00:20
<TabAtkins_>
hober: I think your Change Proposal is good.
00:33
<hober>
thanks! any suggestions for improvements?
02:05
TabAtkins
doesn't know who people are at this talk, so is just hanging out until we have a panel discussion.
02:07
<annevk>
there's a talk first?
02:07
<TabAtkins>
Man, apparently?
02:07
<TabAtkins>
http://voyageofhtml5.eventbrite.com/
02:07
<Dashiva>
So where's our april's fools joke?
02:08
<TabAtkins>
There is no April Fools joke this year.
02:08
<Dashiva>
Is that the joke?
02:08
<TabAtkins>
You'll never know.
02:08
<Dashiva>
"Consider HTML5 is such a joke to begin with, WHATWG has decided to instead commit to a April's Fools serious activity"
02:09
<TabAtkins>
Heh, where's that from?
02:09
<Dashiva>
I made it up
02:09
<Dashiva>
April's fools!
02:09
<TabAtkins>
Ah, gotcha.
02:09
<annevk>
laughing out loud!
02:55
<boblet>
MikeSmithX: you there?
02:55
<MikeSmith>
boblet: yep
02:56
<MikeSmith>
sorry for not replying to e-mail
02:56
<MikeSmith>
I was traveling yesterday
02:56
<boblet>
np :) thought you’re busy so maybe IRC is faster
03:01
<MikeSmith>
boblet: let's ping annevk later when he's on
03:01
<annevk>
I'm here
03:01
<MikeSmith>
oh
03:02
<annevk>
waddup?
03:02
<boblet>
heh
03:02
<MikeSmith>
annevk: boblet's going to Fukuoka also on April 24 for the Web event I mentioned
03:02
<MikeSmith>
we need to decide how many days to stay there
03:02
<boblet>
annevk: hear you’re coming to Japan to partake in Mike’s … unique hospitality style :)
03:03
<boblet>
should be lots of fun
03:03
<annevk>
it was cool last time :)
03:03
<annevk>
my only constrain is that I fly out May 8
03:04
<boblet>
I think Oikawa-san is doing a same-day trip, but Shiraishi-san wants to stay Saturday night. dunno if Yakura-san has mentioned his plans yet
03:05
<boblet>
I’ve got work on 25th, so am trying to decide whether to leave 24th after event or crazy early 25th
03:06
<boblet>
It’ll be my first time to Kyuushuu so thinking to go 23rd, and wondering what your plans are…
03:06
<boblet>
what time constraints do you have Mike?
03:07
<MikeSmith>
none
03:07
<MikeSmith>
so I guess we can arrive on the 23rd
03:08
<MikeSmith>
and then stay over on the 23rd and 24th at least
03:08
<MikeSmith>
maybe head back on the 25th or even 26th
03:10
<boblet>
it sounds like they’re gonna organise a hotel, but I’m guessing that’d be for a night. ok I might stay 23rd then head back on 24th. I’ll miss the after-party but I wouldn’t be able to drink anyhow
03:11
<MikeSmith>
we can have a before-party on the 23rd
03:11
<boblet>
MikeSmith: you know we’re meant to send an indication of costs right? and check the profile page by this w/e?
03:11
<MikeSmith>
yeah, I saw that
03:11
<boblet>
before-party sounds great, esp. as it’s an afternoon event
03:11
<MikeSmith>
but I have not had time to check on flights or hotel
03:12
<MikeSmith>
did Shirakawa-san give any hotel info yet?
03:16
<boblet>
MikeSmith: apparently Shimokawa-san in Fukuoka has offered to book a hotel for us. I guess somewhere reasonable (they’re paying) and near the venue, which hopefully means convenient for you both outside the covered day(2?)
03:16
<boblet>
s/day(2)/day(s)/
03:17
<boblet>
annevk: they’re also wondering if you’d like to have your name associated with the event…
03:18
<boblet>
not sure if that means just saying that you’ll be there, or if they’re sounding you out as a speaker
03:18
<MikeSmith>
boblet: do we have a URL for this thing yet?
