00:02
<Philip`>
erlehmann: That's not a reeducation chamber, it's an incineration chamber
00:02
<Philip`>
You can't fool me that easily
00:04
<erlehmann>
Philip`, its perfecty safe. the WHATWG wishes to remind you that incineration state is not yet part of the spec.
00:10
<TabAtkins>
No one would happen to have an xp home sp3 license key, would they? I'm reimaging a comp for a friend, and of course his windows disk is long gone.
00:11
<TabAtkins>
Alternately, a trustable iso for xp without any service packs. I have a key for that.
00:14
<Philip`>
Don't computers normally have license keys printed in a sticker on the back or something?
00:14
<Philip`>
s/in/on/
00:15
<TabAtkins>
Yeah, but that's for xp home sp0. The key won't work for a higher service pack.
00:15
<TabAtkins>
double alternately, I have an xp pro sp2 disk. I have a surfeit of keys and disks, but none that match.
00:33
<MikeSmith>
can somebody who has IE7 or IE8 running please do a quick test of something for me?
00:33
MikeSmith
too lazy to start his Windows VM
00:34
<TabAtkins>
shoot
00:53
<CryptoQuick>
Reposted from #html (sorry!)-- Hello folks! I have a very specialized problem... while using canvas within foreignobject in SVG, I can't stack two canvases on top of each other and have the one behind it still be visible. one canvas I have markers and cursors and stuff that's updated often, then in the other is CPU-intensive, persistent stuff.
00:53
<CryptoQuick>
trying this doesn't work:
00:53
<CryptoQuick>
ctx.globalCompositeOperation("destination-over");
00:53
<CryptoQuick>
ctx.drawImage(bg, 0, 0, 1000, 800);
00:56
<erlehmann>
CryptoQuick, why don't you just have one canvas?
00:56
<erlehmann>
also, isn't this a browser-specific thing?
00:56
<CryptoQuick>
I'm trying that now, actually-- with save and restore?
00:56
<CryptoQuick>
browser-specific? no, I don't think so, it does the same in Firefox and WebKit
00:56
<erlehmann>
CryptoQuick, tried it with all of webkit, gecko and presto ?
00:57
<CryptoQuick>
I can try presto now, one sec
00:57
<CryptoQuick>
oh lord
00:57
<CryptoQuick>
I hope people don't use opera :P
00:57
<CryptoQuick>
it messed up my unicode! >:I
00:57
<CryptoQuick>
anyway
00:58
<CryptoQuick>
now my new approach is to draw the markers directly on the canvas, then restore it at the end of the function...
00:58
<CryptoQuick>
is that the right way to do it? :I
00:59
<erlehmann>
dunno. i'm no canvas expert.
00:59
<erlehmann>
CryptoQuick, i'd ask on mozilla IRC if i were you.
00:59
<CryptoQuick>
alrighty, thanks :)
01:42
<MikeSmith>
anybody running dev-channel Chrome on Linux? I've got 223.11 and my status bar seems to have completely disappeared.. wondering if it's a known issue
02:04
<MikeSmith>
http://www.indicthreads.com/3625/html-5-websocket-cracks-the-http-request-response-barrier/
03:21
<MikeSmith>
http://trac.tools.ietf.org/area/app/trac/wiki/IriWorkGoals
03:26
Hixie
goes cross-eyed trying to understand some of the e-mails on the hybi list
03:29
<MikeSmith>
Hixie: I just added links to the "Error handling in URIs" discussion thread on the uri⊙wo list from last year
03:30
<MikeSmith>
added to that wiki page, that is
03:30
<MikeSmith>
speaking of HyBi - http://trac.tools.ietf.org/bof/trac/wiki/HyBi
03:56
<othermaciej>
good evening
04:05
<MikeSmith>
othermaciej: so I'll be at both the HyBi and IRI BOFs on November 10 in Hiroshima, nominally representing W3C and the HTML WG. At some point between now and then, would like to make sure I'm synced up with others from the group to make sure I can try to make whatever points we want made
04:05
<MikeSmith>
Hixie also ↑
04:05
<Hixie>
yeah let's meet next week
04:05
<Hixie>
grab me at some point
04:05
<MikeSmith>
OK
04:05
<MikeSmith>
hai
04:05
<Hixie>
i'm gonna be running around like a headless chicken
04:06
<MikeSmith>
