00:02
<Hixie>
jruderman: fixed the routing problem
00:03
<jruderman>
Hixie: which problem?
00:03
<jruderman>
Hixie: oh, the google maps thing with the mozilla office
00:05
<jruderman>
Hixie: now it directs people to building S rather than building K, but that's probably a lot harder to get right
00:06
<jruderman>
Hixie: thanks, now we'll lose fewer moco employees to pedestrian fatalities on the 101 freeway
00:07
<Philip`>
Why does Google Maps choose a certain route when there's an alternative that's two minutes shorter, and that it's happy to pick when I drag the route around a bit?
00:07
Philip`
isn't sure why it'd be optimising for anything except distance
00:09
<gavin>
it takes into account speed limits and such, doesn't it?
00:09
<gavin>
favors highways/major streets too, I think
00:09
<gavin>
I'm sure it's very complex
00:09
<Philip`>
Uh, speed limits for pedestrians?
00:09
<gavin>
oh, well
00:09
<gavin>
that's in beta!
00:09
<Philip`>
(This is the walking mode)
00:09
<Hixie>
jruderman: well the pin can be edited by users (and has been. a lot. maybe other people in your office are moving it to the wrong location. :-) )
00:10
<jruderman>
Hixie: hehe
00:10
<codedread>
funny - i just noticed this pin editing thing 5 minutes ago
00:10
<Hixie>
jruderman: but the problem was that it ended on 101, and that has been fixed for the address you gave :-)
00:10
<Hixie>
Philip`: simplicity is a factor
00:10
<jruderman>
Hixie: did you fix it letting people put building pins on freeways, or did you just move that one pin?
00:11
<Hixie>
jruderman: i can't talk about what the actual fix is, because it's a two phase fix and the second phase hasn't shipped yet.
00:11
<Hixie>
ask me again in a few weeks
00:11
<jruderman>
you guys take the 'future products or services' way too seriously
00:11
<Hixie>
:-)
00:12
gsnedders
passes jruderman a future product or service
00:12
<Philip`>
Hixie: I suppose it could be, but the shorter route is only two extra steps, and I'm too lazy to travel an extra two minutes just for that
01:08
<Hixie>
Philip`: i'm not defending it, just saying what the reason is
01:10
<roc_>
I don't think the 101 problem was the pin location
01:18
<Hixie>
roc_: why not? (as far as i can tell, the pin location caused maps to decide the nearest street was 101.)
01:19
<roc_>
because even if the nearest street *is* 101 (and presumably there are valid locations for which that's true), it should know you can't stop there.
01:20
<roc_>
I'm not ripping on Maps, it's great and the problem is hard. Besides, I have friends who work on it
01:20
<Hixie>
oh i agree that it could try to avoid stopping on 101 in general
01:21
<roc_>
I only blogged about it because I thought it was hilarious. The Street View was particularly good :-)
01:22
Hixie
didn't know you'd blogged about it
01:22
<roc_>
ages ago
01:23
<Hixie>
ah
01:24
<jruderman>
i also tried to report it as a bug using email a long time ago, but telling hixie about it yesterday got it fixed quickly :)
02:14
<roc_>
http://developer.mozilla.org/web-tech/2008/09/04/web-workers-part-1/
02:16
<Hixie>
i love how all the examples of workers are always really really inefficent implementations of solved problems :-)
02:17
<Hixie>
do any of the demos on whatwg.org/demos/workers work?
02:17
<roc_>
to be fair, compilers like to use fibonacci too
02:18
<Hixie>
oh i'm to blame as well
02:18
<roc_>
I wish for a compiler that can optimize fibonacci to the closed form
02:18
<Hixie>
one of my examples uses 11 workers to return a constant
02:18
<roc_>
I don't really know much about the workers work myself, so I couldn't answer your question other than by testing them
02:18
<Hixie>
k
02:18
<Hixie>
i'll grab a build once i'm back on my laptop
02:19
<Hixie>
(my linux box can't run ff3, doesn't have some required library)
02:19
<Hixie>
libgtk-x11 or something
02:19
<Philip`>
roc_: It ought to be easy for the compiler just to do pattern-matching to detect a Fibonacci function and replace it with an efficient library call
02:19
<roc_>
Philip`: yeah, but that's cheating
02:19
<Philip`>
(like what MSVC does with memcpy loops)
02:19
<Philip`>
(but pointlesser)
02:20
<roc_>
recognizing and solving general difference equations is possible, although probably equally useless in practice
02:21
<Philip`>
I guess the compiler would have to work out it needs a "if (n < 0) { for(;;); }" to get the right computation
02:21
<Lachy>
I finally got my public-html email backlog down to zero! -)
02:22
<Lachy>
now I can start on the 485 in the whatwg backlog :-/
02:45
<BenMillard>
Hixie, there's a similar problem with ARIA examples for static web content: existing semantics in HTML4 are often adequate, or the ARIA becomes unnecessary when the feature is changed to follow a usable design pattern.
