00:22
<Hixie>
2 lines out of 160 e-mails so far
00:22
<Hixie>
sigh
00:23
<Philip`>
Does "I want latex2html to give less ugly output" count as a problem?
00:23
<Hixie>
it's closer than most of these e-mails
00:24
<Hixie>
"I have content in LaTeX and want to convert it to HTML and have it retain a high degree of typographical quality" is a problem
00:24
<Hixie>
though you should probably include an explanation of _why_ you want to convert it to HTML
00:24
<Philip`>
Because Adobe Reader is horribly slow, so I don't want to use PDF, and there's no other option
00:25
<Hixie>
well HTML isn't an option either now
00:25
<Philip`>
(Actually I'd just use PDF anyway, because it's people's own faults if they don't have a nicer PDF reader)
00:25
<Hixie>
so really your problem is "I have content in LaTeX and want to view it in a good viewer while retaining a high degree of typographical quality"
00:26
<Hixie>
at which point, HTML really doesn't seem like it has a good place in any solution for that problem
00:26
<Philip`>
It needs to be a widely-distributed reader too, so that most people are able to access my content
00:26
<Philip`>
and it needs to be indexed by search engines, so that people can find my stuff
00:27
<BenMillard>
PDFs do get indexed by search engines?
00:27
<Hixie>
these all seem like LaTeX problems, not HTML problems :-)
00:28
<Philip`>
BenMillard: They do by at least Google, which is all that counts
00:28
<Philip`>
Hixie: They're still problems, and problems that could be solved by HTML
00:28
<BenMillard>
how about "I have content in LaTeX and want to include it on my web page."
00:28
<Hixie>
to be honest i don't really see the problem here
00:28
<Hixie>
BenMillard: yes, that's a different problem entirely
00:29
<Hixie>
Philip`: if you have content in LaTeX, and you publish it to PDF, you already have a widely available viewer, good typopraphy, and searchability
00:29
<Hixie>
Philip`: seems like a solved problem
00:29
<Hixie>
BenMillard: and a valid problem
00:29
<othermaciej>
PDF is its own world
00:30
<Hixie>
BenMillard: though not one that anyone on the mailing list seems to have mentioned so far :-) (in the 163 mails i've read)
00:30
<othermaciej>
it seems nice in principle to be able to do math decently without escaping to a different basic technology for your whole document
00:30
<othermaciej>
I have to agree that MathML syntax seems vile, now that I have looked more closely
00:31
<Philip`>
Hixie: Many people use latex2html now, despite it being slightly ugly, so there must be reasons why they do that, and I'm just not entirely sure what those reasons are :-)
00:31
<BenMillard>
hixie, perhaps you are being to picky about their wording?
00:31
<Hixie>
Philip`: yeah, me either
00:31
<Philip`>
but that doesn't make the reasons not exist
00:32
<Hixie>
BenMillard: no, i'm trying to pull problems out of these e-mails, even if they haven't mentioned them intentionally. Sadly most of these e-mails are arguing about which syntax to use, on the basis of the syntax's own characteristics.
00:32
<Hixie>
not suitability to solving any particular problem
00:32
<BenMillard>
hixie, ah
00:32
<Hixie>
Philip`: oh i agree. but if i don't know what the problems are, i can't solve them, or evaluate proposed solutions for them.
00:41
<Hixie>
man, i'm going to be skipping a lot of these e-mails in my eventual reply
00:41
<Hixie>
so much arguing back and forth about nothing
00:42
<BenMillard>
the inexpressiveness of plain text can bring out the worst in people (including me)
00:42
<Hixie>
oh i've seen these discussions happen in person too
00:43
<Hixie>
the only difference is that those discussions don't end up in my inbox two years later :-P
00:43
<BenMillard>
win!
00:44
<Hixie>
Elliotte Harold wrote:
00:44
<Hixie>
> The specific syntax is important because there's a huge, useful toolchain for processing XML and there's essentially zilch for
00:44
<Hixie>
> processing this strange HTML 5 thing.
00:44
<Hixie>
yay for all of you guys implementing the html5 tools over the past 2 years
00:56
<Hixie>
wohay, e-mail 250/368 is henri's e-mail earlier this month
00:56
<Hixie>
118 e-mails in a few weeks
00:56
<Hixie>
hopefully _some_ will have problem descriptions
00:56
<Hixie>
because the first 249 e-mails had nothing beyond what's on the wiki now
01:04
<Hixie>
ok henri's e-mail more than doubled the size of that wiki page
01:15
<BenMillard>
would the top level of list items be more useful as headings on that page?
01:16
Hixie
tries
01:16
<Hixie>
nah, that looks more ugly i think
01:17
<Hixie>
http://wiki.whatwg.org/index.php?title=New_Vocabularies&oldid=2894 is what it looked like
01:17
<Hixie>
i've reverted it
01:19
<Hixie>
i wonder if mathml would take off more if it were available in text/html, but IE continued to not support it
01:21
<vlad_>
are there any useful tools for producing mathml?
