Options:
- # Session Start: Sun Aug 24 00:00:00 2008
- # Session Ident: #whatwg
- # [00:10] * Quits: Amorphous (i=jan@g227178060.adsl.alicedsl.de) (Read error: 110 (Connection timed out))
- # [00:15] * Joins: Amorphous (i=jan@f048035151.adsl.alicedsl.de)
- # [00:20] * Quits: tantek (n=tantek@dsl001-150-252.sfo1.dsl.speakeasy.net)
- # [00:20] * Quits: eseidel (n=eseidel@c-67-180-49-110.hsd1.ca.comcast.net)
- # [00:43] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [00:49] * Quits: Maurice (i=copyman@cc498817-b.emmen1.dr.home.nl) ("Disconnected...")
- # [00:51] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [01:50] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [02:08] * Quits: billmason (n=billmaso@ip75.unival.com) (".")
- # [02:14] * Quits: webben_ (n=benh@91.85.149.201)
- # [02:24] * Quits: jeremyb_ (n=jeremyb@unaffiliated/jeremyb) (Read error: 104 (Connection reset by peer))
- # [02:24] * Joins: jeremyb_ (n=jeremyb@unaffiliated/jeremyb)
- # [02:51] * Quits: csarven (n=csarven@modemcable144.140-202-24.mc.videotron.ca) (Read error: 110 (Connection timed out))
- # [02:57] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [03:03] * Quits: kangax (n=kangax@ool-182f8118.dyn.optonline.net)
- # [03:32] * Quits: tndH (i=Rob@adsl-77-86-6-71.karoo.KCOM.COM) ("ChatZilla 0.9.83-rdmsoft [XULRunner 1.9/2008061013]")
- # [03:40] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [03:49] * Joins: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [04:51] * Joins: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [05:11] * Quits: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [05:44] * Quits: jeremyb_ (n=jeremyb@unaffiliated/jeremyb) (Read error: 104 (Connection reset by peer))
- # [05:44] * Joins: jeremyb_ (n=jeremyb@unaffiliated/jeremyb)
- # [05:47] * Quits: jeremyb_ (n=jeremyb@unaffiliated/jeremyb) (Read error: 104 (Connection reset by peer))
- # [05:47] * Joins: jeremyb_ (n=jeremyb@pool-162-84-143-143.ny5030.east.verizon.net)
- # [06:35] <Hixie> wow, david is arguing not only for alt= being required but for it being required to never be empty even for decorative images
- # [06:35] <Hixie> and he's blind, as far as i can tell
- # [06:35] <Hixie> so presumably would be aware of the crazyness of that position
- # [06:35] * Hixie is baffled
- # [06:40] * Quits: codedread (n=schiller@c-24-13-43-191.hsd1.il.comcast.net) (Remote closed the connection)
- # [06:55] * Joins: eseidel_ (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [06:57] * Joins: kangax (n=kangax@ool-182f8118.dyn.optonline.net)
- # [06:58] * Quits: eseidel_ (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net) (Client Quit)
- # [07:01] * Quits: jruderman (n=jruderma@c-67-180-39-55.hsd1.ca.comcast.net)
- # [07:07] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [07:13] * Joins: csarven (n=csarven@modemcable144.140-202-24.mc.videotron.ca)
- # [07:15] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [07:35] * Joins: hdh (n=hdh@118.71.120.173)
- # [07:40] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [07:42] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net) (Client Quit)
- # [07:47] * Joins: jruderman (n=jruderma@c-67-180-39-55.hsd1.ca.comcast.net)
- # [07:50] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [08:04] * Quits: csarven (n=csarven@modemcable144.140-202-24.mc.videotron.ca) (Read error: 110 (Connection timed out))
- # [08:08] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [08:09] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [08:35] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [08:37] * Quits: kangax (n=kangax@ool-182f8118.dyn.optonline.net)
- # [08:50] * Joins: shepazu (n=schepers@88.128.86.134)
- # [08:53] * Joins: heycam (i=cam@80.187.209.39)
- # [09:07] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [09:12] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net) (Client Quit)
- # [09:28] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [09:33] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [09:54] * Joins: Maurice (i=copyman@cc498817-b.emmen1.dr.home.nl)
- # [09:56] * Joins: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [10:05] * Quits: eseidel (n=eseidel@c-76-126-222-119.hsd1.ca.comcast.net)
- # [10:09] * Joins: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se)
- # [10:56] * Joins: ROBOd (n=robod@89.122.216.38)
- # [10:57] * Quits: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se) (Remote closed the connection)
- # [11:04] * Joins: weinig (n=weinig@212.145.145.227)
- # [11:32] * Joins: virtuelv (n=virtuelv@163.80-202-65.nextgentel.com)
- # [11:40] * Joins: webben (n=benh@91.85.149.201)
- # [12:02] * Joins: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [12:02] * Quits: weinig (n=weinig@212.145.145.227)
- # [12:03] * Quits: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net) (Client Quit)
- # [12:31] * Joins: weinig (n=weinig@212.145.145.227)
- # [12:44] <Hixie> Philip`: i would be interested in hearing your opinion of http://lists.w3.org/Archives/Public/public-html/2008Aug/0254.html
- # [12:49] * Joins: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [13:12] <Hixie> http://lists.w3.org/Archives/Member/w3c-html-cg/2008JulSep/0073.html sugests that the HCG call two weeks ago consisted almost entirely about talking about me
- # [13:13] <Dashiva> Too bad it's a treehouse-archive