02:45
<BenMillard>
(Re: your comment about workers examples)
03:24
<BenMillard>
Hixie, I sent a follow-up to my review of "@headers issue resolved": http://lists.w3.org/Archives/Public/public-html/2008Sep/0163.html
03:25
<BenMillard>
the blog entry it links to has numerical analysis of the tables cited as supporting evidence versus corrected and redesigned versions I've done
03:29
<BenMillard>
time for lunch :P
07:21
<hsivonen>
Hixie: I guess in the case of Java the problem scenario is that someone has serialization code buried at the end of a pipeline somewhere
07:22
<hsivonen>
Hixie: and then they start sending HTML5 stuff through the pipeline
07:22
<hsivonen>
Hixie: and the output is wrong
07:22
<hsivonen>
Hixie: and ripping out 4 or 5 lines of code and replacing them with 2 lines of code referring to the serializer I wrote is onerous
07:23
<hsivonen>
as they'd need to discover and add Yet Another Jar
07:23
<hsivonen>
and perhaps convince their lawyers that code under the MIT license and copyright Mozilla Foundation is OK to use
07:24
<hsivonen>
I get an impression that the HTML5 content would end up in the pipeline in a semi-uncoordinated way
07:25
<hsivonen>
because if it were coordinated, one would think the serializer swapping could be coordinated, too
09:50
zcorpan
discovers document.implementation.createHTMLDocument
09:58
<annevk>
Hixie, http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2006-November/007719.html was my e-mail
10:00
<annevk>
Hixie, it's in the WF3 folder apparently o_O
10:18
<Lachy>
LOL! I like how John Foliot claims my proposed study of longdesc is flawed based on the fact that it will reveal the exact problems I designed to reveal. :-)
10:19
<Dashiva>
Seems quite in line with most commercial studies :)
10:25
<hsivonen>
one can read quite a bit of browser history in the UA string of Chrome
10:26
<hsivonen>
I wonder if we ever come up with a way to stop the growth of the UA string
10:27
<hsivonen>
does a typical HTTP GET request still fit on one IP packet?
10:27
<Dashiva>
Yeah, one day we'll have RealUserAgent instead
10:27
<Dashiva>
Since everything in the original UA string will be back-compat and talismans
10:31
<hsivonen>
is Netscape 6 the only browser in history that has swam against the UA string cruft stream?
10:31
<hsivonen>
Opera did recently, but it can still vary it per site, right?
10:31
<wilhelm>
No. Mine says "Opera/9.52 (X11; Linux i686; U; en)".
10:31
<wilhelm>
Yes. It identifies as other browsers on some sites.
10:32
<wilhelm>
We maintain a list.
10:32
<Dashiva>
Why on earth does it list X11...
10:32
<hsivonen>
Dashiva: well, it say U and en, too...
10:33
<Dashiva>
Yeah, but my windows version says that too
10:33
<hsivonen>
Chrome/1.0 now that would have been revolutionary :-)
10:33
<Philip`>
Does any browser ever *not* say "U"?
10:33
<Dashiva>
platform, U, language is always there
10:33
Philip`
doesn't know if they can be configured with rubbish encryption instead of U
10:34
<Philip`>
IE8b2 now sends a "UA-CPU: x86" header
10:34
<webben>
what?
10:34
<webben>
what for?
10:35
<roc_>
so that Web pages can send the correct x86 binaries for execution
10:35
<hsivonen>
webben: ActiveX install sites perhaps?
10:35
<roc_>
process separation rules
10:35
<webben>
hsivonen: http://blogs.msdn.com/ie/archive/2006/10/17/accept-language-header-for-internet-explorer-7.aspx#839384 seems so.
10:37
<Philip`>
Oh, IE7 sends UA-CPU too?
10:40
hsivonen
wonders if people who write about "token security" intended the pun...
10:44
<annevk>
so should I do my fragment tests again with ( or so as character instead of ?
10:46
<Philip`>
That should avoid the URL-parsing issues in (at least) Konqueror
10:49
<annevk>
does x' help with that too?
10:49
<annevk>
' seems pretty harmless, and suffix it with x just in case
10:50
<annevk>
prefix, I mean, though i'll suffix it as well
10:53
<annevk>
so far same results in browsers
10:53
<annevk>
I think I'll leave it be
11:44
<hsivonen>
How many Windows ports of WebKit are there with different graphics layers? Safari, Arora, Chrome, AIR. Did the company that ported WebKit to Windows Mobile have a 5th graphics layer?
11:44
<hsivonen>
does AIR use the same Abode vector graphics renderer as InDesign and Acrobat?
11:45
<hsivonen>
does any of them actually delegate Bézier rasterization to GDI+ or WPF?
11:53
<roc_>
I don't know what AIR does
11:53
<roc_>
the others definitely don't
11:53
<roc_>
I'm pretty sure AIR doesn't either
11:53
<zcorpan>
should document.characterSet and document.charset be part of dom core instead of dom html?
11:53
<roc_>
GDI+ is crappy, WPF would be very hard to access
11:59
<hsivonen>
roc_: presumably the WPF renderer is C or C++. Don't they expose a C API for it?
11:59
<roc_>
they don't
11:59
<hsivonen>
roc_: that's odd.
11:59
<roc_>
no it's not
11:59
<hsivonen>
roc_: ok. strategically it may not be odd.
12:00
<roc_>
also, exposing APIs is costly
12:00
<roc_>
because then you have to support them to the end of time
12:00
<hsivonen>
yeah
12:00
hsivonen
just published an API
12:00
<roc_>
sucka
12:03
<annevk>
zcorpan, if they depend on charset sniffing might be easier in HTML
12:04
<annevk>
zcorpan, also, DOM Core already has various encoding features (all completely pointless of course, the DOM shouldn't deal with encoding)
12:04
<hsivonen>
zcorpan: does query string treatment depend on charset in XML?
12:05
<annevk>
eg, inputEncoding and xmlEncoding
12:08
<hsivonen>
which reminds me that I should extend Jing to pass the document's original encoding around totally in violation of sane layering
12:09
<zcorpan>
hsivonen: don't know
12:10
<zcorpan>
annevk: should i drop xmlEncoding and inputEncoding? only firefox supports those
12:11
<annevk>
oh really? ugh
12:11
<annevk>
XMLHttpRequest references one of them :/
12:11
<annevk>
(that feature is not actually tested)
12:11
<zcorpan>
which one?
12:11
<hsivonen>
didn't DOM 3 specifiers look at what browsers already did?
12:12
<zcorpan>
hsivonen: evidently not
12:12
<annevk>
zcorpan, serializing Document for sending
12:12
<annevk>
hsivonen, nono, not at all
12:12
<hsivonen>
is the browser DOM exposed as a Java DOM anywhere?
12:13
<hsivonen>
is the Gecko DOM still available to applets or something?
