Joins: gavin (gavin@74.103.208.221)
# [00:50] * Joins: hober (ted@69.45.6.105)
# [01:09] * Quits: mjs_ (mjs@17.203.14.224) (Ping timeout)
# [01:17] * Joins: heycam (cam@130.194.72.84)
# [01:17] <heycam> ehe
# [01:17] <heycam> oops ww, sorry
# [01:17] <heycam> (damn focus stealing windows)
# [01:27] * Joins: mjs (mjs@17.255.100.25)
# [01:41] * Joins: mjs_ (mjs@17.203.14.224)
# [01:42] * Quits: mjs (mjs@17.255.100.25) (Ping timeout)
# [01:46] * Quits: mjs_ (mjs@17.203.14.224) (Quit: mjs_)
# [01:48] * Joins: olivier (ot@128.30.52.30)
# [02:16] * Joins: karl (karlcow@128.30.52.30)
# [02:16] <karl> Hixie: http://dev.w3.org/cvsweb/libwww/Robot/src/RobotMain.c is it what you are looking for?
# [02:17] <karl> with regards to W3CRobot
# [02:19] * Joins: sbuluf (agzy@200.49.140.211)
# [02:24] * Quits: dbaron (dbaron@63.245.220.241) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [02:41] * Quits: hyatt (hyatt@17.203.15.154) (Ping timeout)
# [02:41] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [02:46] * Joins: gavin (gavin@74.103.208.221)
# [02:57] * Joins: mjs (mjs@64.81.48.145)
# [02:58] * Quits: kingryan (rking3@208.66.64.47) (Quit: kingryan)
# [03:02] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [03:03] * Joins: gavin (gavin@74.103.208.221)
# [03:05] * Quits: hober (ted@69.45.6.105) (Quit: ERC Version 5.2 (IRC client for Emacs))
# [03:05] <Hixie> karl: which line?
# [03:18] <karl> hixie?
# [03:18] <Hixie> i couldn't find the lines that honour the content-type headers. that's what i was asking about.
# [03:19] <karl> ah in this code. From what I understand
# [03:19] <karl> Robot is using libwww
# [03:19] <karl> so I guess Robot is just a driver for libwww processing
# [03:22] <Hixie> yeah i read both and while i found code that determined what the type was, and found code that sniffed for the type, i couldn't find any code that actually honoured the type.
# [03:22] <karl> ah ok I see. I haven't looked closely at that.
# [03:23] * karl is hunting for the last maintainer
# [03:24] <karl> Henrik was the person used to create the beast
# [03:24] <karl> maybe something to ask him
# [03:24] <karl> frystyk@microsoft.com
# [03:25] <karl> Jose Kahan at w3c too
# [03:25] <karl> maintains the code
# [03:25] <karl> kahan@w3.org
# [03:25] * Joins: beowulf (beowulf@87.198.168.38)
# [03:28] <Hixie> i'm not really worried. i was more curious as to whether Roy could answer that question, since i thought maybe he just gave examples at random hoping i wouldn't verify them. since he didn't reply to my question, maybe my initial guess that most spiders, etc, ignore the content-type even more than browsers do is actually true.
# [03:29] * Quits: beowulf (beowulf@87.198.168.38) (Ping timeout)
# [04:11] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [04:13] <karl> http://lists.w3.org/Archives/Public/www-archive/2007Aug/0052
# [04:13] <karl> test with table, tbody and CSS
# [04:19] * Quits: Lionheart (robin@66.57.69.65) (Ping timeout)
# [04:22] <Philip`> karl: Opera 9.2 and IE7 do nothing special on those tables. FF3 is buggier than FF2 - the bottom table's background extends down the page past the bottom of the table
# [04:23] <karl> ouch
# [04:31] * Quits: aroben (adamroben@17.203.15.195) (Quit: aroben)
# [04:45] <karl> hmm the good thing of reading the IRC logs in the morning, more than being live on the IRC log is that the passion in the debates seems to be useless.
# [04:46] <karl> it encourages me to think, to just read thing most of the time, and not reply except if I really think I can help move forward.
# [04:46] <MikeSmith> I wonder if it would be considered appropriate for HTML WG members to use the alt attribute to embed disclaimers in all their blog postings.
# [04:46] <karl> Each time in my head, I hear a "I disagree" or "No. That's not right". I refrain to send an email.
# [04:53] <MikeSmith> karl, fwiw, iCab seems to display those tables as expected
# [04:58] * Joins: dbaron (dbaron@63.245.220.241)
# [05:23] <Lachy> DanC, yt?
# [05:25] <DanC> sorta... I'd rather you gave me a clue what you want me for when you ping me
# [05:34] <MikeSmith> Lachy - if it's about the public-forms-tf list maintenance issue, I'm trying to check up on it now too
# [05:34] <Lachy> MikeSmith, ok.
# [05:35] <Lachy> MikeSmith, DanC suggested I mail sysreq with a new request, so I'm writing that now
# [05:35] * Quits: dbaron (dbaron@63.245.220.241) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [05:49] <olivier> lachy?