- # [13:15] <Philip`> I think the described "gain for one segment of the population (those needing alt text)" comes from having any mechanism that gives a textual description of an image when no equivalent is available (and that requires authors to always write something, and hence to think about alt and (with non-zero probability) write something that's better than nothing);
- # [13:15] <Philip`> and this particular "small pain for another (tool writers)" is specific to the curly-brace syntax and doesn't seem to be a problem in any other proposed mechanism, because the other mechanisms don't add any complexity to the otherwise-pleasant situations where you really do have equivalent text
- # [13:15] <Philip`> I think the gain is worthwhile, so I wouldn't want to go back to alt being optional; but I think the pain is real, so I'd prefer some other solution that has the same gain without the pain
- # [13:15] <Philip`> but that involves tradeoffs against the pains (of varying levels of theoreticality) of other solutions, which is not trivial (if it's even possible) :-(
- # [13:17] <Hixie> k
- # [13:19] <Philip`> (I guess the desire to be able to distinguish equivalent text from descriptive text, so that <img alt="equivalent"> can be rendered the same as <span>equivalent</span>, seems like a mostly theoretical concern to me, since in reality nobody distinguishes alt that way, so a UA will have to cope with the indistinguishability anyway)
- # [13:20] <Hixie> that's basically the argument which is leading me back to alt="" being optional -- whatever we do the data won't be reliable, so UAs are going to have to do heuristics in all cases anyway, so why not just have the simplest syntax
- # [13:20] <Hixie> it's not like categories/roles like the spec suggests using now are actually going to be useful
- # [13:21] <Hixie> ok bed time
- # [13:21] <Hixie> nn
- # [13:21] <Philip`> That's far more of a change than just a simpler syntax, since it deprives the UA/user of even descriptive text about the image
- # [13:21] <Hixie> in practice i doubt that we'd be able to get descriptive text about the image
- # [13:22] <Hixie> it's not like flickr knows what the images are
- # [13:22] <Hixie> and in the case of a computer-generated fractal view, it's not like saying "it's a fractal" is gonna help anyone
- # [13:22] <Hixie> same with a webcam
- # [13:23] <Hixie> or a blind user's photos
- # [13:33] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [13:34] <Philip`> Even if Flickr just said something like <img alt="Main photo">, that would help users tell apart the main photo from e.g. the "next in photostream" button where Flickr forgot to put any alt attribute, so it's better than nothing
- # [13:35] <Hixie> so you want to require that flickr be conforming for one <img> so that the usability of their site is improved because you think they won't be conforming for another one?
- # [13:36] <Dashiva> So once again, we're back at something like notalt="text that isn't alternate"
- # [13:36] <Hixie> if we are assuming we can get them conforming for any images, why would we assume the main image (which is the hardest to get right) would be where they would start in becoming conforming?
- # [13:38] <Philip`> Hixie: The evidence seems to back up that assumption, because Flickr does have alt on its main image whereas it doesn't on some of its other images
- # [13:39] <Hixie> the alt on the main image is pretty bogus right now in most cases
- # [13:40] <Hixie> i would argue that the flickr site right now shows that which images get alt="" attribute is more likely to be near-random than anything else
- # [13:40] <Philip`> Presumably the main image has been there since Flickr first existed, so there's plenty of time for someone who knows 'HTML4 says I need an alt attribute' to have edited the code to generate an alt attribute, while the other images are more recent and nobody has cared to 'fix' them yet
- # [13:40] <Philip`> "pretty bogus" is still better than nothing :-)
- # [13:40] <Hixie> maybe
- # [13:40] <Hixie> and maybe
- # [13:40] <Hixie> or maybe not
- # [13:41] <Philip`> and when our choices seem to be between "nothing" and "better than nothing", the latter seems a better choice
- # [13:41] <Hixie> if it was clearly better i'd agree
- # [13:42] <Philip`> Dashiva: Or just alt="text that isn't alternate" :-)
- # [13:43] <Hixie> we could also require that all such images be in a link pointing at the image itself
- # [13:44] <Dashiva> Philip`: I'm cool with it. You just have to convince public-html
- # [13:44] <Hixie> actually he just has to convince me, and, if it is controversial, the chairs :-)
- # [13:45] <Dashiva> No, he has to convince public-html so my inbox can get peace and quiet at last ;)
- # [13:45] <Hixie> i think it's pretty obvious that there are a number of people who are unwilling to move from their positions and who have positions that are unlikely to ever gain wide acceptance
- # [13:46] <Hixie> e.g. the people who think that even decorative images should have non-empty alt
- # [13:48] <Hixie> Philip`: so what text would you suggest for the following cases? a generated image for a fractal explorer; the images in google maps in map view and satellite view; the street view images in google maps; an image uploaded as part of a batch upload with no information; a webcam
- # [13:50] <Hixie> ok going to bed for real now. will read responses tomorrow.