12:13
<zcorpan>
annevk: xmlEncoding or inputEncoding?
12:13
<annevk>
input
12:13
<zcorpan>
ok
12:13
<zcorpan>
i'll drop xml* and see if anyone complains
12:13
<annevk>
hsivonen, there's this plan to make a new DOM spec for the Web
12:14
<annevk>
hsivonen, we keep the old spec around as standard for Java and to upset nobody, but the spec for browsers would be this other thing
12:14
<hsivonen>
annevk: my point is, will it break something that really assumes that the JDK DOM interfaces and the Web DOM interfaces match?
12:15
<annevk>
aah, dunno
12:15
<annevk>
they already diverge here and there so I don't think it matters much
12:15
<hsivonen>
I wonder if there will be demand for GWT wrapping the browser DOM in org.w3c.dom interfaces
12:16
<hsivonen>
I can imagine someone might want to be able to compile the same code on the server and on GWT
12:16
<zcorpan>
http://simon.html5.org/specs/web-dom-core
12:17
<hsivonen>
although GWT could probably fake the incompatible parts somehow
12:18
<zcorpan>
i've merged in dom2views
12:18
<zcorpan>
i'll drop all dom feature stuff except those that return booleans
12:18
<zcorpan>
(and are implemented)
12:21
<annevk>
dom2views is also merged in http://dev.w3.org/csswg/cssom-view/
12:21
<zcorpan>
oh
12:21
<annevk>
including elementFromPoint
12:21
<zcorpan>
cool
12:22
<zcorpan>
dropped both
12:23
<annevk>
feel free to merge the exceptions from XMLHttpRequest and Web Workers in btw
12:23
<annevk>
if your spec moves fast enough we can just reference that then :)
12:25
<zcorpan>
done
12:26
<hsivonen>
ajaxian.com requires a New York Times -style registration in order to comment :-(
12:28
<hsivonen>
the MSDN custom tag namespace stuff is amusing
12:28
<hsivonen>
<HTML XMLNS:MY>
12:28
<hsivonen>
no URI value
12:29
<hsivonen>
is the documentation for creating MathPlayer and Renesis-compatible DOM nodes via script in IE?
12:30
<roc_>
hsivonen: Java can still access the Gecko DOM, I'm pretty sure
12:31
<roc_>
as can Flash
12:31
<hsivonen>
roc_: ok.
12:31
<roc_>
but not via the org.w3c.dom interfaces AFAIK
12:31
<hsivonen>
oh.
12:31
<hsivonen>
I expect the threading with Java is interesting
12:32
<roc_>
annevk: did you notice that we implemented ClientRect.width/.height on trunk?
12:32
<roc_>
the threading isn't really our problem
12:33
<roc_>
the Java plugin is moving to use NPRuntime scripting, the same plugin two-way scripting API that Flash uses
12:33
<annevk>
roc_, no, somehow missed that, was it a while ago?
12:33
<roc_>
I really have no idea how that looks on the Java side
12:34
<roc_>
annevk: actually very recent
12:34
hsivonen
was unaware of NPRuntime scripting
12:35
<annevk>
roc_, oh ok, I guess I need to add you or someone else to my bugzilla tracking preferences then :)
12:35
<roc_>
http://hg.mozilla.org/mozilla-central/rev/a612d3e553b2
12:36
hsivonen
finds http://www.dessci.com/en/products/mathplayer/author/dynamic.htm#Scripting
12:37
<roc_>
Neil was really just trying to move scroll* and client* up to all elements, but also rolled in height and width
12:37
<annevk>
neat
12:39
<hsivonen>
will IE8 mode break these custom namespace ActiveX controls?
12:39
Philip`
wonders if it's possible for a CGI script to print raw HTTP headers via lighttpd, instead of it being parsed and reserialised before transmission
12:41
<roc_>
I'm still a bit nervous about returning fractional values from ClientRect. People see it and assume it's a bug. However, I still think it's the right thing to do and it doesn't seem to have caused any actual damage
12:41
<Philip`>
hsivonen: I think they're meant to still work in most cases, except for e.g. https://connect.microsoft.com/IE/feedback/ViewFeedback.aspx?FeedbackID=333905 if you're using CSS to assign namespace behaviour
12:43
<hsivonen>
Philip`: can the <?IMPORT stuff be inserted via JS or does the parser need to see it literally?
12:43
<annevk>
roc_, the spec does say float :p
12:44
<hsivonen>
Microsoft, Renesis and DesignScience aren't particularly forthcoming about this
12:44
<Philip`>
hsivonen: Do you count document.write as being inserted via JS?
12:44
<hsivonen>
Philip`: I do
12:44
<hsivonen>
Philip`: although I'd like to have a purely parserless way
12:45
<hsivonen>
Philip`: document.write should be OK for packaging the V.nu parser as a JS shim
12:45
<hsivonen>
Philip`: however, for the HTML5 Live DOM, I'd like to have a fully scripted API
12:47
<Philip`>
gsnedders: It works fine if I put the <?import...?> in document.write
12:47
<Philip`>
Uh
12:47
<Philip`>
s/gsnedders/hsivonen/
12:48
<hsivonen>
Philip`: OK. thanks
12:48
<hsivonen>
Philip`: do you know if it persists if the document contents are removed after that?
12:49
<hsivonen>
does IE even support emptying the document and inserting a new root?
12:50
<Philip`>
hsivonen: You can also do document.namespaces.add('v', 'http://www.w3.org blah blah it doesn\'t matter what you put here', '#default#VML');
12:50
<hsivonen>
Philip`: excellent. thanks
12:51
<Philip`>
http://msdn.microsoft.com/en-us/library/ms535932(VS.85).aspx - woah, it's even documented
12:52
<zcorpan>
hsivonen: it does so long as the new node is html and you use replaceChild (iirc)
12:53
<hsivonen>
zcorpan: thanks. (the HTML5 Live DOM doesn't use replaceChild)
12:54
Philip`
discovers a new way to crash IE8
12:54
<zcorpan>
hsivonen: or wait, you can empty the document and then append a new child, too
12:54
<zcorpan>
hsivonen: but you can't just empty the document and leave it empty
12:55
<hsivonen>
zcorpan: how can it force me to insert something and not leave it empty?