# [05:49] <olivier> what's your web account username?
# [05:49] <Lachy> olivier, it's lhunt
# [05:50] <olivier> k
# [05:50] <olivier> try http://cgi.w3.org/member-bin/list_mgnt.pl now?
# [05:50] <Lachy> works now
# [05:50] <Lachy> thanks
# [05:50] <olivier> cool
# [05:50] <olivier> np
# [05:53] <Lachy> olivier, you might want to close the other thread I just sent to sysreq also
# [05:53] <Lachy> just before you fixed it
# [05:53] <olivier> sure, thanks
# [05:54] <olivier> it hasn't been distributed to sysreq yet, but I'll think of closing it when it does
# [05:58] <DanC> 79 answers in http://www.w3.org/2002/09/wbs/40318/dprv/results
# [05:58] <DanC> 412 non-respondents.
# [05:58] * Joins: dbaron (dbaron@71.198.189.81)
# [05:59] <DanC> dbaron, is that your brain, or just your machine maintaining a connection?
# [05:59] <DanC> I'm interested in your thoughts on the design principles survey http://www.w3.org/2002/09/wbs/40318/dprv/
# [05:59] <dbaron> well, it's me reconnecting to IRC
# [06:01] <DanC> hmm... several others in the non-respondent list that I would have liked to get on their radar
# [06:02] <dbaron> If I fill out the form, will it blow away whatever anyone else from Mozilla filled out?
# [06:02] <DanC> no
# [06:03] <DanC> this is an informal thing where every person's response is separate, not a formal question.
# [06:04] <DanC> 16% tricky to call that representative
# [06:08] * Quits: heycam (cam@130.194.72.84) (Quit: bye)
# [06:09] <karl> danc: it depends on if you align that with http://junkyard.damowmow.com/291
# [06:10] <karl> the list of main (nb of mails) contributors to the mailing list. Though it is quantity, not quality.
# [06:11] <karl> I think I have done a Frenchism
# [06:11] <karl> quality not in the sense of product quality, but in the sense of property
# [06:13] <karl> quality = "a distinctive attribute or characteristic possessed by someone or something"
# [06:13] <karl> ok works in English
# [06:26] <dbaron> DanC, filled out
# [06:26] <DanC> :)
# [06:27] * DanC wanders off...
# [06:30] <dbaron> I'm not sure why the last question used checkboxes rather than radio buttons, though.
# [06:30] <dbaron> (For what the editors can change.)
# [06:32] <DanC> yeah... wasn't very well thought-out
# [06:32] * DanC wanders off better this time...
# [06:36] * Quits: dbaron (dbaron@71.198.189.81) (Quit: 8403864 bytes have been tenured, next gc will be global.)
# [06:43] * Joins: schepers (schepers@128.30.52.30)
# [07:10] * Joins: aroben (adamroben@67.160.250.192)
# [07:10] * Quits: aroben (adamroben@67.160.250.192) (Quit: aroben)
# [07:36] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [07:41] * Joins: gavin (gavin@74.103.208.221)
# [07:47] * Quits: schepers (schepers@128.30.52.30) (Client exited)
# [07:50] * Quits: laplink (link@193.157.66.214) (Ping timeout)
# [07:50] * Quits: xover (xover@193.157.66.5) (Ping timeout)
# [07:55] * Joins: schepers (schepers@128.30.52.30)
# [07:56] * Joins: xover (xover@193.157.66.5)
# [07:56] * Joins: laplink (link@193.157.66.214)
# [08:03] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: Less talk, more pimp walk.)
# [08:32] * Joins: MikeSmith (MikeSmith@mcclure.w3.org)
# [08:59] * Joins: beowulf (beowulf@87.198.168.38)
# [09:00] * Quits: gsnedders (gsnedders@86.145.188.203) (Connection reset by peer)
# [09:01] * Joins: Lionheart (robin@66.57.69.65)
# [09:02] * Joins: gsnedders (gsnedders@86.145.188.203)
# [09:02] * Quits: beowulf (beowulf@87.198.168.38) (Ping timeout)
# [09:11] * Quits: Lionheart (robin@66.57.69.65) (Connection reset by peer)
# [09:35] * Quits: karl (karlcow@128.30.52.30) (Quit: Where dwelt Ymir, or wherein did he find sustenance?)
# [09:43] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [09:44] * Quits: sbuluf (agzy@200.49.140.211) (Ping timeout)
# [09:46] * Joins: ROBOd (robod@86.34.246.154)
# [09:48] * Joins: gavin (gavin@74.103.208.221)
#
# Session Start: Fri Aug 24 09:57:17 2007
# Session Ident: #html-wg
# [09:57] * Now talking in #html-wg
# [09:57] * Topic is 'HTML WG telcon 16 Aug 1700Z http://www.w3.org/html/wg/il16 also logged: http://krijnhoetmer.nl/irc-logs/'
# [09:57] * Set by DanC on Thu Aug 16 18:49:57
# [10:05] * Quits: olivier (ot@128.30.52.30) (Quit: Leaving)
# [10:21] <krijnh> Ping
# [10:21] * Disconnected
# [10:21] * Attempting to rejoin channel #html-wg
# [10:21] * Rejoined channel #html-wg
# [10:21] * Topic is 'HTML WG telcon 16 Aug 1700Z http://www.w3.org/html/wg/il16 also logged: http://krijnhoetmer.nl/irc-logs/'
# [10:21] * Set by DanC on Thu Aug 16 18:49:57
# [10:21] * Quits: Philip` (philip@80.177.163.133) (Ping timeout)
# [11:57] * Disconnected
# [11:57] * Attempting to rejoin channel #html-wg
# [11:57] * Rejoined channel #html-wg
# [11:57] * Topic is 'HTML WG telcon 16 Aug 1700Z http://www.w3.org/html/wg/il16 also logged: http://krijnhoetmer.nl/irc-logs/'
# [11:57] * Set by DanC on Thu Aug 16 18:49:57
# [11:57] * Joins: Philip` (philip@80.177.163.133)
# [12:14] * Disconnected
# [12:14] * Attempting to rejoin channel #html-wg
# [12:14] * Rejoined channel #html-wg
# [12:14] * Topic is 'HTML WG telcon 16 Aug 1700Z http://www.w3.org/html/wg/il16 also logged: http://krijnhoetmer.nl/irc-logs/'