01:21
<Hixie>
hm, http://www.w3.org/mid/ABFEED64-8A5F-43C9-981B-416C6F65EA50⊙if is a very good point
01:22
<Hixie>
vlad_: dunno
01:22
<vlad_>
and potentially some way to easily put in normal-looking ascii math (perhaps with some script)?
01:22
<Hixie>
yeah, i really don't want to require that people write out raw mathml
01:22
<Hixie>
that's just insane stuff
01:22
<vlad_>
like, being able to type x^2/y^2 and have it do the right thing would be nice
01:22
<vlad_>
even if it's something like <span class="math">x^2/y^2</span> and then there's some JS that translates
01:23
<othermaciej>
JS translating latex math expressions to mathml would probably be doable
01:23
<Hixie>
i'd rather not rely on JS here
01:23
<Pavlov_>
vlad_: thats a font feature
01:23
<Hixie>
but that's another issue
01:26
<Philip`>
vlad_: The OO.o equation editor produces MathML and could be considered useful
01:27
<BenMillard>
hixie, the headings version seems a more conventional wiki page, to me.
01:27
<Hixie>
BenMillard: i'm not a conventional kind of guy :-)
01:28
<vlad_>
Pavlov_: e.g. http://www.zvon.org/HowTo/Output/MathMLtestsuite_abs1.php
01:29
<vlad_>
is a pretty crappy way to write |H/K| (granted, you also want italic H and K, but still)
01:29
<Pavlov_>
well
01:29
<Pavlov_>
its not great
01:30
<Pavlov_>
maybe some simple stuff would be OK
01:31
<BenMillard>
feedback on HTML5 from a friend I knew when I did GTA modding some years ago: "Damn you guys are fuckin smart"
01:32
<BenMillard>
I guess a poor first impression of the HTML5 effort is not a universal experience
01:32
<Hixie>
heh
01:33
Philip`
also happened to do GTA modding some years ago, where 'some' = 9
01:33
<BenMillard>
ooh, I modded the early ones too :)
01:34
<Hixie>
speaking of GTA, is GTA4 out yet?
01:34
Hixie
has some cash earmarked for a PS3 and GTA4 and is just waiting
01:37
<BenMillard>
http://www.rockstargames.com/IV/events/ - 29th April 2008
01:37
Hixie
calendars it
01:38
<Philip`>
http://web.archive.org/web/20020121215533/http://gta2.games.ultrastore.com/ - yay, my site sort of still exists
01:39
<BenMillard>
OMG I REMEMBER THAT WEBSITE
01:41
<BenMillard>
Philip`, this is a muliplayer level I made for GTA2: http://projectcerbera.com/gta2/multislayer/
01:41
<Philip`>
Unfortunately I barely knew what I was doing, and didn't know how to write GUI programs in anything other than Visual Basic
01:42
<BenMillard>
I still don't know how to write GUI programs in anything other than VB :$
01:42
<Philip`>
Later I learnt how to write GUI programs in Perl, which wasn't really any better
01:43
<Hixie>
heh
01:44
<Philip`>
(The only ones I've liked were in C++ / wxWidgets)
01:44
<Hixie>
delphi baby
01:45
<othermaciej>
I like to write GUI programs in HTML+CSS+JS
01:46
<othermaciej>
everything else seems annoying
01:46
<othermaciej>
(though ObjC+Cocoa is not so bad as such things go either)
01:46
<vlad_>
othermaciej: seriously.. though HTML could use some love
01:46
<Philip`>
BenMillard: The world looks oddly familiar there - the textures and shapes are all familiar, but I got bored and moved away before the map editor was released so the perspective is odd
01:46
<vlad_>
(cough xul box model cough)
01:46
<othermaciej>
vlad_: we got your xul box model
01:47
<othermaciej>
I wish the CSS WG would show it some love
01:47
<vlad_>
yeah
01:47
<othermaciej>
instead of figuring how to full-justify arabic text in the presence of chinese vertical writing that uses double-width characters
01:47
<Hixie>
all it would take is an editor
01:48
<vlad_>
othermaciej: we should really fedex those people a nice pen and some parchment
01:48
<Philip`>
othermaciej: HTML+CSS+JS doesn't work so well when you're trying to decode and render graphics from complex binary files on the user's computer, and trying to do it fast
01:48
<vlad_>
and they can have at it
01:48
<othermaciej>
Philip`: oh, well, I wouldn't use it for the non-GUI part
01:50
<Philip`>
othermaciej: Do you mean you'd have a non-GUI part in a proper language (C++ or Python or whatever) and expose a web interface to it for all interaction?
01:51
<othermaciej>
if I were called on to write a GUI app I would probably do a desktop/web hybrid with most of the UI in HTML and compute-intensive or system-integrating code in a lower level language
01:52
<BenMillard>
Philip`, yeah seeing proper perspective in a 2.5D sprite-based game is pretty odd :)
01:53
<BenMillard>
Philip`, I remeber using M1win for GTA1, which had no draw distance limit. that was cool
01:54
<BenMillard>
Philip`, the textures for MultiSlayer are actually from the Vice City level in GTA1
02:02
<dglazkov>
hey enough reminiscing :) I have a question about the state of the HTML5 spec
02:03
<dglazkov>
would you say that HTML5 is an "upcoming spec"?