12:59
<Philip`>
http://philip.html5.org/tests/ie8/cases/document-close.html - yay for IE8
13:05
<zcorpan>
hsivonen: it throws if you don't insert anything
13:05
<zcorpan>
hsivonen: and leaves it in the state it was before emptying
13:06
<hsivonen>
zcorpan: ah.
13:06
<hsivonen>
(without testing, I'm assuming this is one of those weird situations where the scripted view to the DOM and the engine understanding of the DOM are synced only upon spin through the event loop)
13:08
<zcorpan>
Philip`: what happens?
13:10
<Philip`>
zcorpan: IE8b2 just freezes permanently
13:11
<Philip`>
While trying to report the bug, something's broken Opera so it's using 100% CPU and making my fans very loud, but at least Opera is still being responsive
13:12
<Philip`>
(To be more specific, IE8b2 freezes permanently and uses 100% CPU too)
13:13
<Philip`>
(To be more accurate, IE8b2 freezes permanently and uses 100%-divided-by-number-of-CPU-cores CPU too)
13:21
<annevk>
btw, http://tc.labs.opera.com/html/img/usemap/001.htm demonstrates that name in combination with usemap is not affected, despite usemap taking a URL reference originally...
13:22
<annevk>
tested in Firefox 3, Internet Explorer 6, and Opera 9.5
13:25
<zcorpan>
annevk: matches the spec
13:25
<annevk>
right
13:26
<annevk>
just wondering name can always be normalized or not
13:26
<annevk>
^^ whether
13:29
<Philip`>
hsivonen: http://hsivonen.iki.fi/introducing-sax-tree/ - s/in and/in an/
13:30
annevk
was just about to mention that
13:41
Philip`
wonders what's the most efficient way to read a load of HTML pages from disk
13:41
<Philip`>
I suppose the best way is to not actually read them from disk, so if I can compress them so they'll fit in RAM then that'd be good
13:41
<zcorpan>
can i drop normalize() and normalizeDocument() from dom core?
13:41
<Philip`>
but then I'll probably want more pages until it's too much to fit in RAM again :-(
13:45
<annevk>
zcorpan, prolly
14:01
<annevk>
hahaha, http://lists.w3.org/Archives/Public/public-html/2008Sep/0175.html
14:02
<annevk>
the original message was already pointless, and now it's more
14:02
<annevk>
amazing
14:02
zcorpan
drops setAttributeNodeNS
14:02
<hasather>
annevk: :D
14:06
zcorpan
adds Element.children
14:07
<annevk>
zcorpan, you have time to work on this now?
14:07
<annevk>
zcorpan, also, when say "add" you mean "add in local copy"?
14:08
<zcorpan>
annevk: yes and yes :)
14:08
<annevk>
great and not so great :)
14:11
<zcorpan>
hmm, children is an HTMLCollection
14:12
<zcorpan>
layering problem
14:12
<annevk>
I once heard Ian suggest HTMLCollection should be part of the DOM specs
14:12
<zcorpan>
but under another name?
14:12
<annevk>
no
14:12
<zcorpan>
ok
14:13
<annevk>
but having references from DOM Core to HTML5 and vice versa doesn't seem bad eiter
14:13
<annevk>
either*
14:14
<zcorpan>
i think dom core should be standalone
14:14
<zcorpan>
but for now i'll reference html5
14:14
<annevk>
i agree with the sentiment, not sure if it'll work out
14:18
<zcorpan>
uploaded new revision
14:22
<annevk>
any chance you can do HTML5/XMLHttpRequest-style IDL fragments rather than oldschool-DOM-spec-style IDL fragments?
14:23
<annevk>
eg, dropping "raises(DOMException)", dropping a lot of space characters and newlines
14:24
<zcorpan>
yeah i'll clean it up second-pass
14:25
<annevk>
you should prolly also keep some notes in the source of what is left out for now
14:25
<annevk>
at least, I would find that useful
14:27
<zcorpan>
ok
14:53
<zcorpan>
r3: add comments about what is dropped; added Text, Comment, DocumentType, ProcessingInstruction
14:53
<zcorpan>
i think i have all interfaces now; is there something i've missed?
14:56
<annevk>
all the interfaces you want I assume? several interfaces appear to be dropped
14:57
<annevk>
DocumentPosition is needed fwiw; we recently had to add support for that
15:00
<zcorpan>
yeah. ok
15:06
<annevk>
http://validator.nu/?doc=http%3A%2F%2Ftc.labs.opera.com%2Fhtml%2Fimg%2Fusemap%2F002.htm ?
15:06
<annevk>
hmm
15:07
<zcorpan>
map is flow content
15:08
<zcorpan>
perhaps it should be changed to transparent phrasing?
15:11
<annevk>
I feel to see a good reason for why I should move the <map> out of the <p>
15:15
<annevk>
wow, fail
15:21
<annevk>
in case people were not convinced that proprietary is bad: http://uk.techcrunch.com/2008/09/04/startups-in-chaos-as-adobes-flashpaper-discontinues/
15:24
<smedero>
It feels like an edge use-case but is there a reason to allow <map> inside of <p> for something like a sparkline with an image map that had pointers to expanded info about different data points? (http://en.wikipedia.org/wiki/Sparkline)
15:24
<smedero>
(i'm trying to bring up the whatwg list to see if that has been brought up before, but uh.. it is list.whatwg.org isn't responding for me...)
15:26
<smedero>
(ahh, now it is)
15:27
<zcorpan>
since <map> doens't behave like a block in the parser it feels unintiutive to disallow it in p
15:28
<zcorpan>
(and it's rendered inline)
15:28
<zcorpan>
(iirc)
16:14
<zcorpan>
Hixie: why is "pre, code { font-size: inherit;" in http://www.whatwg.org/style/specification ?