# [12:14] * Set by DanC on Thu Aug 16 18:49:57
# [12:23] * Joins: myakura (myakura@210.227.200.158)
# [12:32] * Quits: MikeSmith (MikeSmith@mcclure.w3.org) (Quit: Less talk, more pimp walk.)
# [12:53] * Joins: tH_ (Rob@87.102.90.11)
# [12:53] * tH_ is now known as tH
# [13:28] * Quits: myakura (myakura@210.227.200.158) (Ping timeout)
# [13:31] * Joins: myakura (myakura@210.227.200.158)
# [13:32] * Joins: Lionheart (robin@66.57.69.65)
# [13:53] * Joins: Lionhear1 (robin@66.57.69.65)
# [13:53] * Quits: Lionheart (robin@66.57.69.65) (Ping timeout)
# [13:58] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [14:03] * Joins: gavin (gavin@74.103.208.221)
# [14:10] * Parts: Lionhear1 (robin@66.57.69.65)
# [14:47] * Joins: polin8 (polin8@64.81.134.176)
# [15:25] <Philip`> "I think it is reasonable to expect people to upgrade their technology within a 5 year period" - IE6 is six years old and still has about 50% of the market, and it seems there are about as many IE5 (~7 years old) users as Opera users
# [15:29] <robburns> Philip: I was thinking that if we actually picked the version of each of the major browsers we're aiming that might be a good idea (not necessarily in the design principle, but in applying that design principle).
# [15:31] <robburns> IE5 and up. Mozilla 6 and up. Any Safari. IE5Mac? and up, iCab (version n?), Opera (version n?)
# [15:32] <robburns> Just to give us a solid metric to know whether we're meeting the degrade gracefully principle (or at least know where we're failing to degrade gracefully)
# [15:34] <Philip`> It would be good if there were statistics of how much each version of each browser is used - almost all the data I've seen combines all the versions together
# [15:35] <Philip`> At least http://www.thecounter.com/stats/2007/August/browser.php splits out IE and Netscape major version numbers
# [15:38] * Joins: matt (matt@128.30.52.30)
# [15:38] <robburns> I guess I should have included Konqqueror as well. I wonder what's in that unknown category (links, lynx, amaya icab, etc)
# [15:39] <Philip`> http://encarta.msn.com/xStylesheets/inline.css - that COLOR:<%=headlinecolor%> is totally not going to work
# [15:40] * Quits: Hixie (ianh@129.241.93.37) (Ping timeout)
# [15:42] <Philip`> I'd guess Unknown might largely include IE/Firefox users with configurations/firewalls/etc that block the user-agent header
# [15:42] <Philip`> but that's totally just a guess
# [15:43] <Philip`> (though based on the evidence that pretty much nobody uses Konqueror, so probably nobody uses Links/Lynx/Amaya/iCab either, so that many Unknowns must involve users of other browsers)