03:19
MikeSmith
shuold probably go back and read the e-mail
03:43
<annevk>
for continuity, the above was handled in /msg
03:47
<annevk>
so for the CSS value API we need some kind of hashmap and some kind of list
03:47
<annevk>
e.g. for 'margin' you need a hashmap and for 'cursor' you'd need a list
03:48
<annevk>
the hashmap can be immutable, but the list needs to be mutable
03:49
<annevk>
another constrain is that both the hashmap interface and list interface might need to be implemented on the same object in certain scenarios (for backwards compatibility; if something changes from being just a hashmap into a comma-separated list of hashmaps)
03:49
<annevk>
current sketch I have:
03:49
<annevk>
map { mapLength, mapKey(), getter getMapItem() }
03:51
<annevk>
list { listLength, insertListItem), deleteListItem(), getter getListItem() }
03:52
<annevk>
since getListItem() takes an int and getMapItem() takes a string the getter thing ought to work
03:52
<annevk>
thoughts?
04:04
<annevk>
http://twitter.com/Bess/status/11402770234 -- 'Tab Atkins says "sexiest" thing in HTML5 is CSS transformation. CSS camp is more peaceful than JS camp. Too much fight over JS camp'
04:04
<annevk>
o_O I wonder what was actually said
04:25
<djbpython>
anyone have any good examples? having trouble parsing some html into an lxml etree
05:30
<othermaciej>
Hixie: I am sad that the YouTube HTML5 beta no longer works :-(
05:42
<MikeSmith>
djbpython: using html5lib?
05:54
<shepazutoo>
hsivonen: ping
06:03
<MikeSmith>
othermaciej: my WebKit keeps crashing
06:03
<MikeSmith>
r56855
06:04
<othermaciej>
MikeSmith: hmmm
06:04
<othermaciej>
I am running my own build of r56882
06:04
<othermaciej>
it does not seem particularly crashy
06:04
<othermaciej>
is there any particular site that reproduces it?
06:06
<MikeSmith>
I got a bunch of tabs open
06:06
<MikeSmith>
and it seems to be crashing not on any particular site
06:07
<MikeSmith>
I'll mess around with it some more and see
06:22
<MikeSmith>
othermaciej: crashing on the "Thank you for testing WebKit" page
06:22
<othermaciej>
MikeSmith: consistently?
06:22
<othermaciej>
http://nightly.webkit.org/start/about is not crashing for me in my build
06:24
<MikeSmith>
so it opens http://nightly.webkit.org/start/trunk/56855 , then if I close that window (so there are no more windows open), then go to my History and select the "Thank you for testing WebKit" page, then let it open, then close it, then it crashes, consistently
06:25
<MikeSmith>
it gives me a "Do you want to resent application settings for WebKit?" dialog
06:25
<othermaciej>
MikeSmith: can't repro on my current build; will try it with an actual nightly
06:26
<MikeSmith>
OK
06:26
<MikeSmith>
I tried the "Reset and relaunch" option, but it seems to have made no difference
06:28
<hsivonen>
shepazutoo: pong
06:29
<hsivonen>
today is the day when you can't trust news :-(
06:29
<MikeSmith>
othermaciej: maybe it's an OS problem.. I've not restarted my machine for a long time
06:30
<othermaciej>
tried with the nightly - cannot repro
06:31
<MikeSmith>
ok
06:31
<MikeSmith>
I guess I'll try restarting my machine
06:37
<Lachy>
Has anyone come across any obvious or funny april fools day hoaxes yet?
06:37
<othermaciej>
I hope the YouTube HTML5 beta is an April Fool's joke
06:37
<othermaciej>
er the HTML5 beta being broken
06:57
<Lachy>
http://youtube-global.blogspot.com/2010/03/textp-saves-youtube-bandwidth-money.html
06:58
<Lachy>
I wish I could find a video that that new feature works on.