- # [13:50] <Hixie> nn
- # [13:51] <Philip`> Dashiva: There's an easier solution to the inbox problem: create a new mailing list for alt discussion
- # [13:51] <Philip`> Even better, create a Task Force for it
- # [13:51] <Dashiva> Philip`: Oh boy
- # [13:51] <Dashiva> Hixie: 04:45 on a sunday? That's dedication :)
- # [13:52] * Joins: tndH (i=Rob@adsl-77-86-6-71.karoo.KCOM.COM)
- # [14:10] <krijnh> http://fronteers.nl/congres/2008/speakers#anne-van-kesteren
- # [14:34] * Quits: othermaciej (n=mjs@c-69-181-42-194.hsd1.ca.comcast.net)
- # [14:38] * Quits: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [15:06] * Joins: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [15:07] * Quits: starjive (i=beos@213-66-217-32-no30.tbcn.telia.com)
- # [15:08] * Joins: shepazutoo (n=schepers@88.128.86.134)
- # [15:09] * Quits: shepazu (n=schepers@88.128.86.134) (Read error: 104 (Connection reset by peer))
- # [15:18] * Quits: weinig (n=weinig@212.145.145.227)
- # [15:26] * Joins: javaskript (n=chatzill@c182222.adsl.hansenet.de)
- # [15:27] * javaskript is now known as cramforce
- # [15:30] * Quits: cramforce (n=chatzill@c182222.adsl.hansenet.de) ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070206]")
- # [15:30] * Joins: cramforce (n=chatzill@c182222.adsl.hansenet.de)
- # [15:32] * cramforce is now known as malde
- # [15:32] * malde is now known as malde_
- # [15:34] <malde_> Hi all, is there a recommended way to use the database api to retrieve the field names of a table without selecting any data?
- # [15:36] <Dashiva> describe?
- # [15:37] <Dashiva> Maybe that's a mysql thing
- # [15:37] <malde_> describe might work. I'll take a look at the result set.
- # [15:39] <malde_> With the Gears API I can do SELECT * FROM tableName WHERE 0 = 1
- # [15:42] <Philip`> SQLite doesn't have "describe" or anything equivalent, I believe
- # [15:43] <Philip`> but it has a table named "sqlite_master" that contains the database schema, apparently
- # [15:49] <Dashiva> SQL-92 seems to require a COLUMNS table
- # [15:50] <malde_> Is sqlite_master a table in every database?
- # [15:54] <malde_> Ah, seems like it is
- # [16:06] <Philip`> Dashiva: But that's just a standard, you can't expect it to work :-p
- # [16:07] <gsnedders> malde_: yes
- # [16:07] <gsnedders> Dashiva: What actually implements any SQL standard?
- # [16:08] <Philip`> The SQL standard has things like column types, which is a feature SQLite seems quite happy to ignore
- # [16:08] <Dashiva> gsnedders: Most of them implement parts of it
- # [16:08] <gsnedders> "parts" :)
- # [16:08] <Philip`> so I'd be surprised if it implemented all the other features exactly right :-)
- # [16:14] * Quits: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [16:29] * Joins: nullpoet (n=nullpoet@aazi65.neoplus.adsl.tpnet.pl)
- # [16:31] * Joins: myakura (n=myakura@p3216-ipbf5106marunouchi.tokyo.ocn.ne.jp)
- # [16:33] * Quits: myakura (n=myakura@p3216-ipbf5106marunouchi.tokyo.ocn.ne.jp) (Client Quit)
- # [16:39] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [16:43] * Joins: weinig (n=weinig@212.145.145.227)
- # [16:45] * gsnedders looks up that Varsity survey with medicine at the top and maths at the bottom
- # [16:47] <Philip`> http://www.varsity.co.uk/archive/669.pdf ?