# [15:46] * Quits: xover (xover@193.157.66.5) (Connection reset by peer)
# [15:46] * Quits: laplink (link@193.157.66.214) (Connection reset by peer)
# [15:47] <robburns> on that shortcut link you indicated, I'm getting a MIMe type of image/x-icon
# [15:47] <robburns> http://images.hi5.com/images/favicon.ico
# [15:47] <robburns> that's in the content header anyway. or did you mean from somewhere else?
# [15:50] <Philip`> I get "Content-Type: text/plain"
# [15:51] <Philip`> (according to Firebug and wget and LWP)
# [15:53] <zcorpan_> i get image/x-icon with http://www.rexswain.com/httpview.html and according to opera's info panel
# [15:54] <robburns> Philip: I'm using curl. Weird.
# [15:55] <robburns> MiAl:~ robburns$ curl -I http://images.hi5.com/images/favicon.ico
# [15:55] <robburns> HTTP/1.1 200 OK
# [15:55] <robburns> Server: Apache
# [15:55] <robburns> Last-Modified: Tue, 07 Aug 2007 21:15:10 GMT
# [15:55] <robburns> ETag: "f5c-37e-4372285337380"
# [15:55] <robburns> Accept-Ranges: bytes
# [15:55] <robburns> Content-Length: 894
# [15:55] <robburns> Content-Type: image/x-icon
# [15:55] <robburns> Date: Fri, 24 Aug 2007 13:55:19 GMT
# [15:55] <robburns> Connection: keep-alive
# [15:55] <robburns> Cache-Control: max-age=31536000
# [15:55] <robburns> Cache-Control: max-age=31536000
# [15:55] <robburns> Expires: Wed, 06 Aug 2008 21:22:05 GMT
# [15:55] <Philip`> What IP address do you get for image.hi5.com?
# [15:55] <robburns> Philip: that's the input and output from curl
# [15:56] <Philip`> $ curl -I http://images.hi5.com/images/favicon.ico
# [15:56] <Philip`> HTTP/1.1 200 OK
# [15:56] <Philip`> Content-Type: text/plain
# [15:56] <Philip`> Server: TUX/2.0 (Linux)
# [15:56] <Philip`> Content-Length: 894
# [15:56] <Philip`> ETag: "894-eglioanj"
# [15:56] <Philip`> Accept-Ranges: bytes
# [15:56] <Philip`> Last-Modified: Tue, 07 Aug 2007 21:15:05 GMT
# [15:56] <Philip`> Date: Fri, 24 Aug 2007 13:56:48 GMT
# [15:56] <Philip`> Connection: keep-alive
# [15:56] <Philip`> Cache-Control: max-age=31536000
# [15:57] <robburns> Philip: IP is 204.13.51.238
# [15:57] <Philip`> Oops, I meant images.hi5.com
# [15:57] <robburns> maybe it's a proxy issue? Or load balancing with an errant member?
# [15:58] <Philip`> Oh, it's Akamai
# [15:58] <Philip`> images.hi5.com is an alias for images.hi5.com.edgesuite.net.
# [15:58] <Philip`> images.hi5.com.edgesuite.net is an alias for a269.g.akamai.net.
# [15:58] <Philip`> which would probably explain why I see something totally different to what you see
# [15:59] * Quits: Lachy (chatzilla@124.170.133.138) (Ping timeout)
# [15:59] <robburns> Philip: I get this response on lookup
# [15:59] <robburns> a269.g.akamai.net. 12 IN A 204.2.225.42
# [15:59] <robburns> a269.g.akamai.net. 12 IN A 204.2.225.17
# [16:00] <Philip`> Okay - if I try downloading that file from that IP, then it does give image/x-icon
# [16:00] <Philip`> but if I try from 84.53.179.65 then it gives text/plain
# [16:00] <robburns> Yeah, I was thinking maybe Akamai. Does that mean they have a misconfigured server somewhere?
# [16:01] <Philip`> (and I get Server: TUX/2.0 from 84..., and Apache from 204...)
# [16:02] <Philip`> It sounds like they do
# [16:03] <robburns> Philip: so there we have a mapping configuration problem. The same file, but the linux server doesn't know the .ico filename extension
# [16:04] <Philip`> From hi5.com's front page: http://media01.crackle.com/1/w/kd/8jpkb_c300.flv - text/plain video, hooray
# [16:04] <robburns> this is the kind of investigation I was raising on list. Trying to determine whether authors aren't getting the correct filename extensions on their files (like the .html CSS file you found) or servers not getting configured with the right mapping (like this case), when new filename extensions come along
# [16:06] <zcorpan_> i would presume that the latter is more common
# [16:06] <robburns> text/plain video. nice. Does that work in any browser (as the main resource or as a sub-resource)?
# [16:06] <zcorpan_> e.g. .wmv files as text/plain is pretty common
# [16:06] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [16:07] * Joins: laplink (link@193.157.66.214)
# [16:07] * Joins: xover (xover@193.157.66.5)
# [16:07] <robburns> zcorpan: Philip did find a CSS file with a .html extesnsion
# [16:07] <Philip`> http://ad.accelerator-media.com/st?ad_type=pop&ad_size=0x0&section=17629&banned_pop_types=29&pop_times=1&pop_frequency=43200 - no Content-Type at all
# [16:07] * Joins: Lachy (chatzilla@124.170.133.138)
# [16:07] <Philip`> robburns: The video is just used by Flash, which I assume ignores the content-type equally in any browser
# [16:07] * Joins: Hixie (ianh@129.241.93.37)
# [16:07] <zcorpan_> robburns: right, i'm not saying the former doesn't occur
# [16:08] <robburns> zcorpan: No,I didn't mean to imply that. I tend to think the mapping is the bigger problem too. But that's just a hypothesis
# [16:08] <zcorpan_> ok
# [16:09] <Philip`> http://media01.crackle.com/1/d/0b/6spgb_tnw.jpg - text/plain image
# [16:09] <Philip`> hi5.com is an interesting example of everything you can do wrong
# [16:10] <robburns> Philip: but they're getting Akamai's help with that exploration of everything you can do wrong (which is a little surprising to me)
# [16:11] * Quits: Lachy (chatzilla@124.170.133.138) (Ping timeout)
# [16:11] * Joins: Lachy (chatzilla@124.170.133.138)
# [16:11] * Joins: gavin (gavin@74.103.208.221)
# [16:12] <Philip`> media01.crackle.com seems to be non-Akamai
# [16:12] <Philip`> so it's not entirely Akamai's fault :-)
# [16:14] <robburns> No, there's definitely other problems, I'm just surprised to see it from a company whose only purpose is basically to serve up content.
# [16:17] * Quits: Lachy (chatzilla@124.170.133.138) (Ping timeout)
# [16:22] * Joins: Lachy (chatzilla@124.170.133.138)
# [16:25] * Joins: billmason (billmason@69.30.57.156)
# [16:40] <robburns> the louvre example you gave is weird. Would there be any security reason, that they might return a different content-type for a header alone than with GET? I'm getting a different mangled content type almost every time, but when I do GET I get the correct one . ("http://www.louvre.fr/llv/commun/home.jsphas a half-random string of bytes like "�:l/css" ")
# [16:57] <Philip`> I can't imagine any intentional reason for handling GET and HEAD differently in that way - it looks like broken C code that accidentally overwrites the string buffer before it's transmitted
# [16:58] <Philip`> They do the same for HEAD requests on other resources, seemingly always replacing the first 4 bytes with random junk