07:00
<othermaciej>
Lachy: works on this: http://www.youtube.com/watch?v=iDe4v318f64&feature=featured&textp=fool
07:01
<Lachy>
doesn't work for me
07:02
<othermaciej>
worked for me
07:02
<othermaciej>
in fact, works every time I visit that page
07:04
<Lachy>
damn geo-ip restrictions. Only works for me when I connect to a VPN in the US
07:18
<Lachy>
hmm, maybe it's not intentional geo-ip restrictions after all. It works from London too. It just doesn't work within Norway. It might just be something to do with their CDN caching, and not having the updated Flash video distributed everywhere yet.
08:00
<hsivonen>
Hixie: Did you make defer on VBScript invalid on purpose?
08:00
<hsivonen>
Hixie: everything that's not JS is a data block, right?
08:55
<MikeSmith>
othermaciej: restarted my machine but still crashing
09:17
<MikeSmith>
hmm, "lxml.etree.XMLSyntaxError: htmlParseCharRef: invalid xmlChar value 55349, line 59559, column 2233"
09:17
<MikeSmith>
when running the splitter
09:19
<Philip`>
That's a surrogate thingy?
09:19
<Philip`>
Sounds like the Anolis serialisation bug
09:19
<Philip`>
creating invalid HTML which lxml is therefore unable to load
09:20
<Hixie>
hsivonen: no, what makes you think that?
09:20
<Philip`>
(i.e. it's not my fault)
09:20
<Hixie>
hsivonen: executable code = script. non-executable = data.
09:20
<jgraham>
MikeSmith: You obviously don't read the logs :)
09:21
<MikeSmith>
jgraham: I read them, but I skip all the boring parts
09:22
<jgraham>
I stand by my original statement then
09:23
<jgraham>
Anyway it is a bug in the python stdlib that makes the html5lib serializer spit out the wrong character references for non BMP characters
09:23
<jgraham>
gsnedders promised to fix it yesterday, so complain to him
09:24
<jgraham>
(and then to me to update pms I guess)
09:27
<zcorpan>
Hixie: should we invent a mime type specifically for <script> data blocks?
09:27
<jgraham>
(and, then, for completenes, at Hixie to regen the spec)
09:27
<MikeSmith>
jgraham: Ok, thanks
09:27
<MikeSmith>
btw, I did just find it in the logs also
09:28
<Hixie>
zcorpan: wouldn't the mime type depend on what type the data block uses?
09:28
<zcorpan>
Hixie: so that people know what to use and newer scripting languages are less constrained by legacy
09:28
MikeSmith
wonders what to do in the mean time, because this is breaking the build of the multipage W3C copy of the spec
09:29
<zcorpan>
Hixie: a parameter or data-* could be used to distinguish different types of data blocks
09:29
<MikeSmith>
"some of the named character references are in fact combining characters".. great stuff
09:30
<jgraham>
I'm not sure that is insane from a MathML point of view
09:30
<Hixie>
zcorpan: i don't understand what problem that would solve, though it seems like it would make things more complicated in general
09:31
<zcorpan>
Hixie: i don't know what to put in type='' if i want to use a custom data block
09:32
<zcorpan>
Hixie: if i make something up, it might clash with some future scripting langauge that makes up the same mime type as i did
09:32
<MikeSmith>
jgraham: MathML should really be discouraging people from used named character references, instead of facilitating it
09:32
<Hixie>
the same type you would use for content-type if you were sending that block over http
09:32
<Hixie>
MikeSmith: i'm not sure if the characters being combining is intentional -- we should ask the mathml wg about that
09:33
<zcorpan>
hmm, i guess that would be text/plain
09:33
<MikeSmith>
Hixie: I see
09:34
<jgraham>
MikeSmith: If hand authoring is a goal there is no other sane way to do it
09:34
<jgraham>
If had authoring is not a goal we get into "the tools will save us" territory
09:35
<MikeSmith>
god save anybody who tries to do hand authoring of any serious mathml
09:36
<zcorpan>
indeed
09:36
<jgraham>
Well yeah it is painful
09:36
<jgraham>
But moving it from painful to effectively impossible is still not good
09:40
<hsivonen>
Hixie: so a validator can't check the requirement because the validator can't know which MIME types are executable by some UA
09:40
<hsivonen>
Hixie: I thought data blocks were defined to be things that aren't JS
09:41
<hsivonen>
Should I expect class selector matching to be sufficient proof of what encoding was used for a CSS file in IE6?