17:10
<Philip`>
How lovely - BeautifulSoup is giving me a <div> element whose parent is a <div> element whose only child is an <a> element
17:12
<Philip`>
Also, html5lib fails a zillion tests, so there's no way I can change anything without being quite likely to unknowingly introduce a regression
17:42
<Dashiva>
So earlier it was all "@headers is only used in generated content, so it doesn't matter that it's complex and easy to mess up" and today it's "but the USER wants to use @headers in handwriting, so we should let him"
17:49
<takkaria>
http://mail.gnome.org/archives/xml/2008-September/msg00018.html -- icky
17:57
<smedero>
Dashiva: Well it is also funny because Gez said he modified his example to "simplify the problem"
17:58
<smedero>
I don't think he redesigned it.... but it was modified in some form and I asked for a pointer the original so I can understand what was changed and I got zilch.
17:58
<smedero>
(it wasn't just changes to anonymize the data... columns and such were removed to "make it easier to understand")
17:59
<Dashiva>
Simplify the problem of getting people to accept it
18:25
<sicking>
annevk, ping
18:37
<smedero>
FWIW, Gez got me a pointer to the unmodified complexdatatable example: http://juicystudio.com/wcag/tables/altcomplex.html
18:38
<smedero>
(More or less, includes extra columns at the end...)
18:42
<hsivonen>
whoa. Renesis has an ASV compat Quirks Mode
19:06
<hsivonen>
how do I create an element in IE via JS with a particular scopeName and tagUrn?
19:07
<Philip`>
Maybe by declaring a prefix with namespaces.add(...) and then createElement('prefix:name')?
19:07
Philip`
doesn't know any other way
19:08
<hsivonen>
Philip`: thanks
19:28
<hsivonen>
Philip`: that works
19:28
<hsivonen>
to the point of getting scopeName and tagUrn right
19:29
<hsivonen>
can I associate the scope name with an ActiveX control without inserting an <object> into the document?
19:33
<hsivonen>
giving a clsid: URL as the third parameter of namespaces.add does not work
19:37
<Lachy>
I don't know how to make it any clearer to John Foliot, he seems to be totally missing the point of what I'm saying
19:38
<Lachy>
I don't get why he can't comprehend that the proposed study is about seeing how effective longdesc is for users that currently have the ability to access it, nor why he doesn't understand why that would be useful
19:45
<hsivonen>
it seems crazy if <?import has a JS equivalent but <object>-based ActiveX hook doesn't
19:46
<hsivonen>
Philip`: "in and" fixed. thanks
19:47
<hsivonen>
Philip`: HtmlParser Javadoc fixed locally. thanks
19:56
<Dashiva>
What? There are AT apps that don't expose longdesc?
19:57
<webben>
Dashiva: There are AT apps that don't do ... lots of things.
19:58
<webben>
Dashiva: JAWS has exposed it since version 4.something.
19:58
<webben>
although in some recent versions, you've had to enable announcements for it.
19:58
<Hixie>
Lachy: as far as i can tell he's not interested in improving accessibility, only in sounding like he's an expert and getting his ego stroked by getting his way
19:58
<webben>
Dashiva: that behavior appears to have changed back to announcing by default in current JAWS 10 Beta.
19:59
<Lachy>
Hixie, fair enough. But honestly, did you clearly understand what the purpose of my proposed study, and why it would be useful?
20:00
<Lachy>
I just want to make sure its not me who's missing something
20:00
<Dashiva>
webben: Wasn't it so that most JAWS users rarely, if ever, upgraded? So if it's gone in the latest version, it's not a big deal
20:01
<webben>
Dashiva: "most". I'm not sure that's true. Certainly the lag time for upgrades is greater for commercial AT at hundreds of $ a shot than for free web browsers.
20:01
<Hixie>
Lachy: yes
20:02
<Hixie>
Lachy: your proposal was exactly the kind of study we need to get objective data
20:02
<webben>
Dashiva: I'm not sure what the referent of "gone" is though.
20:02
<Dashiva>
webben: default announcement
20:02
<Hixie>
Lachy: though next time i would phrase your hypothesis to sound like you agree with the side that isn't yours
20:03
<Hixie>
Lachy: since the hypothesis doesn't matter (we get the same results either way) and the people who are clearly unable to understand how to do research would at least not get a knee jerk reaction against you
20:03
<Philip`>
Now all we need is infinite resources so we can carry out reliable user studies without worrying about their cost
20:04
<Dashiva>
Hixie: They'd probably knee-jerk just from the from address
20:04
<webben>
Dashiva: No. It's the other way round, I think. Default announcement has come _back_.
20:04
<Philip`>
Hixie: But then they'll say the hypothesis is self-evident and there's no point doing new studies on it
20:04
<Dashiva>
So it was added, removed and then re-added? Quirky
20:05
<Lachy>
Hixie, ok. I'll keep that in mind. I'm rewriting it as clearly as I possibly can so that there can be absolutely no confusion about what I mean
20:05
<webben>
Lachy: FWIW I agree that your proposed study would be a very good addition to information collected so far, though I don't think studies of this (human-tech interactions) sort tend to produce "irrefutable" results.
20:05
<Lachy>
webben, ok, fair enough.
20:06
<Hixie>
Philip`: then you just respond "well then you wouldn't object to a study would you"
20:06
<Hixie>
studies never provide irrefutable results
20:06
<Lachy>
webben, my point was that the results would take us from having only observation and hypothesis closer to having a theory
20:06
<Hixie>
they can always be refuted by more studies that contradict them
20:06
<webben>
Hixie: Absolutely. It's not my word.
20:07
<Philip`>
Hixie: They should object to resources being spent on that study instead of on something more interesting and informative and insightful
20:11
<hsivonen>
does IE support adding methods to Element prototype?