- # [16:51] <Philip`> (Correlation is not causation, so it would be unwise to base your application decisions on such a survey :-p )
- # [16:51] <gsnedders> Philip`: Yeah, exactly :P
- # [16:51] * Philip` wonders how much correlation there is between correlation and causation
- # [16:51] <gsnedders> Philip`: So, I'll apply for Clare because it has the highest recreational drug use
- # [16:52] <gsnedders> Hmm, interesting: a greater number of people identify as bi than gay.
- # [16:52] * gsnedders notes that is unusual
- # [16:52] * gsnedders has now dragged this channel totally off-topic :D
- # [16:53] <Philip`> It's fortunate that nobody ever lies in online surveys
- # [16:53] <gsnedders> Of couse.
- # [16:53] <gsnedders> *course
- # [16:53] * Quits: Lachy (n=Lachlan@85.196.122.246) ("This computer has gone to sleep")
- # [17:01] * Joins: sverrej (n=sverrej@89.10.27.245)
- # [17:07] * Quits: weinig (n=weinig@212.145.145.227) (Read error: 104 (Connection reset by peer))
- # [17:07] * Joins: weinig (n=weinig@212.145.145.227)
- # [17:11] * Quits: sverrej (n=sverrej@89.10.27.245) (Read error: 104 (Connection reset by peer))
- # [17:12] * Joins: sverrej (n=sverrej@89.10.27.245)
- # [17:15] * Quits: nullpoet (n=nullpoet@aazi65.neoplus.adsl.tpnet.pl) (Read error: 110 (Connection timed out))
- # [17:21] * Quits: malde_ (n=chatzill@c182222.adsl.hansenet.de) (Remote closed the connection)
- # [17:22] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [17:40] * Quits: weinig (n=weinig@212.145.145.227)
- # [17:42] * Joins: kangax (n=kangax@ool-182f8118.dyn.optonline.net)
- # [18:00] * Joins: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se)
- # [18:00] <zcorpan> Hixie: "A corollary to this is that the alt attribute's value should never contain text that could be considered the image's caption, title, or legend." -- is that intended to be a SHOULD?
- # [18:10] * Joins: weinig (n=weinig@212.145.145.227)
- # [18:25] * Quits: kangax (n=kangax@ool-182f8118.dyn.optonline.net)
- # [18:33] * Joins: PHPechowiec (n=Adrian@aabq172.neoplus.adsl.tpnet.pl)
- # [18:33] <PHPechowiec> hi
- # [18:33] <PHPechowiec> <meta http-equiv=Creation-Date content= "Sun, 24 Aug 2008 10:27:54 GMT"> ?
- # [18:34] <PHPechowiec> can i use this line in html5?
- # [18:41] * Quits: PHPechowiec (n=Adrian@aabq172.neoplus.adsl.tpnet.pl) ("Lost terminal")
- # [18:41] <webben> PHPechowiec: based on http://www.w3.org/html/wg/html5/#pragma , I do not believe that would conform.
- # [18:41] <gsnedders> It wouldn't.
- # [18:46] * Joins: sverrej__ (n=sverrej@89.10.27.245)
- # [18:53] * Joins: Lachy (n=Lachlan@85.196.122.246)
- # [18:56] * Quits: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se) (Remote closed the connection)
- # [18:57] * Quits: sverrej_ (n=sverrej@89.10.27.245) (Read error: 113 (No route to host))
- # [18:57] * Quits: sverrej (n=sverrej@89.10.27.245) (Read error: 110 (Connection timed out))
- # [18:58] * Joins: sverrej (n=sverrej@89.10.27.245)
- # [19:02] <gsnedders> Dashiva: Are you one of the people under your name on Facebook?
- # [19:02] <Dashiva> No
- # [19:02] <gsnedders> Dashiva: Are you someone on Facebook not under your name? :D
- # [19:02] <Dashiva> No
- # [19:03] <gsnedders> (Yes, I am bored)
- # [19:03] <gsnedders> (Yes, I am stalking you all)
- # [19:04] <Dashiva> I was on orkut for a few weeks before realizing it was a mistake
- # [19:04] <Dashiva> And I'm on myopera out of necessity
- # [19:04] <gsnedders> Why is it needed?
- # [19:05] <Dashiva> I did work for opera in a user-facing role :)
- # [19:05] <gsnedders> user-facing?
- # [19:05] <Dashiva> Widgets
- # [19:05] <gsnedders> No wonder I only ever saw your back.
- # [19:06] <gsnedders> (I know that is only a very vaguely witty remark, but it'll suffice)
- # [19:06] <Dashiva> Maybe you should get glasses
- # [19:07] <gsnedders> I have some in my pocket, and I'm getting new ones next week.