09:42
<Hixie>
hsivonen: well, you can hard-code that text/javascript and a few other script types are script types and assume most other types are not, but yeah, there's no way to reliably check that
09:42
<Hixie>
hsivonen: but then, you have the same problem with reliably checking the syntax of the scripts themselves
09:42
<Hixie>
hsivonen: you can only check the syntax of things you know about
09:42
<hsivonen>
right
09:42
<zcorpan>
Hixie: but it's easy to check async/defer
09:42
<zcorpan>
Hixie: and src
09:43
<hsivonen>
oh. right. you aren't supposed to use src with data blocks
09:43
<hsivonen>
Hixie: the newly-introduced requirement is redundant, then
09:43
<hsivonen>
which confused me
09:44
<Hixie>
how so?
09:44
<Hixie>
i made a bunch of changes today
09:44
<Hixie>
i'm not sure which you're talking about
09:44
<zcorpan>
async and defer must not be present if src isn't
09:44
<zcorpan>
src is not allowed on data blocks
09:44
<hsivonen>
Hixie: the one that says async and defer are prohibited on data blocks
09:45
<hsivonen>
r4927
09:45
<Hixie>
oh, oops, i missed the requirement on async and defer needing src
09:45
<Hixie>
only saw it for charset
09:45
<boblet>
Dear cabal, for your amusement: http://twitter.com/boblet/status/11414105037 /cc TabAtkins
09:48
<hsivonen>
is the REPLACEMENT CHARACTER supposed to match in a selector just like any other character?
09:48
<Philip`>
boblet: That actually has been proposed on the WHATWG list, I believe
09:50
<Philip`>
boblet: http://www.w3.org/Bugs/Public/show_bug.cgi?id=6999
09:50
<Philip`>
(I was wrong)
09:52
<boblet>
Philip`: omg, not even in April!
09:52
<boblet>
wow, reality rules
09:53
<hsivonen>
yeah, reality is always crazier than April Fools.
09:54
<jgraham>
Also from WHATWG "Should we have attributes to indicate whether <video> content [...] is NSFW"
09:54
<hsivonen>
this XHTML 3.0 announcement is an old joke, but it's pretty much like XSL-FO: http://www.cs.tut.fi/~jkorpela/html/xhtml3.html
09:55
<Hixie>
hsivonen: looks like what a lot of people in public-html are suggesting should be valid
09:56
<hsivonen>
Hixie: I don't know if that's the criterion the chairs will apply.
09:56
<Hixie>
which criterion?
09:57
<asmodai>
heh, got to love xkcd's april's fool
09:57
<hsivonen>
Hixie: that things that a lot of people suggest to be valid should be
09:58
<jgraham>
hsivonen: One of the principal suggesetees is himself a chair
09:59
<Hixie>
hsivonen: oh i wasn't suggesting that it would pass or fail or that it was a good or bad idea, i was just pointing out that that april fools was another example of something that some people were suggesting seriously
09:59
<hsivonen>
Hixie: oh
10:00
<hsivonen>
jgraham: well, previously one complaint was that Hixie had opinions of his own. at least Maciej and Paul participate in gauging if Sam's opinion enjoys consensus
10:01
MikeSmith
puts a s/5[5-6][0-9][0-9][0-9]/32/g into his makefile
10:02
<Hixie>
we haven't actually seen what the chairs base their decisions on yet
10:03
<Hixie>
so far as far as i'm aware we haven't seen any non-editorial decisions from the chairs
10:03
<Hixie>
and their editorial decisions have been more or less random
10:03
<zcorpan>
MikeSmith: won't that affect other parts of the spec? e.g. the websocket handshake examples have some random numbers
10:04
<zcorpan>
maybe they don't match the pattern
10:05
<Hixie>
(e.g. microdata is in its own draft despite being pretty tightly integrated with html, but a random meta name keyword is in the same draft as <meta> despite being mostly useless and completely independent)
10:05
<hsivonen>
Hixie: even though the basis isn't overt, all the three chairs are supposed to participate in the decision
10:05
<Hixie>
(have there been other decisions?)