# [17:44] * Joins: oedipus (oedipus@70.21.186.99)
# [18:14] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [18:19] * Joins: gavin (gavin@74.103.208.221)
# [18:27] * Joins: Goubil (Fr@86.220.48.7)
# [18:28] * Quits: Goubil (Fr@86.220.48.7) (Quit: Goubil)
# [18:44] * Quits: myakura (myakura@210.227.200.158) (Quit: Leaving...)
# [18:46] * Quits: gsnedders (gsnedders@86.145.188.203) (Quit: gsnedders)
# [18:50] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
# [18:56] * Joins: gsnedders (gsnedders@86.145.188.203)
# [19:08] * Joins: kingryan (rking3@208.66.64.47)
# [19:14] * Quits: matt (matt@128.30.52.30) (Quit: matt)
# [19:29] * Joins: aroben (adamroben@17.203.15.195)
# [20:06] * Joins: matt (matt@128.30.52.30)
# [20:21] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [20:26] * Joins: gavin (gavin@74.103.208.221)
# [20:36] * Joins: ROBOd (robod@86.34.246.154)
# [20:48] * Joins: dbaron (dbaron@63.245.220.241)
# [21:02] * Quits: oedipus (oedipus@70.21.186.99) (Quit: oedipus)
# [21:17] * Joins: oedipus (oedipus@70.21.186.99)
# [21:21] * Quits: oedipus (oedipus@70.21.186.99) (Quit: oedipus)
# [21:22] * Quits: aroben (adamroben@17.203.15.195) (Quit: aroben)
# [21:30] * Philip` wonders if he should bother adding voting support to his version of the issues list
# [21:38] * Joins: oedipus (oedipus@70.21.186.99)
# [21:43] * Quits: oedipus (oedipus@70.21.186.99) (Quit: oedipus)
# [21:44] * Joins: hober (ted@68.107.112.172)
# [21:45] <gavin_> robburns: "Well why target an accessibility feature other than because its an accessibility feature." ?
# [21:45] * Joins: hyatt (hyatt@24.6.184.144)
# [21:46] <gavin_> surely you're not seriously making the argument that it was "targeted" only because it was an accessibility feature?
# [21:48] <Philip`> "it works in those browsers [IE and WebKit] and it provides accessibility benefits" - is that meaning that the current implementations in current versions of those browsers provides accessibility benefits?
# [21:48] <robburns> gavin: I'm merely trying to point out that the features most subject to ridicule by Maciej and others just happen to be features crucial to disabled users. Then we hear comments all over the place that: "there's just this big misunderstanding. we're have nothing against accessibility". So I'm just wondering WHY, these bizarre pronouncements of eliminating things that haven't been eliminated always happen to relate to accessibility. What's going on there?
# [21:49] <robburns> Philip: yes I believe they do
# [21:50] <robburns> Philip: it doesn't provide the title attribute on mouse over of the area (which is the bug that iI reported). However, that's a very minor bug in those implementations.
# [21:50] <Philip`> What are those benefits? I've not seen IE or WebKit ever do anything at all with it (except that IE parses it to "useMap")
# [21:50] <Philip`> As far as I can tell, they simply don't have implementations of usemap
# [21:51] <robburns> I'd have to double-check on this, but my understanding is that an aural browser user can activate the input by activating an area instead.
# [21:51] <mjs> robburns: the current HTML5 draft drops a lot of things that were in HTML4 from the conforming language
# [21:51] <mjs> robburns: there just aren't a lot of vocal complaints for the ones that aren't nominally accessibility-related
# [21:51] <gavin_> robburns: I don't think your first statement is true, nor your last
# [21:52] <Philip`> robburns: The usemap gets ignored when tabbing through the page
# [21:52] <robburns> mjs: well I'm not even sure if anything has been done with <input usemap> in HTML5. I never said it had. However the discussion of <input usemap> here clued me in to the fact that it is broken in one way or another in every major browser. Whenever I hear about a broken feature like that I report it to at least WebKit and Mozilla (which are the what I use the most)
# [21:53] <Philip`> (Firefox and Opera do handle tabbing usemaps, but pretty buggily even for <img usemap>)
# [21:53] <robburns> gavin: I'm not sure what I said first.
# [21:54] <gavin_> robburns: your first was "the features most subject to ridicule [...] happen to be crucial to disabled users"
# [21:54] <Philip`> style="" was dropped, and that's unrelated to accessibility
# [21:54] <robburns> Philip: well I'd be curious to hear how something like Emacspeak handles it though.
# [21:54] * Joins: aroben (adamroben@17.203.15.195)
# [21:54] <robburns> That's where the accessibility benefit would come in handy.
# [21:54] <gavin_> robburns: the last was "these bizarre pronouncements [...] always happen to relate to accessibility"
# [21:54] <robburns> Philip: yeah, I thought of that just as I hit enter.
# [21:55] <mjs> dropping style="" did generate some objections, but not a conspiracy theory
# [21:55] <robburns> Philip: However, even the dropping of style= seems to have more of a pedantic rationale than anything reasonable.
# [21:56] <robburns> gavin: Both those statements reflect my impressions. And I know they reflect the impressions of others too.
# [21:58] <jgraham> robburns: I personally think that dropping style is a bad idea. Unlike some members of the html-wg I have not felt the need to flame anyone or try to suppress communication of the fact that style has been dropped
# [21:58] <jgraham> that is, some members of the group about other issues
# [21:59] <jgraham> Indeed, if people say "HTML 5 doesn't have the style attribute" and there are real reasons to retain it I think it helps my case
# [21:59] <robburns> mjs: I'm not suggesting any conspiracy theories. However, we never hear reasonable discussion about these things. We go back and forth about whether there's a use case and such: even though use cases have been provided. But you have to wonder what could be so important as to drop a feature like say TH@abbr form HTML5. If that somehow got in there, what damage could it do to you and others? I can think of some who would be hurt by it not being there, but i
# [21:59] <jgraham> Because more people will come forward and present those reasons.
# [21:59] <robburns> jgraham: style has not been dropped. No design decisions have been made. That's what our chair has said
# [21:59] <jgraham> I have documented some of those in the wiki. I trust that in due course the reasons will be considered
# [22:00] <mjs> robburns: your last message got cut off
# [22:00] <jgraham> I have not flamed anyone because that analysis has not happened yet
# [22:01] <mjs> robburns: complaining to the chairs about what I say in my own bug tracker is completely out of line
# [22:01] <jgraham> robburns: I think that's an incredibly petty distinction to make
# [22:01] <mjs> robburns: please stop your abusive behavior regarding bugs.webkit.org
# [22:01] <robburns> " I can think of some who would be hurt by it not being there, but it's hard for me to understand what problem is created by including it."