- # [19:08] <Dashiva> You can find me on mixi.jp though :)
- # [19:11] * Dashiva notices another friend invite on myopera
- # [19:11] <Dashiva> I don't get these people who want several thousand friends they never even talk to
- # [19:14] * Quits: jeremyb_ (n=jeremyb@unaffiliated/jeremyb) (Read error: 104 (Connection reset by peer))
- # [19:17] * jgraham doesn't understand gsnedders joke
- # [19:17] <jgraham> Presumably if Dashiva was in a user facing role and you only ever saw his back that would mean that you were inside Opera
- # [19:18] <gsnedders> Or I'm just not a user
- # [19:21] <jgraham> Oh I see.
- # [19:21] <Dashiva> gsnedders isn't a user, he's an addict. *rimshot*
- # [19:21] <jgraham> I don't think that's true though. Presumably you jut wouldn't see him at all
- # [19:22] <gsnedders> Well, I'm around enough to be reverse-engineering Opera, even if I'm not really using it
- # [19:22] <jgraham> Hmm John seems to have missed my point somewhat
- # [19:22] <Dashiva> Why am I not surprised?
- # [19:22] <gsnedders> http://media.tumblr.com/1vnWCPZWQ9fd1s1k1zTvXUJM_500.png
- # [19:23] <Dashiva> 4/10
- # [19:23] <Dashiva> gsnedders: You forgot to supply alternate text
- # [19:26] <gsnedders> alt="A line, labelling you, heading in a curve away from the marked point"
- # [19:27] <Dashiva> So the initial heading being towards the point is not significant?
- # [19:27] * gsnedders shrugs
- # [19:32] * Quits: weinig (n=weinig@212.145.145.227)
- # [19:34] <Dashiva> "The issue boils down to this: should conformance be worth paying any attention to without the law requiring it? We argue no."
- # [19:35] * Joins: hasather (n=hasather@90-231-107-133-no62.tbcn.telia.com)
- # [19:58] * webben thinks that this is a bit of a storm in a teacup giving the lack of general interest in conformance, and authors' happiness to ignore conformance problems they judge irrelevant to them.
- # [19:58] <webben> *given
- # [19:59] <webben> i'd be much more worried about trying to explain weirdnesses of HTML to people trying to validate their content.
- # [19:59] <webben> at least "provide a text equivalent" can be explained in human language
- # [20:00] <webben> unlike say entity errors in current HTML validation, which needs to be explained in geekspeak :)
- # [20:02] * Joins: csarven (n=csarven@modemcable144.140-202-24.mc.videotron.ca)
- # [20:02] * webben wonders if HTML5 make & in URLs in href conforming or still requires conversion to &
- # [20:02] <webben> newbies ever understands that.
- # [20:02] <webben> *never understand
- # [20:03] * Joins: malde_ (n=chatzill@d123124.adsl.hansenet.de)
- # [20:10] <Philip`> webben: That's still non-conforming, because it's likely to result in errors (e.g. when someone wants to add a "gt" parameter to their query and writes <a href="foo.cgi?a=b>=c"> and it all goes horribly wrong), and so it's still worthwhile repeatedly telling authors not to do that
- # [20:13] <Philip`> Hopefully the conformance checker would detect when you have that error in a URL attribute, and would say something relatively friendly like "It looks like you are using '&' to separate values in this URL - you have to write them as '&' instead", rather than saying something confusing and meaningless like "Error: Entity reference was not terminated by a semicolon."
- # [20:13] <webben> Hopefully.
- # [20:14] <Dashiva> Philip`: Only if it outputs the message as RDF
- # [20:16] <webben> apparently, however, http://www.flickr.com/ don't bother encoding to & even now.
- # [20:17] <Dashiva> webben: Considering all the hate they get on public-html, maybe they decided conformance wasn't worth trying for :)
- # [20:17] <webben> which I suppose points at the problem of requiring things that only help sometimes.
- # [20:17] <webben> Dashiva: Flickr's non-conformance is long-standing; I'm not sure the public-html discussions about Flickr have very much visibility.
- # [20:19] <webben> I see people regularly use & instead of & because they think it will work, even when they care about other aspects of valid HTML.