10:06
<Hixie>
hsivonen: they might be using dice for all i can tell :-)
10:07
<Hixie>
we don't really have enough data to be able to tell yet
10:08
<MikeSmith>
zcorpan: I meant it just as a workaround until it's fixed or worked around in html5lib
10:09
<hsivonen>
argh. the rejection of non-text/css sheets makes stuff harder to test
10:09
<hsivonen>
I need to configure my server to have a separate set of extensions for misconfigured versions of each MIME type
10:10
<hsivonen>
I already map .html to text/html; charset=utf-8 and .htm to text/html
10:11
<hsivonen>
and yes, apparently REPLACEMENT CHARACTER works like any other character for selector matching
10:19
<gsnedders>
Yay for 4933 giving us a load to update in html5lib
10:19
<hsivonen>
has Opera started rejecting non-text/css sheets even in the quirks mode?
10:19
<zcorpan>
hsivonen: no
10:19
<Hixie>
ok i should sleep now
10:19
<Hixie>
nn
10:20
<hsivonen>
zcorpan: oops. I forgot to change my default encoding in Opera back to Windows-1252 after setting it to a Cyrillic encoding
10:24
<hsivonen>
It would be very interesting to test this document.open() stuff in IE5
10:24
<hsivonen>
and Netscape 4
10:26
<hsivonen>
Hixie: do you happen to remember why you made the spec say that the encoding for document.open()ed docs is UTF-16?
10:26
<hsivonen>
oh. Hixie went to sleep
10:30
<zcorpan>
"You can tell if a browser has so designated its content handler
10:30
<zcorpan>
by trying to display a JPEG in <embed>."
10:30
<zcorpan>
browsers support jpeg in embed
10:50
<asmodai>
Hixie sleeps?
10:50
<asmodai>
I thought he was a script.
10:53
<zcorpan>
scripts can sleep
10:57
<Dashiva>
They just lock up the application when they do
10:58
<zcorpan>
Hixie locks up HTML5 when he sleeps
11:05
<asmodai>
heh
11:47
<Lachy>
today's xkcd is brilliant :-)
11:47
<asmodai>
Lachy: tried sudo apt-get dist-upgrade yet?
11:48
<Lachy>
asmodai, no. Just tried (sudo) make me a sandwich
11:50
<asmodai>
interesting read about that privacy issue with :visited
11:50
<asmodai>
http://hacks.mozilla.org/2010/03/privacy-related-changes-coming-to-css-vistited/
11:52
<gsnedders>
dbaron posted something too
11:52
<gsnedders>
On his own site
11:53
<Lachy>
haha, try the cat command :-)
11:54
<Dashiva>
It doesn't jive well with having browser shortcuts on . and -
11:54
<Lachy>
yeah, it doesn't seem to work in opera well at all
11:58
<asmodai>
always Opera... ;)
11:59
<asmodai>
gsnedders: but it's linked to from that link I pasted
11:59
<gsnedders>
Oh, I didn't notice that when I read it yesterday
12:01
<Philip`>
It saves a lot of effort if you just read http://github.com/chromakode/xkcdfools/blob/master/xkcd_cli.js instead of typing in commands
12:02
<Dashiva>
I just put the page source into jslint
12:04
<asmodai>
rofl, even a small adventure clone built in
12:04
<Dashiva>
I didn't get this one for the asl: '6/M/Battle School'
12:05
<asmodai>
oneliners['emacs'] = 'You should really use vim.';
12:05
<asmodai>
oneliners['vi'] = oneliners['vim'] = 'You should really use emacs.';
12:05
<Philip`>
Ender?