20:11
<Hixie>
as i understand it IE doesn't have prototypes
20:12
<hsivonen>
Hixie: IIRC, it does have for objects that don't count as host objects
20:12
<hsivonen>
like Array.prototype
20:12
<webben>
Lachy: I think a useful addition to your proposed study would be to try and work out, if there's a difference, why.
20:12
<hsivonen>
I know this stuff is not supported for ActiveX-based XHR DOM nodes
20:12
<hsivonen>
I guess I should just test
20:13
<Lachy>
webben, any suggestion how?
20:13
<webben>
Lachy: well, qualitative observation during the process and asking people at the end would help I guess.
20:13
<Hixie>
hsivonen: right i mean for host objects
20:13
<Lachy>
ok
20:14
<Hixie>
hsivonen: i take it your validator instance wouldn't be able to handle the load of a google blog pointing at it?
20:14
<hallvord>
hsivonen: AFAIK Element.prototype and friends is coming in the very latest beta of IE8
20:14
<hallvord>
..well, should have used past tense: came .. thing is I haven't tested it yet :)
20:15
<webben>
Lachy: e.g. this might uncover whether (for example) people don't use longdesc because they don't expect it to work because of bad UA implementation (as with the guy in Joshue's video) or whether (for example) they miss the description link because of their navigation strategy.
20:15
<Hixie>
hallvord: really? wow
20:15
<hsivonen>
Hixie: it probably wouldn't handle it :-(
20:15
<Hixie>
k
20:15
<webben>
Hixie: Couldn't Google host an instance of it (if that were okay with hsivonen)?
20:16
<hsivonen>
it would be interesting to see how badly it would melt down, though :-)
20:17
<Hixie>
webben: hsivonen might be able to port it to appengine, i don't know
20:17
<hsivonen>
Hixie: if you had appengine for Java...
20:17
<hsivonen>
which I suspect Google already has internally for deploying its own Java-based apps
20:18
<Hixie>
i don't think we've deployed any java-based apps on appengine
20:18
<Hixie>
i don't know what the plan is for that team, either, so i don't even know if java is on the cards :-)
20:18
<hsivonen>
Validator.nu is well-suited for scaling on an infrastructure that replicates the app automatically
20:19
<hsivonen>
after all, it doesn't need write storage except for logs
20:19
<hsivonen>
and it doesn't need a login mechanism
20:20
<hsivonen>
so it could be put on a cloud computing platform without the usual BigTable or Amazon SimpleDB lock-in issues
20:20
<hsivonen>
it does have the issue that it wants to do *a lot* of initialization up front to build read-only data structures that are shared among threads
20:21
<hsivonen>
I figured that Validator.nu on a usual day requires less CPU and RAM than one EC2 compute unit provides
20:22
<hsivonen>
so currently, it doesn't make sense to invest in automatic scaling on EC2
20:23
<hsivonen>
hallvord: thanks.
20:24
<hsivonen>
I'm trying to come up with a way to route setAttributeNS calls to something else in IE without imposing a per-call cost on other browsers
20:25
<hsivonen>
Hixie: I've been speculating about what's next for App Engine
20:26
<hsivonen>
Hixie: and since Google has a small set of supported languages, Java is the obvious runner up after Python
20:26
<Hixie>
that does seem like a good guess
20:26
<Hixie>
(i really don't know what that team is doing)
20:26
<hsivonen>
Hixie: but I figured providing App Engine for Java would need new restrictions on the VM
20:26
<hsivonen>
and probably a class loading model that isn't transparently compatible with a normal JVM
20:27
<smedero>
mostly unrelated but there's a 20% project going to get Perl into App Engine... so there's a chance that Java could be a 20% as well.
20:27
<hsivonen>
but I wouldn't be too surprised if Google hacked up yet another environment that uses the Java language without the full JVM runtime semantics
20:30
<hallvord>
hsivonen: see
20:30
<hallvord>
http://www.microsoft.com/windows/internet-explorer/beta/readiness/developers-new.aspx#mutabledom
20:30
<hallvord>
http://www.quirksmode.org/blog/archives/2008/09/ie8b2_roundup.html
20:30
<hsivonen>
smedero: I'd expect Java to require more drastic VM hacking than Perl
20:30
<smedero>
indeed.
20:30
<hsivonen>
hallvord: thanks
20:33
<Hixie>
"comments are HTMLCommentElement interfaces and inherit from Element" lol
20:34
<Hixie>
"Early on, we had to decide if we wanted to implement a "fake" layer for the constructor/prototype objects that looked identical to the W3C model, or to expose what we had, warts and all."
20:34
<Hixie>
...and they picked the warts... because of their commitment to standards?
20:35
<hsivonen>
"Object doesn't support this property or method"
20:35
<hsivonen>
sigh
20:35
<hsivonen>
I'd appreciate it if IE said which property
20:36
<Hixie>
this one
20:36
<Hixie>
duh
20:39
<hsivonen>
anyway, I think this HTML5 parsing thing can be programmed on top of IE eventually, although I'm failing ATM
20:39
<hsivonen>
my proof of concept does work: http://hsivonen.iki.fi/test/moz/ie-namespaces.html
20:39
<hsivonen>
now if WebKit rendered dynamically inserted SVG...
20:41
<webben>
hsivonen: What are you trying to do exactly?
20:41
<webben>
implement a HTML5 parser as an activex control?
20:42
<hsivonen>
webben: I'm trying to emulate createElementNS for MathML and SVG namespaces in IE7+MathPlayer+Renesis
20:42
<hsivonen>
webben: so that the Validator.nu HTML Parser when compiled with GWT could work in IE as well
20:43
<webben>
ah okay
21:36
<hsivonen>
reloading my test page crashes IE7
21:50
<zcorpan>
hsivonen: iirc, attributes can't be namespaced in ie, though i don't know how the plugins handle it
21:50
<Lachy>
Hixie, webben yt?