- # [20:26] * Quits: tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net)
- # [20:33] * Joins: starjive (i=beos@213-66-217-32-no30.tbcn.telia.com)
- # [20:34] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [20:35] * Quits: malde_ (n=chatzill@d123124.adsl.hansenet.de) (Remote closed the connection)
- # [20:39] * Parts: hasather (n=hasather@90-231-107-133-no62.tbcn.telia.com)
- # [20:40] * Joins: hasather (n=hasather@90-231-107-133-no62.tbcn.telia.com)
- # [20:47] * Joins: malde_ (n=chatzill@d123124.adsl.hansenet.de)
- # [20:50] * Joins: othermaciej (n=mjs@c-69-181-42-194.hsd1.ca.comcast.net)
- # [20:51] * Quits: othermaciej (n=mjs@c-69-181-42-194.hsd1.ca.comcast.net) (Read error: 104 (Connection reset by peer))
- # [21:23] * Quits: malde_ (n=chatzill@d123124.adsl.hansenet.de) (Read error: 110 (Connection timed out))
- # [21:26] * Joins: hasather_ (n=hasather@90-231-107-133-no62.tbcn.telia.com)
- # [21:27] * Parts: hasather_ (n=hasather@90-231-107-133-no62.tbcn.telia.com)
- # [21:42] * Joins: sverrej_ (n=sverrej@89.10.27.245)
- # [21:53] * Joins: othermaciej (n=mjs@c-69-181-42-194.hsd1.ca.comcast.net)
- # [21:55] * Quits: sverrej (n=sverrej@89.10.27.245) (Read error: 110 (Connection timed out))
- # [21:57] * Quits: sverrej__ (n=sverrej@89.10.27.245) (Read error: 113 (No route to host))
- # [21:58] * Joins: sverrej (n=sverrej@89.10.27.245)
- # [22:07] * Quits: gsnedders (n=gsnedder@host217-44-35-200.range217-44.btcentralplus.com) ("Killin' teh intarwebs")
- # [22:08] * Joins: annevk (n=annevk@77.163.243.203)
- # [22:13] * Quits: sverrej_ (n=sverrej@89.10.27.245) (Read error: 110 (Connection timed out))
- # [22:14] * Joins: sverrej_ (n=sverrej@89.10.27.245)
- # [22:17] * Quits: sverrej (n=sverrej@89.10.27.245) (Read error: 110 (Connection timed out))
- # [22:18] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [22:18] * Joins: sverrej (n=sverrej@89.10.27.245)
- # [22:18] <jgraham> Sigh
- # [22:19] * annevk notices people have been having fun over the weekend
- # [22:19] * annevk played Super Mario World and Call of Duty 4
- # [22:20] <jgraham> heh
- # [22:21] <annevk> and I used my friends high upload bandwidth to get my Cambridge photos on flickr
- # [22:21] <Dashiva> annevk: I'm sure you could make a game out of public-html. They managed to come up with buzzword bingo, after all :)
- # [22:22] * jgraham wonders if the a11y people only play text based adventures
- # [22:24] <Dashiva> jgraham: Making textual alternates available would make most games very prone to macroing/botting/etc
- # [22:24] * Joins: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se)
- # [22:24] <jgraham> annevk: Did you have fun punting in the rain?
- # [22:24] * Quits: hdh (n=hdh@118.71.120.173) (Remote closed the connection)
- # [22:25] * Quits: ROBOd (n=robod@89.122.216.38) ("http://www.robodesign.ro")
- # [22:25] <zcorpan> hsivonen: fwiw, i now didn't get that error message when filing a bug
- # [22:26] <Philip`> I like how some games (at least Portal, and probably the Half Life 2 games too) have closed captioning, so you can get all the audio cues even without any audio
- # [22:27] <hsivonen> webben: There's indeed tension between saying "this stuff sucks so you shouldn't use it for new content" and "this stuff sucks so badly that you should go and remove if from the content you are upgrading"
- # [22:27] <hsivonen> webben: I think we don't have a good solution that respects both the time of upgraders and of authors starting from a clean slate
- # [22:28] <webben> well, one could actually differentiate between the two
- # [22:28] <annevk> btw, if you like Mario, http://www.youtube.com/watch?v=in6RZzdGki8 is quite funny
- # [22:28] <webben> whether it's good to do so or not I don't know.
- # [22:28] <annevk> jgraham, there wasn't too much rain actually
- # [22:29] <webben> hsivonen: for example one could error in response to >= but only warn for &name=
- # [22:29] <annevk> jgraham, it was fun to do
- # [22:29] <jgraham> Ah, I saw Hakon had his umbrella out
- # [22:30] <annevk> it's from Bert iirc
- # [22:30] <annevk> but the drops were hardly noticable for most of it
- # [22:36] <jgraham> Sigh**2
- # [22:38] <annevk> maybe you should set up a proxy that filters out WHATWG and HTML5 related e-mail until your thesis is done :)
- # [22:38] <annevk> though I guess it gives some distraction
- # [22:41] * Joins: gsnedders (n=gsnedder@host217-44-35-200.range217-44.btcentralplus.com)
- # [22:42] <hsivonen> webben: the problem is how to distinguish them in a way that doesn't fail like HTML 4.01 Strict vs. Transitional?