12:06
<Philip`>
http://en.wikipedia.org/wiki/List_of_Ender's_Game_series_organizations#Battle_School
12:06
<asmodai>
Never read it (yet).
12:07
<gsnedders>
asmodai: Uh, yeah. That comes from a xkcd comic.
12:08
<asmodai>
gsnedders: Read them all but can't remember that one
12:20
<gsnedders>
hsivonen: http://validator.nu/?doc=http%3A%2F%2Fjacobrask.net%2Fcontenteditable-style.htm
12:21
<gsnedders>
hsivonen: it doesn't seem to like <!doctype html><body> <style scoped></style>
12:21
<gsnedders>
(note the spac)
12:21
<Creap>
<- jacobrask
12:21
<gsnedders>
Oh, the space doesn't matter
12:22
<gsnedders>
<!doctype html><title></title><body><style scoped></style> is invalid if you believe the validator
12:22
<gsnedders>
Creap: Do you need different nicks on different servers!?
12:22
<brucel>
c'mon, who believes a validator? puh-lease!
12:22
<gsnedders>
I don't ;P
12:23
<jacobrask>
gsnedders: work and non-work ;)
12:23
gsnedders
just gets confused because his Opera and W3C usernames are "gsneddon"
12:27
<brucel>
nice glass dolphin, eh, gsnedderws
12:30
<nessy>
nice new currency ;)
12:31
<hsivonen>
gsnedders: thanks. I have no idea how that bug arose.
12:36
<asmodai>
hsivonen: Curious thought: when do you anticipate the mathml to stop throwing errors? :)
12:38
<hsivonen>
asmodai: I want to say "later this month", but because I don't want to make promises I can't keep, I'm not promising anything
12:38
<asmodai>
hsivonen: Ok, but it's on a fairly soonish schedule, which is cool
13:03
hsivonen
notes that the commenters on the Chromium blog and the commenters on the Flash Player blog have very different attitudes towards bundling Flash with Chrome
13:13
<shepazutoo>
hsivonen: yt?
13:13
<hsivonen>
shepazutoo: yes
13:18
<shepazutoo>
hsivonen: I'm starting to write up wording for allowing @href/@src instead of/in addition to @xlink:href, and I was wondering if you had any thoughts on how we can resolve the dual-attribute conflicts...
13:24
<hsivonen>
shepazutoo: the most obvious thing that comes to mind would be to use xlink:href it it is present and the new thing otherwise
13:31
<jgraham>
Happy mailman reminder day everyone
13:33
<shepazutoo>
hsivonen: what about the DOM value reflection for say, @href and @src?
13:45
<hsivonen>
shepazutoo: is the DOM property now href?
13:48
<hsivonen>
hmm. it is
13:49
<hsivonen>
adding no-namespace href and src may be more trouble than they are worth
13:50
<hsivonen>
shepazutoo: I don't have a good idea for reflecting no-namespace href and src consistenly
13:50
<asmodai>
When did Firefox stop giving clipboard access by default? 3.6?
13:51
<hsivonen>
shepazutoo: since .href setter will have to continue to set xlink:href
13:51
<hsivonen>
or I'd assume that has to happen for compat
13:53
<hsivonen>
shepazutoo: for setters to work sanely, you should probably have new DOM properties for no-name href and src
13:53
<hsivonen>
shepazutoo: but I can't think of a sane name for the href property
13:54
<hsivonen>
shepazutoo: why do you want to introduce new attributes with duplicate behavior?