21:51
<Lachy>
Hixie, webben, anyone else, could you review this before I send it? http://lachy.id.au/temp/study.txt
21:55
<webben>
Lachy: I think the distinction between a user agent requiring users to change their configuration to be aware of long descriptions (e.g. earlier versions of JAWS) and announcing them by default (e.g. current JAWS 10 Beta on reading, iCab on hover) is significant.
21:56
<webben>
although I realize John only says "natively support" in the quote.
21:57
<Lachy>
webben, I wasn't aware of that. What should I say about it?
21:59
<webben>
Lachy: I guess a prelude to this study would need to be some sort of agreement about what a decent implementation of longdesc might do and whether any current implementations do it.
21:59
<webben>
If none do, you'd need to introduce an test implementation that does (e.g. by hacking NVDA perhaps).
21:59
<Lachy>
I was just going to have the test run with each user's normal software and setup, to make it as realistic as possible
22:00
<webben>
realistic to what?
22:02
<Lachy>
to the way they use it use it normally
22:02
<Lachy>
it would be a bit unfair to have the test performed by a user who is unfamiliar with the software being used
22:03
<webben>
Lachy: I agree that it introduces other complications. But I don't think a test of bad implementations would tell one much about the actual point at issue (at least from my understanding of the point at issue).
22:04
<webben>
Lachy: So I guess I'm saying you _might_ have to take software users are familiar with and modify it somewhat.
22:04
<webben>
I'm not saying: make a brand new screen reader or browser or something ;)
22:06
<Lachy>
I don't like that idea, because that would require informing the user of what changes have been made, which would bias the results due to having additional information about the test that they would not otherwise have
22:07
<webben>
Lachy: Hmm. Couldn't you tell them that you've modified it but not tell them how?
22:07
<Lachy>
in what way would you want to modify it?
22:07
<webben>
for that matter, actually, what's wrong with them knowing about the feature?
22:08
<webben>
otherwise you're not just testing whether the feature can do a given task, but also the initial intuitiveness of the feature's presentation, which could bias the study in both directions
22:09
<Lachy>
because if the users didn't know about the longdesc before hand, and are supposedly representative of the general population, then informing them about the feature would influence the result
22:09
<webben>
e.g. people might think, hmm, what's this long description thing, must try that. or people might think, hmm, dunno what that is, better not try that.
22:10
<webben>
yes, but you're not trying to test people's knowledge of features, as far as I can tell.
22:10
<Lachy>
I'm trying to test how useful a feature is in practice given the current, near the top of the range ATs
22:11
<webben>
okay, but that's not the actual point being debated.
22:12
<webben>
Lachy: again, that's basically testing implementations not the concept.
22:14
<Lachy>
hmm. ok. So then we would have to set a baseline for the ATs that can be used.
22:14
<webben>
Lachy: yeah, exactly, it very much depends on working out what a "good" implementation would do.
22:15
<webben>
which isn't necessarily trivial.
22:15
<Lachy>
fair enough. I suppose that's fair given how we would test the proposed headers algorithm
22:15
<Lachy>
which would need modified UAs/ATs to do it
22:15
<webben>
that is to say, it's easy to distinguish very bad from better implementations, but hard to distinguish very good from better implementations.
22:16
<webben>
e.g. Firefox native longdesc implementation is pants. the longdesc firefox add-on clearly improves it.
22:16
<webben>
iCab is an improvement again (because of the on-hover indicator).
22:17
<webben>
- that's for sighted users.
22:17
<webben>
I'm pretty sure that defaulting to announcing longdesc is "better" than defaulting to not announcing it.
22:19
<Lachy>
I added a paragraph saying "We need to set a set of baseline feature requirements for the assistive technologies that can be used. I'll leave this for those who are more familiar with them,"
22:26
<Lachy>
I sent it
22:36
<annevk>
'Jonas Jacobi, Kaazing’s CEO, will be delivering a three day course through Kaazing’s partnership with SkillsMatter in London on October 6th, 2008. The course is entitled “Comet Evolved: HTML 5 Web Sockts & Server-Sent Events.”' -- http://thepeninsulasedge.com/blog/?p=156
22:36
<annevk>
for features with no browser implementations, that's pretty impressive
22:44
<annevk>
hsivonen, http://www.tbray.org/ongoing/When/200x/2003/12/13/MegaXML (credit: tag)
22:53
annevk
added it to http://wiki.whatwg.org/wiki/Namespace_confusion
23:17
<Hixie>
we want to test implementations and not the concept
23:18
<Hixie>
because users interact with the implementations, not the concepts
23:20
<Lachy>
Hixie, yeah. But we should let them use the optimal implementation. If the beta of jaws 10 notifies the user by default as webben said, then that's fair enough
23:21
<Lachy>
making them use implementations they know to be flawed isn't likely to be accepted
23:21
<webben>
Hixie: Testing implementations is a worthwhile activity as long as one is clear about what that actually means.
23:23
<Lachy>
but I agree, modifying the implementation well beyond would be considered a typical setup to could bias the results too much, so there needs to be limits
23:24
<webben>
I suspect that radical implementations aren't necessary, but I'm not quite sure what other people's view of an optimal implementation would be.
23:25
<Lachy>
I would prefer it to be a near default setup, with additional options only set based on each users' normal configuration
23:26
<webben>
e.g. I think JAWS 10 Beta seems reasonable for a blind user (not sure about deafblind) and iCab 4.x seems reasonable for a mouse-wielding sighted fellow.
23:26
<webben>
the problem with JAWS 10 Beta is ... it's beta.
23:26
<Lachy>
how is iCab considered assistive technology?
23:28
<webben>
From my perspective, these are just tools and sets of tools users with various abilities use to consume content.
23:28
<Lachy>
but they need to be users for whom long descriptions are considered the most useful
23:29
<webben>
I don't think long descriptions are necessarily more useful to blind people than colorblind people etc.