- # [22:42] <hsivonen> webben: it seems a bit pointless to make the validator emit messages at all where the messages are meant to be ignored
- # [22:42] <webben> hsivonen: warnings aren't "meant to be ignored".
- # [22:43] <jgraham> annevk: Not such a bad idea I guess. But I think my thesis will get done. It's just annoying to try and explain something clearly and then realise that you didn't get your message across at all
- # [22:43] <hsivonen> webben: I'm getting into the game of analyzing millions of pages in order to get data about what kind of errors are the ones that would particularly waste the time of upgraders
- # [22:44] <webben> hsivonen: I think the "failure" (if there was one) had to do with choices about what went into strict that disagreed with what authors wanted to use and that worked in the browsers they cared about.
- # [22:44] * Quits: starjive (i=beos@213-66-217-32-no30.tbcn.telia.com)
- # [22:44] <annevk> jgraham, maybe you should try to enforce your points by law
- # [22:45] <webben> hsivonen: I suppose target and iframe being the most important examples.
- # [22:45] <annevk> jgraham, they'll probably be familiar with that
- # [22:45] * annevk hides
- # [22:45] <hsivonen> webben: as far as HTML tables and IE installed base are concerned, we have that kind of problem with HTML5
- # [22:45] <Dashiva> webben: Didn't stop target from being a point of contention in html5 as well :)
- # [22:46] <webben> Dashiva: Well of course not. Authors wanting to use doesn't make it a great idea.
- # [22:46] <Dashiva> Whether it's a good idea or not doesn't mean much if it leads to scrapping your strict version by authors
- # [22:47] <webben> depends on what the point of the versioning is
- # [22:47] <webben> if the point of conformance is to get "good practice" but you can only get conformance by sanctioning "bad practice" that's a largely pointless exercise.
- # [22:48] <hsivonen> fwiw, I'm not so much concerned about pushing an ideal of good practice
- # [22:48] <Dashiva> webben: Not if allowing one bad practice means they fix a dozen others
- # [22:48] <hsivonen> I'm more interested in producing a tool that allows authors catch mistakes they didn't make
- # [22:49] <hsivonen> although Validator.nu does have political good practice aspects
- # [22:49] <webben> what are "mistakes they didn't make"?
- # [22:49] <hsivonen> oops
- # [22:49] <hsivonen> didn't *intend* to make
- # [22:50] <webben> how can you intend to make a mistake?
- # [22:50] <Dashiva> E.g. if you're using a bad practice on purpose
- # [22:50] <webben> is that good or bad? using a "bad practice" on purpose.
- # [22:50] <hsivonen> webben: if you use <embed> in HTML 4.01 on purpose, for example
- # [22:51] <webben> that's not a "mistake". that's a conformance error.
- # [22:51] <webben> or an instance of deliberate non-conformance
- # [22:52] <webben> validators can't really do much about instances of deliberate non-conformance
- # [22:52] * annevk wonders what the point of this conversation is
- # [22:52] <webben> the point for me, is increasing puzzlement about what conformance in HTML5 is, and its integrity as a single thing.
- # [22:53] <hsivonen> annevk: I guess the point on my side is that I don't believe as much in the wishful part of conformance as I think Hixie does
- # [22:53] <hsivonen> annevk: when it comes to removing "bad" legacy features from the conforming language
- # [22:54] <webben> it seems to me the one bit of conformance everyone (?) would buy into is where you recommend features that work in lots of user agents over features that do the same thing but in fewer user agents.
- # [22:56] <webben> Hixie seems keen to use conformance to protect future extension points, which is a good idea, but of less widespread appeal.
- # [22:56] * Quits: svl (n=me@ip565744a7.direct-adsl.nl) ("And back he spurred like a madman, shrieking a curse to the sky.")
- # [22:56] <annevk> works for CSS
- # [22:57] <hsivonen> webben: that's not what making align on td non-conforming is about, though
- # [22:57] <webben> annevk: CSS has extension points built into the language.
- # [22:57] * Quits: zcorpan (n=zcorpan@c-cb21e353.1451-1-64736c12.cust.bredbandsbolaget.se) (Remote closed the connection)
- # [22:57] <webben> annevk: I don't think CSS validation/conformance has been a particularly powerful force for authors.