I know Larry has been working on try to put together some f2f time for IRI discussion next Thursday, but don't know that it's been decided yet or not (I've got some e-mail in my inbox to read yet)
04:07
<Hixie>
no decision that i've seen
04:07
<Hixie>
unicode is meeting at the same time in cupertino or something
04:07
<MikeSmith>
yeah
04:07
<Hixie>
some of us might have to take a ride down there at some point
04:07
<Hixie>
dunno what the story is
04:08
<MikeSmith>
I'll look into to.. figure out where it is at least
04:09
<MikeSmith>
anyway, I have to step out for bit now. back later
04:11
<othermaciej>
MikeSmith: I would be glad to chat about it sometime during the next week
04:11
<othermaciej>
Hixie: I'm sending Alexey to the Unicode meeting
04:11
<othermaciej>
at least, to the part that's supposed to be relevant to IDNA
04:13
<Hixie>
I hope they fix the mess that is IDNA2008
04:13
<Hixie>
making it incompatible with IDNA2003 is just so stupid
04:14
<othermaciej>
yeah the meeting is all about how to fix the security and compatibility issues caused by IDNA2008 changes
04:14
<othermaciej>
which led both Alexey and me to think, "in that case, why are we changing it?"
10:24
<gsnedders|work>
hsivonen: I realized I wasn't really going to be able to do it quickly enough for the script to really be worthwhile, and didn't do it
10:26
<hsivonen>
:-(
10:38
<hsivonen>
whoa. google paper-spams me to advertise Web advertising
10:44
<erlehmann>
┌─┐
10:44
<erlehmann>
┴─┴
10:44
<erlehmann>
ಠ_̼ರೃ
10:44
<erlehmann>
GENTLEMEN
10:47
<jgraham>
erlehmann: BTW the answer is both
10:49
<erlehmann>
jgraham, please elaborate on that in the context of the original-fake distinction
10:51
<erlehmann>
if the object you believed to me orininal turns out to be fake, does the object lose value? or does the idea of the object lose value and the objects value stays the same?
10:51
<erlehmann>
s/to me/to be/g
10:57
<jgraham>
erlehmann: The idea of the object loses value (where value is the personal worth I abscribe to it) because in the case of art it is the concept that is valuble (in the case of e.g. an oven it is the object itself that is valuble since the idea of an oven is useless for making food)
10:58
<erlehmann>
got it. thanks
11:34
<hsivonen>
I wonder if I should take the Google ad bait just to burn the 75 euros of "free" advertising credit on Validator.nu
11:35
<hsivonen>
there must be a catch, but I fail to see the catch
11:37
<gsnedders|work>
hsivonen: When I have more time I may look closer again
12:02
<Dashiva>
hsivonen: You get addicted to the added traffic and start paying for it
12:32
<hsivonen>
is http://www.w3.org/TR/MathML3/ the latest public MathML 3 draft?
12:36
<MikeSmith>
hsivonen: you mean WD? or editor's draft?
12:41
<hsivonen>
MikeSmith: whatever is the latest. probably ED
12:42
<MikeSmith>
hsivonen: I'm trying to look and see if I can find an ED
12:43
<hsivonen>
(it just seems useless to read WDs if EDs are public)
12:43
hsivonen
tried to google with site:dev.w3.org but failed
12:43
<zcorpan_>
their WG page didn't seem to point to any EDs
12:44
<hsivonen>
ok
12:44
<hsivonen>
maybe I should just read the WD, then
12:48
<MikeSmith>
hsivonen: they don't use dev.w3.org
12:48
<MikeSmith>
but I can't find where elsewhere they store them
12:48
<hsivonen>
MikeSmith: OK. thanks
12:48
<MikeSmith>
but anyway, I checked their member-only list
12:49
<MikeSmith>
there's no mention of an updated ED in their list archives
12:49
<hsivonen>
MikeSmith: I don't want to read anything secret, because I want to be able to discuss stuff if I find something to comment on
12:49
<MikeSmith>
hsivonen: yeah
12:49
<MikeSmith>
anyway, it does seem the WD is the latest
12:50
<zcorpan_>
hsivonen: you can discuss it in secret
12:51
<jgraham>
You could talk to yourself about it, for example
12:51
<hsivonen>
zcorpan_: I want to be able to whine here on IRC without going to #secrettreehouse.