# [22:02] <gavin_> who said that?
# [22:02] <robburns> mjs: come on. You really believe I'm being abusive?
# [22:02] <mjs> robburns: yes
# [22:02] <robburns> gavin: I was finishing the part of my message that got cut off. I said that.
# [22:02] <jgraham> robburns: How familiar are you with the dynamics of large bug tracking systems
# [22:03] <gavin_> oh
# [22:03] <jgraham> ?
# [22:03] <gsnedders> robburns: yes.
# [22:03] <gavin_> I don't think "I see no problem with including it" should be the criteria for adding things to the spec
# [22:03] <robburns> gsnedders: yes?
# [22:03] <robburns> jgraham: I'm fairly well familiar. I've been following WebKit and Mozilla for some time now.
# [22:04] <gsnedders> robburns: re: being abusive
# [22:04] <gavin_> (and that ignores problems like scope-creep)
# [22:04] <gavin_> anyways, I need to go
# [22:04] <mjs> robburns: "fix my bug or I'll tell the teacher you were bad" is an abusive behavior
# [22:04] <robburns> gavin: it's not the criteria. The justifications have been given. But the response is usually something like: " but I just don't like it"
# [22:04] <jgraham> Then you will have noticed that comments that do not lead to the resolution of a bug are not appreciated
# [22:04] <robburns> mjs: who said that
# [22:05] <robburns> jgraham: I don't know if I've noticed that. It's hard to gauge what effect any comment has on the final resolution of the bug
# [22:05] <mjs> what else is the point of complaining that I provided reasons that maybe it shouldn't be fixed?
# [22:05] <robburns> jgraham: just like with out WG, I see a lot of misunderstanding and talking past each other in bug tracking comments
# [22:06] <mjs> if you think the current draft of HTML5 should be ignored in deciding what to do about your bug, you can say so in the bug
# [22:06] <mjs> complaining to the chairs is crazy and abusive
# [22:06] <mjs> the HTML WG chairs are not in charge of WebKit development
# [22:06] <gsnedders> robburns: quite frankly, as a member of the HTML WG, I don't care what Apple decide to do within their own project regarding the current draft. I don't feel that you need to email 450 people about it.
# [22:06] <jgraham> robburns: I'm not familiar with webkit but in bugzilla.mozilla.org discussion should in general be held outside the bug report and summarised in the bug report
# [22:06] <robburns> mjs: you're misrepresenting the work of our WG. That's the problem. If you weren't a member of the WG, I wouldn't say anything to anyone. However, you are and I expect the members of the WG to not go around misrepresenting the work of the WG. Especially since you even said you don't even care about this bug.
# [22:07] <mjs> and your actions are making a bad impression on those who are actually in charge (such as me)
# [22:07] <robburns> fgsnedders: I emailed two people about it.
# [22:07] <robburns> mjs: I really don't care what impression I make on you. It just doesn't matter to me at all
# [22:07] <gsnedders> robseveral emails in the "Answering the question..." thread theme to be about this
# [22:08] <gsnedders> *robburns:
# [22:08] <mjs> robburns: do you care about having your bugs addressed or do you just want to grandstand?
# [22:08] <robburns> jgraham: that's not how I see mozilla's bugtracker being used.
# [22:08] <robburns> mjs: I don't really trust you to address my bugs or any bugs that Hixie says you can't address
# [22:09] <gsnedders> robburns: why do you think we all treat Hixie's word as ultimate truth?
# [22:09] <Hixie> wait, you don't treat my word as the ultimate truth? crap.
# [22:09] <robburns> gnsedders: I responded in that thread, but I didn't start that thread. The chair actually brought the issue to the WG
# [22:09] <mjs> robburns: if you think Hixie is in control of what code changes we make then you are sadly mistaken
# [22:09] <Hixie> awww
# [22:09] <robburns> gnsedders: mostly
# [22:09] * Hixie pouts
# [22:09] <mjs> robburns: but in any case, if that is your attitude, don't file bugs
# [22:10] <robburns> mjs: I think you're not in as much control of code changes at WebKit as you seem to think you are
# [22:10] <mjs> filing them just to generate conflict is abusive
# [22:10] <robburns> mjs: why should I not file bugs?
# [22:11] <gsnedders> robburns: I don't think any of us will say we agree with everything in the draft.
# [22:11] <gsnedders> Hixie: I know, disappointing for you :P
# [22:11] <jgraham> robburns: Bug trackers are tools for people who work on code to organise their work
# [22:12] <mjs> robburns: if you file bugs with the expectaction that they won't be addressed, but then you complain about them, then you are doing it just to generate conflict
# [22:12] <jgraham> they are not there for you to make political statements
# [22:12] <mjs> which is not an appropriate use of the bug tracker
# [22:12] <mjs> it's there to help us track problems and organize our work
# [22:12] <robburns> gsnedders: I know. However several times I've seen a line get added to the draft and suddenly everyone's in PR mode: "got to get the word out about the latest pronouncement". And I've seen that even after its clear you don't agree with the change.
# [22:13] <robburns> mjs: who said I don't expect the bugs I file to get addressed?
# [22:13] <mjs> Hixie: sorry Hixie - maybe we should have discussed this at the last cabal meeting
# [22:13] <mjs> robburns: you did: <robburns> mjs: I don't really trust you to address my bugs or any bugs that Hixie says you can't address
# [22:14] <robburns> mjs: I meant you personally. I'm not even sure I want you coding against the bugs I file. Not with the logic I see you apply in these discussions.