13:54
<hsivonen>
setAttributeNS sucks, but if script authors use the href setter, they don't need the NS-aware attribute setter
13:55
<hsivonen>
and writing xlink:href is not *that* terrible when factoring in the backwards compat benefit
13:57
<shepazutoo>
hsivonen: it's been a pain point for several developers (though I didn't mind so much myself), and for it would be nice for SVG newbies to have more consistency between SVG and HTML
13:57
<hsivonen>
shepazutoo: if we're making incompatible changes to SVG, let's kill the load event
13:57
<hsivonen>
shepazutoo: I see
13:58
<shepazutoo>
hsivonen: maybe that (the load event) should be on the table, too)
13:58
<hsivonen>
shepazutoo: anyway, it seems to me that the hardest part is figuring out the DOM property setter story for the no-namespace href
13:58
<hsivonen>
shepazutoo: sorry I don't have a solution
13:58
<hsivonen>
shepazutoo: has the SVG WG already agreed to make the load event async?
13:59
<hsivonen>
shepazutoo: zcorpan said Opera implements it as async
13:59
<shepazutoo>
hsivonen: re: load, I don't recall... we did discuss it
13:59
<shepazutoo>
that will be resolved in SVG 2, I imagine
14:00
hsivonen
would love to have a preliminary resolution before we ship the HTML5 parser
14:00
<shepazutoo>
hsivonen: re: href setter, yeah, it's tricky
14:00
<shepazutoo>
hsivonen: what's your timeframe?
14:00
<hsivonen>
shepazutoo: no hard timeframe yet
14:00
<hsivonen>
turning it on by default was a Q1 goal but now it's a Q2 goal...
14:01
<hsivonen>
shepazutoo: anyway, resolving the bug is scheduled for Q2
14:01
<hsivonen>
the bug being that the HTML5 parser doesn't dispatch the SVG load event
14:04
<shepazutoo>
hsivonen: then let's focus on getting something in the SVG 2 spec in that tiemframe
14:42
<brucel>
Can I just grab some svg and paste it into my html5 code and have it render? I know I will be able to, but can I now?
14:47
<jgraham>
brucel: In Gecko with html5.enable, yes
14:47
<jgraham>
For most values of svg
14:48
<hsivonen>
FWIW, double hyphens in comments are truly interop-sensitive: https://bugzilla.mozilla.org/show_bug.cgi?id=556037
14:49
<Philip`>
brucel: And in IE9
14:54
<jgraham>
brucel: Obligatory example: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/441
14:54
<hsivonen>
jgraham: I get an Internal Error
14:55
<jgraham>
Oh
14:55
<jgraham>
Yeah
14:55
<jgraham>
Well nevermind, make up your own demo :)
14:55
<jgraham>
(it was probably too long)
15:05
<brucel>
thanks. So I need an html5 parser (or IE9!) to have SVG in HTML5 in practice, tho its allowed in the spec as is
15:06
<beowulf>
http://jonnycampbell.com/ # i think the sections should be articles…
15:09
<jgraham>
brucel: Yes, presumably most browsers are going to add SVG-in-text/html support as part of larger parser changes for HTML5 compliance
15:10
<brucel>
thx.
15:32
<beowulf>
ok, that link i posted is a little weird, for me it showed code, for someone else it's showing a weblog. ah well.
15:37
<jgraham>
beowulf: Seems to be a design decision
15:37
<jgraham>
s/incorrect spellings/correct spellings/
16:22
<JonathanNeal>
Mornin!
17:23
<gsnedders>
So, apparently I should make some decision about university.
17:23
<gsnedders>
I don't like decisions.
17:24
<jgraham>
Well it seems like you don't have that many options so it should be really easy
17:25
<othermaciej>
gsnedders: what kind of decision? which one to attend?
17:28
<gsnedders>
othermaciej: So I have an offer from Glasgow for CS (almost everywhere else rejected me). Do I bother to go to Glasgow or not, will I gain enough from it for it to be worthwhile?