12:55
<hsivonen>
aargh. need to find a way to kill the "click and hold Dock icon for Exposé" feature on SL
12:55
<hsivonen>
I've now triggered Exposé a zilloin times on SL without wanting to trigger it
12:56
<hsivonen>
s/oin/ion/
12:57
<Rik`>
hsivonen: i've never such an option
12:57
<Rik`>
and it's for your own good, they know what you want :)
12:57
<MikeSmith>
hsivonen: wait, how do you know about #secrettreehouse !?
12:58
<zcorpan_>
MikeSmith: he's part of the cabal
12:58
MikeSmith
shuts down #secrettreehouse and moves to #reallysecrettreehouse
13:12
<gsnedders|work>
MikeSmith: Where's that? Same location in the Amazon as the previous one, just higher up that tree?
13:18
<Philip`>
gsnedders|work: If he said, it would hardly be secret
13:19
<jgraham>
I thought Amazon keps all its treehouses in the Colouds these days
13:19
<jgraham>
*Clouds
13:20
<MikeSmith>
gsnedders|work: the old #secrettreehouse was actually in Romania
13:20
<gsnedders|work>
Surely they aren't treehouses then any more?
13:20
<MikeSmith>
former Bush administration extraordinary-rendition site
13:20
<MikeSmith>
the thing about the Amazon was a ruse
13:20
gsnedders|work
wonders whether to make some really obscure Twilight reference to the fact it used to be in Romania
13:21
<gsnedders|work>
I'd probably get it wrong though.
13:22
MikeSmith
reads some e-mail and wonders, Why can't people just love me for who I am?
13:22
<MikeSmith>
the only person in this whole thing who is more lovable than me is gsnedders|work
13:22
<MikeSmith>
it's downhill from there
13:23
<gsnedders|work>
Obviously I should lose all my old-world romanticism and become more like everyone else.
13:23
<MikeSmith>
we should have a ranking
13:24
<MikeSmith>
Que es mas encantador?
13:24
<MikeSmith>
¿Senor gsnedders|work o Senor MikeSmith?
13:27
<MikeSmith>
dammit... 22:30 and I'm getting kicked of the Starbucks terrace at Ebisu station
13:27
<MikeSmith>
now have to find somewhere else to steal electricity
13:28
<gsnedders|work>
"There are no tears/Just pity and fear"
13:30
<MikeSmith>
that's a bright thought
13:30
<MikeSmith>
gsnedders|work: see, that's why people love you
13:34
<MikeSmith>
whiskey
14:01
<Dashiva>
I didn't see this one posted anywhere: http://www.flickr.com/photos/mrlastweek/4014327734/
14:06
<TabAtkins>
Dashiva: ::sigh::
14:45
<Dashiva>
The ARIA discussions are so wobbly
14:45
<Dashiva>
On one hand, ARIA is supposed to be really good and useful. But on the other hand, there's a constant refrain of "We can't do that, because it would require tools to support ARIA"
14:46
<Dashiva>
You'd think supporting core accessibility technology would be a given for accessibility tools
14:57
<mikekelly>
hi fans
14:58
<mikekelly>
what was the upshot of the discussion about http authorization - did anything change?
15:35
<gsnedders|work>
Wow, it's dark outside.
15:36
<gsnedders|work>
Halloween must be upon us!
15:36
<Phae>
you up in the highlands?
15:36
<gsnedders|work>
Phae: Linköping
15:36
<Phae>
oh. i thought you were UK somewhere.
15:36
<Phae>
fair dos.
15:36
<gsnedders|work>
Phae: I have more sense than to be in the highlands over Halloween
15:36
<Phae>
:)
15:36
<gsnedders|work>
Phae: Until June I was at secondary school in St Andrews, where I'd lived me entire life. Then some idiot at Opera decided to give me a job. :P
15:37
<gsnedders|work>
(So, yes, I was in the UK somewhere)
15:39
<Phae>
ah, ok. that's what i thought
15:54
<TabAtkins>
Hmm, I didn't realize that anchor links counted as "relative" for the purpose of <base>.