# [22:14] <gsnedders> robburns: and that''s another straight insult.
# [22:14] <Hixie> mjs: i'm already putting it on the agenda for our next meeting!
# [22:15] <robburns> gsnedders: it's insulting to have mjs tell me about better bow to him! that's about the most insulting thing a person has ever said to me
# [22:16] <jgraham> robburns: wtf are you talking about?
# [22:16] <gsnedders> robburns: when did he ever do that!?
# [22:16] <Hixie> robburns: telling people about changes is a good way to get feedback on them -- and sometimes there are things we don't agree with (like the <br/> thing, which i still think is stupid, but the arguments were rational and convincing and so i had to put it in the spec), but we still want to get feedback on them
# [22:18] <mjs> robburns: insulting my coding skills is (a) laughable and (b) not a particularly effective technique for persuasion
# [22:18] <robburns> <mjs> said: " and your actions are making a bad impression on those who are actually in charge (such as me)" and then "robburns: do you care about having your bugs addressed or do you just want to grandstand?"
# [22:19] <gsnedders> neither are an insult. what you said most certainly is.
# [22:19] <jgraham> robburns: But those things are statements of fact.
# [22:19] <robburns> mjs: why do you think I'm trying to persuade you.?
# [22:19] <mjs> yes, I am expressing the shocking opinion that if you want to ask an open source project to do something, you should be polite to the developers
# [22:19] <robburns> gsnedders: I was insulted, so to me that makes it an insult
# [22:20] <jgraham> Seriously if you think that pissing people off will get them to work harder for you're much mistaken
# [22:20] <robburns> mjs: you're one developer in a larger group. I think I can afford to lose one.
# [22:21] <jgraham> That's astonishing
# [22:21] <robburns> jgrham: who said I was trying to get mjs to work for me at all?
# [22:21] <robburns> jgrham: for all I know mjs spends all his time on IRC
# [22:22] <mjs> robburns: I also happen to be one of the most prolific and influential developers on this particular project
# [22:22] <gsnedders> robburns: if you're attitude is just "ack, I can afford to fall out with one person" you aren't doing that well sociably.
# [22:22] <mjs> picking a pointless fight with me is likely to make others assume you are a troll
# [22:22] <jgraham> I'm sorry, you seem to be so far beyond my comprehension that I'm not sure I can say anything useful
# [22:22] <Hixie> yeah, i don't think it's possible to "lose one" if that one is mjs
# [22:22] <Hixie> since if you lose mjs you've probably lost the whole webkit team
# [22:22] <robburns> mjs: misrepresenting the work of our WG will definitely make others think your a troll (especially if it gets out)
# [22:23] <robburns> Hixie: what do you mean? Like some sort of employee rampage?
# [22:23] <robburns> :-)
# [22:23] <jgraham> robburns: No one outside the working group _actually_ cares about WG politics
# [22:24] <Hixie> robburns: the webkit community tends to follows maciej's lead, if they see him not respecting someone, they tend to all not respect him
# [22:24] <jgraham> they care about their experience as HTML developers/ end users / etc.
# [22:24] <robburns> Hixie: if you say so, I guess I have to believe it, right?
# [22:25] <jgraham> robburns: I'm not a member of the webkit community but I have to say I respect you less after this discussion. So it seems plausible.
# [22:25] <Hixie> robburns: no, you don't have to take my word for it. i'm just giving you advice, it's up to you what you do with it.
# [22:25] <jgraham> I hope you rise in my estimation again
# [22:26] <Hixie> robburns: i'm just making an observation based on my last few years of working closely with the webkit team
# [22:26] <robburns> graham: I agree with you. Which is just another reason why WG members shouldn't be misrepresenting our work and causing our internal politics to spill out onto these other venues.
# [22:26] <robburns> jjgraham: I'm not worried about that.
# [22:26] <robburns> Hixie: ok
# [22:26] <jgraham> robburns: I am though.
# [22:29] <gsnedders> robburns: if you don't care about falling out with one person, do you care about falling out with hundreds of people? You should really try to coexist with people with other views from yourself.
# [22:29] * Quits: gavin (gavin@74.103.208.221) (Ping timeout)
# [22:30] <gsnedders> I'm getting very close to leaving the WG in part due to people with attitudes like yours, and just things getting less and less polite.
# [22:30] <robburns> gsnedders: I do try to coexist. However, if you or mjs or anyone else tell me I have to bow down to them or they won't like me or will do poorly in their job just to spite me, well then so be it.
# [22:31] <gsnedders> robburns: I don't think any of us think that. I don't think that you need to insult us, either.
# [22:31] <robburns> gsnedders: I'm here on this IRC channel to try to understand you guys. And that's despite the fact that most people who disagree with you, think of this IRC as a hostile place.
# [22:32] <robburns> gsnedders: how did I insult you?
# [22:32] <gsnedders> robburns: maybe you didn't insult me, but you've definitely insulted others here. I don't like having such attitudes within a WG.
# [22:33] <jgraham> This is all so depressing
# [22:33] * Philip` thinks IRC should have more rainbow colours to make people happy
# [22:33] <robburns> gsnedders: My understanding of what just transpired here was that someone asked me about the WebKit bug tracker issue.. I tried to politely answer their questions. mjs did not like me saying he should not misrepresent our WG on the WebKit bug tracker. mjs then told me that if I kept insisting he stop with the misrepresentation he would make sure no bug I filed got fixed.
# [22:34] <mjs> robburns: I don't want anyone to bow down to me, I just want you to be polite
# [22:34] <robburns> gnsedders: only after that did I question whether I'd want mjs fixing any bug i filed.
# [22:34] <robburns> gsnedders: and that is just a statement of fact
# [22:34] <gsnedders> I also don't like that a majority of public-html email at the moment isn't about technical issues, which the mailing lists exists to discuss, but questioning how people act outside the WG.