17:29
<othermaciej>
gsnedders: I don't have a good picture of what schools are good in the UK
17:29
<othermaciej>
gsnedders: this is undergrad, or graduate?
17:29
<gsnedders>
undergrad
17:29
<jgraham>
gsnedders: FWIW I think you are underestimating the amount of useful stuff in the degree course
17:30
<jgraham>
I can't imagine you already know all the useful theory + mathematics + so on
17:30
<othermaciej>
gsnedders: I think what you learn in an undergrad CS program is likely to be very useful, assuming you are interested in a career in programming or related professions
17:30
gsnedders
isn't sure about taht
17:30
<gsnedders>
*that
17:31
<othermaciej>
I came into college thinking I knew a lot, but I learned a lot, including important frameworks for how to think about things
17:31
<jgraham>
gsnedders: Which part? The career interest or the fact that the material is useful?
17:31
<gsnedders>
The career
17:31
<jgraham>
Well if you're not interested in it you are doing a very poor job of demonstrating it
17:31
<othermaciej>
in the US you don't have to pick your field of study before you start
17:32
<jgraham>
othermaciej: Typically here you do
17:32
<jgraham>
(although of course there is always more flexibility than they let on)
17:32
<othermaciej>
gsnedders: are there other things you might be interested in pursuing as a career?
17:33
<gsnedders>
othermaciej: Quite a few English related careers (writing, journalism, editing, publishing I all would quite like) as well aeronautical engineering
17:35
<othermaciej>
gsnedders: at MIT, the aero/astro program was famous for being brutally hard, and yet with not so great career prospects
17:35
<othermaciej>
it was widely believed to be the hardest major
17:36
<othermaciej>
not saying that should dissuade you, just reporting
17:36
gsnedders
didn't get the impression it had that bad career prospoects
17:36
<jgraham>
If you are mainly interested in career prospects (which are not the only, or necessarily even the main consideration) all the things you listed are pretty poor
17:36
<othermaciej>
I expect writing is something you could pursue as a minor or simply by taking the right classes while doing another degree
17:37
<othermaciej>
gsnedders: it's not that getting a job is impossible, it's just that it apparently doesn't pay very well compared to other engineering fields, in light of how hard it is
17:39
<jgraham>
(in particular all beginning English students want to be writers and/or journalists. They all end up being teachers)
17:39
<jgraham>
(note: there is some degree of generalisation there)
17:40
<beowulf>
i studied geology, it has something like 3% employment in the field rate for grads
17:40
<gsnedders>
jgraham: I know, that's part of the reason why I'm not quite taht enthusastic
17:40
<beowulf>
you can always do another one :)
17:42
<gsnedders>
othermaciej: I guess I'll do Gaelic at university then...
17:51
<gsnedders>
So yeah, I have no idea.
20:29
<Lachy>
http://www.trainsignal.com/WindowsTraining.aspx
20:41
<gsnedders>
jgraham: Will you murder me if I stay in Sweden?
20:47
<gsnedders>
No RFCs published yet today. :(
21:01
<jgraham>
gsnedders: Possibly. And do what?
21:47
<AryehGregor>
Bwahaha, http://annevankesteren.nl/test/contenteditable-style.htm is great.
21:50
<AryehGregor>
Of course, the cool part is the execution, not the actual functionality.
21:50
<AryehGregor>
The latter would be pretty easy.
22:34
<gsnedders>
jgraham: Hide my body?
22:48
<jgraham>
gsnedders: I doubt that would work. At least that's what TV taught me
22:49
<jgraham>
(there could be an element of selection bias though)
22:49
<jgraham>
Anyway it clearly wasn't what I meant :)
22:49
<gsnedders>
jgraham: More seriously, stay at pera
22:49
<gsnedders>
*Opera
22:49
<gsnedders>
jgraham: Since when did I ever give you answers straight away which were what you meant when there was an obvious mis-interpretation?
22:52
<jgraham>
gsnedders: Since it took you 1.5 hours to come up with that response I would hardly call it straight away