15:54
<TabAtkins>
That sort of sucks.
16:07
<AryehGregor>
That is a pain.
16:08
<Philip`>
TabAtkins: Why would they be anything other than relative links?
16:08
<TabAtkins>
Philip`: Because they're anchor links, which in my mind is a third category. No page is named at all.
16:08
<AryehGregor>
The point is that clicking them shouldn't navigate to a different page if <base> is specified. That's unexpected.
16:09
<AryehGregor>
Of course, <base> is kind of evil anyway except for testing.
16:09
<gsnedders|work>
I don't think using it for testing makes it less evil, it's just there's no less evil way of doing the testing
16:09
<TabAtkins>
Heh, <base> is really useful to me, actually. It lets me rewrite the hierarchy of my site without caring about all the links in content.
16:46
Philip`
decides to append his initials to bug reports, so that he has a non-zero chance of finding them all again
18:01
MikeSmith
wonders if [pt] is a representation for the sound of spitting
18:03
<MikeSmith>
e.g., "the script would run when the attribute was set, not when the element was ; appended" [sound of spitting]
18:08
<Philip`>
Quite possibly
18:09
<Philip`>
It's useful as a unique string to search on, since not many other people spit on their bugs
18:18
<MikeSmith>
Philip`: I'm going to start to put [pt] on my bugs
18:18
<MikeSmith>
or [ppppt]
18:18
<MikeSmith>
for extra "p"
21:46
<ap>
hsivonen: is http://www.w3.org/Bugs/Public/show_bug.cgi?id=7059#c20 still what we need to do?
22:31
<ap>
Hixie: DTD parsing is optional, but if a UA supports entities in normal parsed source (as we do), why not support them in innerHTML?
22:31
<Hixie>
you shouldn't support them in normal parsed source either, imho
22:34
<ap>
Hixie: besides, an entity could be declared in internal subset, which isn't optional to parse IIRC. if the spec just said that the DTD should be taken from target, both cases would work automatically
22:34
<Hixie>
i guess
22:36
<ap>
Hixie: not supporting &nbsp; and friends is not really an option for us, because they work in Firefox
22:36
<Hixie>
it bugs me that xml left this massive interop hole
22:37
<Hixie>
i don't really mind how we fill it
22:37
<ap>
Hixie: you could document what WebKit does :)
22:37
<Hixie>
but we need to make it clear what browsers and other XHTML UAs are to do
22:37
<ap>
Hixie: i.e. hardcode known XHTML DTDs and support character entities
22:37
<Hixie>
i guess i have to bite the bullet and "profile" XML
22:40
<Dashiva>
That sounds like the author-friendly solution
22:41
<Hixie>
ap: can you file a bug or send a mail or something asking me specifically to define how entities are to be parsed by XHTML5 UAs?
22:41
<Hixie>
ap: (if you send mail, send it to public-html so that all the "you can't profile xml!!!!!" people see it and can raise a stink now rather than once i put it in the spec "without discussion")
22:44
<ap>
Hixie: ok
22:44
<ttepasse>
I declare: We need an appendix of the HTML5 parsing algorithm expressed in State Chart XML.
22:44
<Hixie>
ttepasse: if you can provide one (and instructions on how to keep it up to date when the spec changes), i'm happy to add it to the spec :-)
22:45
<ttepasse>
That sound's like work. I just like to see it for my personal amusement. ;)
22:47
<Dashiva>
I'm sure Philip` can drum something up eventually
22:47
<Dashiva>
He's the go-to guy on weird stuff like that :)
22:48
<ttepasse>
Can he do EmotionML on HTML5?
22:48
<ttepasse>
(Sorry, catching up on W3C news ...)
23:08
<Philip`>
It's easy to encode the HTML5 parser as a state machine
23:08
<Philip`>
Start with an initial, then for every HTML document add a transition labelled with that document as a string, leading to a node which is associated with the appropriate DOM
23:08
<Philip`>
s/initial/initial state/
23:14
<Dashiva>
Actually
23:14
<Dashiva>
By what metric are we judging?
23:14
<Dashiva>
Because I don't think that parser would fit on my hard drive