# [22:34] * Joins: gavin (gavin@74.103.208.221)
# [22:34] <robburns> mjs: I have been polite.
# [22:35] <mjs> I don't think I misrepresented anything, I said "likely to be removed", which is my own judgment of what is likely to happen
# [22:35] <mjs> does anyone else here think robburns has been polite?
# [22:35] <robburns> mjs: that is a misrepresentation when no design decisions have been made
# [22:35] * jgraham changes his IRC colour scheme
# [22:35] <gsnedders> robburns: I don't think you have. I think mjs has been more polite than you throughout. He's also several times made it clear that things are his own opinion.
# [22:35] <robburns> mjs: why don't you show me where I've been impolite. quote something.
# [22:36] * Philip` wonders why http://emacspeak.sourceforge.net/info/html/ puts the logo image in a <table>
# [22:36] <robburns> mjs: to take a straw poll about politeness when I come in here alone to try to understand you guys is very impolite.
# [22:37] <gsnedders> robburns: you stated "I have been polite.". Is it really that impolite to question whether other people agree with your statement about yourself?
# [22:38] <mjs> ok, I'm going to go back to real work
# [22:38] <robburns> Philip: I don't think it's necessarily a bad thing to put an image in a table. I have been using that move to create the figure element (on my own cowpath) that HTML5 proposes. That way you get the caption-side property.
# [22:38] <gsnedders> I'm leaving, and if we don't get back on-topic (to technical issues) on the mailing list in the next 24 hours, I'm leaving the WG completely.
# [22:39] <zcorpan_> gsnedders: :(
# [22:39] <jgraham> Indeed
# [22:39] <hober> gsnedders: That's unfortunate, but certainly understandable.
# [22:40] <robburns> Philip: though that's not how emacspeak is using it I see (no caption). But almost. It could be that it was used that way at some point.
# [22:41] * Philip` tries to work out how to make Emacspeak run
# [22:43] <robburns> Philip: I can't help you. I've never used it myself.
# [22:45] * Quits: ROBOd (robod@86.34.246.154) (Quit: http://www.robodesign.ro )
# [22:47] <robburns> Philip: I was just rereading the thread here. As long as IE and Safari place the usemap attribute value into the DOM, then accessibility solutions can make use of it (I'm not aware of any that do). However, imagine if Fire Vox was ported to use Safari. Then Fire Vox itself could ensure the form was submitted with the right coordinates upon activation of an area.
# [22:48] <robburns> I'm not saying this is all working great, but there are just a lot of little nagging bugs that need to be fixed to make it work.
# [22:50] <Philip`> Aha, now Emacs talks to me
# [22:52] * Joins: aroben_ (adamroben@17.255.110.16)
# [22:53] * Quits: aroben (adamroben@17.203.15.195) (Ping timeout)
# [22:54] * Quits: polin8 (polin8@64.81.134.176) (Quit: polin8)
# [22:54] * Quits: hyatt (hyatt@24.6.184.144) (Quit: hyatt)
# [23:09] <Philip`> Hmm, it sort of works with w3m but it looks like just reads the screen and I can't see it doing anything HTML specific
# [23:11] <robburns> Philip: really? Well maybe I'll email T.V. Raman. He's on the WG and the developer of the app
# [23:13] <Philip`> It looks like w3 is the preferred browser
# [23:14] <Philip`> but Gentoo doesn't make that trivial to install with Emacspeak, since one uses xemacs and the other uses emacs
# [23:14] <Philip`> so I'll keep fiddling with it a bit :-)
# [23:15] <robburns> Philip: thanks.
# [23:15] <robburns> I feel a bit like I'm grasping at straws here. I'd like to find out this works somewhere and I"m curious about whether it does. On the other hand, I think its a feature worth getting right and getting implemented. There seems to be some confusion that making it work with the image map will make it fail with the form submission. But I don't see any reason those are mutually exclusive. the only content that could possibly be hurt by fixing this, is conten
# [23:16] <Philip`> "is conten<truncation>"
# [23:16] <robburns> ok
# [23:16] <Philip`> ?
# [23:16] <robburns> "by fixing this, is content that meant to use <img usemap> and was not targeted at IE or Safari (where it wouldn't work the way that author intended.)."
# [23:16] <Philip`> Ah
# [23:19] <robburns> My reading of HTML 4.01 on this is that an input that uses an image map should still be an input. The areas help someone navigate the input image map. When activating the area, the form submits, sending the coordinates to the server. The part that's not quite specified clearly in HTMl 4.01 is that some coordinates need to be sent even if the image map is activated from an area with the keyboard (and not with a mouse click). I'm just trying to make sense o
# [23:20] <robburns> to make sense of what's written in the recommendation. It does look to me like, late in the game, someone said shouldn't image maps apply to <input image> too?
# [23:24] * Quits: matt (matt@128.30.52.30) (Quit: matt)
# [23:32] * Joins: hyatt (hyatt@24.6.184.144)
# [23:55] <Philip`> robburns: Do you know of particular cases where server-side image maps are used? All I've been able to find is a load of <input type=image> used as simply a button (ignoring the coordinates), and http://terraserver-usa.com/image.aspx using it to select a point on a map
# [23:56] <Philip`> (I'm trying to look for cases where somebody would want to provide tooltips on their server-side image map)
# [23:58] <robburns> Philip: I think that this gets done more often through more cumbersome means. Like using image slices to create many <input image> elements. Then there's not necessarily a server-side image map. However, it requires a resource download for every slice in the map
# [23:58] <robburns> Philip: I don't actually have any URLs to point you to (off of the top of my head), But I will try to dig some up.
# Session Close: Sat Aug 25 00:00:00 2007

The end :)