- # [22:57] <webben> e.g. non-validating hacks, opacity, filter, vendor-specific extensions
- # [22:58] <webben> with the last one built-in, but forbidden
- # [22:58] <hsivonen> I stopped paying attention to CSS validation when I couldn't configure the validator to grok Prince extensions
- # [22:58] <annevk> the validator should probably group extensions as a single error
- # [22:59] <hsivonen> annevk: this was the CSS validator running inside Oxygen
- # [22:59] * Joins: malde_ (n=chatzill@d123124.adsl.hansenet.de)
- # [22:59] * Quits: malde_ (n=chatzill@d123124.adsl.hansenet.de) (Client Quit)
- # [22:59] <annevk> my impression has always been that quite a few authors do use the CSS validator, but that's anecdotical
- # [23:00] <webben> Then there's conformance for trying to change authoring practices that could lead to errors, even when they in fact aren't errors that cause problems. e.g. &name in href.
- # [23:00] <hsivonen> or spaces in URLs most of the time
- # [23:00] <webben> While I'm glad of a tool to check that, I suspect that will reduce the amount of conformance considerably.
- # [23:01] <webben> e.g. lots of relatively conforming webpages fail to encode ampersands now
- # [23:01] <annevk> &name could cause a problem going forward
- # [23:01] <annevk> though that is somewhat avoided by requiring ; for new entities
- # [23:01] <webben> annevk: yeah well I'm talking about &name= not &name;
- # [23:02] <webben> and if one didn't require ; that would break _old_ content using &name=
- # [23:02] <webben> of which there's plenty
- # [23:02] <annevk> &name is slightly suboptimal from a parser point of view
- # [23:02] <annevk> so an author might want to optimize that
- # [23:03] <webben> yeah. they _might_.
- # [23:03] <annevk> anyway, if you want &name to be conforming, you want whatwg@whatwg.org
- # [23:04] * Joins: svl (n=me@ip565744a7.direct-adsl.nl)
- # [23:04] <webben> personally, I don't care whether it's conforming or not particularly (I just use & ). I just think it's interesting as an aspect of the emerging picture of HTML5 conformance.
- # [23:05] <annevk> ok
- # [23:07] <webben> my guess (it's only a guess) is that there will be more websites failing conformance because of ampersand rules than would if HTML5 required alt (for example), since I see lots of websites with alt but that fail to encode ampersands
- # [23:07] <webben> Flickr probably won't conform whatever the conformance rules are (alt isn't a "problem" for Flickr now, it just uses alt="" on the main photo page; it doesn't encode ampersands however).
- # [23:08] <webben> given those speculations, I find much of the alt debate rather odd
- # [23:08] <annevk> hsivonen has stats on that
- # [23:09] <annevk> I find it rather odd for very different reasons :D
- # [23:09] <webben> oh there are other odd things said by most correspondents
- # [23:12] <hsivonen> annevk: what do I have stats on?
- # [23:12] * Joins: BenMillard (n=cerbera@cpc1-flee1-0-0-cust285.glfd.cable.ntl.com)
- # [23:13] <webben> hsivonen: websites with/without alt, with/without ampersands encoded properly, combinations thereof, I think?
- # [23:13] <annevk> common conformance errors on pages
- # [23:13] <webben> oh, pages is somewhat different
- # [23:13] <webben> e.g. photo sharing sites have lots of pages, few links.
- # [23:13] <hsivonen> when I ran the alexa 500 through the validator, absent alt wasn't a machine-checkable conformance error
- # [23:14] <webben> note I said "websites failing conformance"
- # [23:14] <annevk> nm then hsivonen, seems I missed his point
- # [23:15] <hsivonen> hmm. looks like my mass validation harness is broken or the disk image containing the test data has a corrupt file system...
- # [23:15] <hsivonen> or both...
- # [23:17] <annevk> http://serialseb.blogspot.com/2008/08/proposing-syntax-to-attach-behaviors-to.html#c8752145182107351356 ?
- # [23:17] <annevk> zcorpen ? ^^
- # [23:18] <annevk> http://ajaxworld.com/general/sessiondetail1008.htm?id=57
- # [23:19] <annevk> people from Kaazing Corporation will speak about HTML5...
- # [23:28] * Joins: roc (n=roc@202.0.36.64)
- # [23:47] * Quits: heycam (i=cam@80.187.209.39) ("bye")
- # [23:50] * Quits: Maurice (i=copyman@cc498817-b.emmen1.dr.home.nl) ("Disconnected...")
- # [23:51] * Quits: shepazutoo (n=schepers@88.128.86.134)
- # [23:53] * Joins: eseidel (n=eseidel@c-24-130-13-197.hsd1.ca.comcast.net)
- # [23:54] * Joins: jcranmer_ (n=jcranmer@ltsp1.csl.tjhsst.edu)
- # [23:57] * Quits: jcranmer (n=jcranmer@ltsp1.csl.tjhsst.edu) (Nick collision from services.)
- # [23:57] * jcranmer_ is now known as jcranmer
- # [23:59] <jcranmer> I prefer i.m.o's GHOST quit message
- # Session Close: Mon Aug 25 00:00:00 2008
The end :)