23:29
<Lachy>
John's complaint is that since no-one uses longdesc on the web, and because it's a relatively new feature in ATs, most users don't know about it, nor query for it.
23:30
<Lachy>
While I see that as a fundamental flaw with its design, he seems to think there is hope for it
23:31
<webben>
longdesc (as opposed to long descriptions) is useful (well, in _theory_) to people navigating non-linearly or any other case where you might want automatic extraction
23:32
<Lachy>
but if the default configuration was to notify the user about its presence, then that could potentially fix one of the problems. But I'm still not sure how to fix the problem of authors not using it
23:33
<webben>
I think that's like trying to fix the problem of Flickr not using alt.
23:33
<webben>
Lachy: In so far as authors might _want_ to use it, they've repeatedly been given guidance not to use it because of broken implementations.
23:34
<webben>
I think that guidance was wrong (because facilitating programmatic association is helpful, I think).
23:34
<webben>
I also agree that providing links to things or including descriptions on the page is good.
23:36
<Lachy>
I'm just not convinced that longdesc has any advantages over ordinary links, and ordinary links have the benefit of being available to everyone
23:38
<webben>
Lachy: I'm quite interested, on behalf of _future_ user agents, in potential mechanisms to avoid redundancy during linear processing while still facilitating the automatic addition of context during non-linear processing.
23:38
<webben>
though I think link text is a much more important case for this than img/longdesc
23:38
<Hixie>
authors do use longdesc
23:38
<Hixie>
but they almost uniformly use it in harmful ways
23:39
<webben>
Hixie: I suspect if more implementations were like iCab there would have been less harmful use. Hard to prove though :)
23:40
<webben>
Hixie: It might be interesting to know why authors who did things wrong, did things wrong.
23:41
<webben>
the most confusing aspect of longdesc seems to be that it's a URI. I guess better naming might have helped there - descriptionref or something.
23:42
<Lachy>
what does iCab do exactly?
23:43
<webben>
Lachy: 1. If you hover over an img with longdesc it changes the cursor to a arrow with a little doc icon. (iCab uses cursor hints for various things.)
23:46
<webben>
2. If you open the context menu on such an img, one of the options is Show long description of Image. If you click that, it opens the referenced URL in a new window. (not that keen on the new window as opposed to new tab business myself, but there we go)
23:46
<webben>
So it's roughly similar to (say) viewing an image in its own tab in Firefox.
23:46
<Hixie>
is the description more useful than just zooming the image and giving a high-contrast vie?
23:46
<Hixie>
view
23:47
<webben>
Hixie: I'd imagine that would depend on the image and the user.
23:48
<Hixie>
can you name a case where an author would create a description that was detailed enough that it was more useful than zooming the picture?
23:49
<webben>
well, where the picture isn't very good to start with would be one example.
23:49
<webben>
i'm not sure how good high-contrast conversions are.
23:49
<Hixie>
wouldn't a description in such cases be better given inline for everyone to enjoy?
23:49
<webben>
i.e. how bad the information loss is
23:50
<Hixie>
seems to me that whenever an image could be described in detail, that the description becomes generally useful and should just be an integral part of the page, not hidden away in a dedicated attribute
23:50
<Lachy>
webben, I don't think low such quality images are a typical use case
23:51
<webben>
Hixie: dedicated attributes are an integral part of the page.
23:51
<webben>
Hixie: Where are the visible links to open images in their own tab? Or print them? Or download them? Or copy them to the clipboard?
23:51
<Lachy>
I agree with Hixie. I think it would be better to encourage authors to provide information about images designed for everyone, which has the effect of helping those who can't see the image well either
23:51
<Hixie>
webben: there's no markup for those either
23:52
<webben>
Hixie: Sure there is. It's called src.
23:52
<Hixie>
webben: well then linking to a description on another page is called <a href=""> and including a description inline is called <p>...</p>.
23:53
<Lachy>
sure, but src isn't meant to be human consumable information. The image is
23:53
<webben>
Lachy: likewise with the description.
23:53
<Hixie>
webben: basic language design -- accessibility should be inherent, people don't think about it and if you make it separate they will screw it up.
23:53
<Hixie>
anyway
23:53
<Lachy>
<a href="description"><img src="foo"></a>
23:53
<Hixie>
longdesc="" has been shown to be a lost cause
23:54
<Hixie>
unless new information comes along, there's no point arguing it
23:55
<webben>
Hixie: Well, yeah I just don't buy the "shown" bit. Can't really help that. :)
23:55
<Hixie>
have you read http://blog.whatwg.org/the-longdesc-lottery ?
23:55
<webben>
yeah
23:55
<Hixie>
well if you don't think that shows it's a lost cause, i don't know what to tell you
23:56
<webben>
Hixie: Yes, I don't see how you can separate the concept from the implementations and the guidance given.
23:56
<webben>
*how you can't, rather
23:57
<Hixie>
doesn't matter what the implementations do
23:57
<Hixie>
over 96% _of images with longdesc_ have bogus longdesc
23:57
<webben>
I think that statistic is heavily determined by the implementations and the guidance given.
23:57
<Hixie>
there is no implementation stategy that can turn crap into usable content
23:58
<Hixie>
that statistic is actual web content.
23:58
<Hixie>
doesn't matter what causes it
23:58
<Hixie>
it's already caused.
23:58
<Hixie>
we could argue about whether to introduce a _new_ attribute
23:58
<Hixie>
and maybe implementations could support that better and we could avoid this problem
23:58
<webben>
Hixie: Oh, yeah, when I'm talking about the "concept", I'm talking about the "concept".
23:59
<Hixie>
however, one has to wonder why it would take 10 years for implementations of longdesc="" to come along, if it's such a great idea
23:59
<webben>
Hixie: iCab's implementation has been around longer than that.
23:59
<webben>
it's not a recent innovation I mean
23:59
<webben>
what i'm not clear about is when JAWS didn't read longdesc by default and why