02:03
dglazkov
is looking for shiny words
02:04
<Hixie>
upcoming?
02:05
<dglazkov>
yes, as in "of the upcoming HTML5 specification"
02:06
<dglazkov>
would "HTML5 specification draft" be better?
02:06
<Hixie>
if it's not already "upcome", for lack of a better term, i wonder what it will take for it to do so :-)
02:06
<Hixie>
what's the whole sentence?
02:06
<Hixie>
i would just say "HTML5"
02:06
<Hixie>
but that's just me
02:18
dglazkov
thinks the peeps here are awesome
02:27
<BenMillard>
oh crikey it's 2:30am
02:45
<BenMillard>
Marco Zehe blogged about labelledby and describedby. I chatted to him in Mozilla's #accessibility channel and blogged an HTML4 alternative: http://projectcerbera.com/blog/2008/03#day24
04:24
<Hixie>
hsivonen: yt?
04:24
<Hixie>
i don't understand your use case in http://www.w3.org/mid/3733150B-0B02-496D-8517-A9A6D040B3E8⊙if
04:24
<Hixie>
can you give an example of a page doing something like that?
04:26
<Hixie>
in particular, i don't understand why <img src="icon.svg" alt=""> isn't the best solution for that (even in xhtml today)
04:34
<Hixie>
hm, people want to be able to extract svg snippets from html documents and edit them in svg editors
05:23
<Hixie>
i am seriously considering not individually replying to all these e-mails
05:23
<Hixie>
even the ones sent to whatwg
05:23
<Hixie>
there is so much bikeshedding in these threads it's insane
05:23
<Hixie>
especially since what we're actually talking about here is the nuclear plant
05:53
<Hixie>
holy crap
05:53
<Hixie>
Philip` provides the first really detailed problem description: http://www.w3.org/mid/47DC70FB.1060205⊙cau
05:56
<Hixie>
ok i've read every e-mail
05:56
<Hixie>
http://wiki.whatwg.org/wiki/New_Vocabularies summarises what i found
08:42
<hsivonen>
Hixie: http://lists.w3.org/Archives/Member/w3c-wai-pf/2008JanMar/0507.html
10:48
<jgraham__>
re: typing <div class="math">latex</div> http://www.math.union.edu/~dpvc/jsmath/
10:50
<hsivonen>
someone who knows HTML5 drag&drop might want to look into the ARIA drag&drop stuff
14:10
<hsivonen>
Hixie: the icon case is basically Sam's blog
14:11
<hsivonen>
Hixie: reasons for not using <img>: 1) Fallback behavior. 2) manageability as part of the markup text in a CMS.
16:50
<andersca>
Hixie?
17:32
<Hixie>
hsivonen: i honestly think that sam's blog would be better off using <img>
17:32
<Hixie>
andersca?
17:33
<andersca>
Hixie: hey
17:33
<andersca>
Hixie: about the cache update process
17:33
<andersca>
Hixie: is the order of the files in the file list important
17:33
<andersca>
Hixie: (the file list described in step 13)
17:38
<Hixie>
um
17:39
<Hixie>
it just says "For each URI in file list" on step 20, without specifying order, so i guess it doesn't matter
17:41
<Hixie>
the only time order matters is during parsing
17:41
<Hixie>
because it affects duplicate elimination in the fallback mode
17:41
<andersca>
yeah
17:56
Hixie
tries to decode "Content-reinforcing contextual progressive-enhancement eye-candy"
17:57
<Hixie>
i really have no idea what that means
17:57
<andersca>
good luck with that ;)
17:57
<Hixie>
maybe hsivonen is just again referring to sam's icons
18:04
<Lachy>
Hixie, yeah, Sam's icons would be a good example of that. I think it means content that is related to the surrounding content, but doesn't provide any additional information, and is done using progressive enhancement techniques
18:04
<Hixie>
yeah so i really don't see why you wouldn't want to use <img> for that
18:05
<Hixie>
duplicating the icon every time it is used seems like an anti-pattern
18:05
<Hixie>
bad for caching, bad for maintainability, etc
18:08
<Lachy>
yeah, I don't know why <img> wouldn't be better either, but I don't know why Sam does it the way he does
18:08
<Lachy>
maybe SVG in <img> isn't as widely supported yet
18:09
<Hixie>
that's obviously not a relevant argument when discussing new features in a spec with an ETA of more than 10 years from now :-)
18:09
<Philip`>
SVG in <img> only works in Opera 9.5, so it's useless in practice
18:10
<Hixie>
i thought i heard it worked in webkit now too
18:10
<Philip`>
Oh, okay, that could be true
18:10
<Philip`>
but if it doesn't work in Firefox, it's still useless in practice
18:11
<Hixie>
well sure
18:11
<Hixie>
but like i said
18:11
<Hixie>
not a problem when we're discussing future technologies
18:12
<Lachy>
yeah, I know. It's not an argument, merely a possible explanation
18:18
<Hixie>
http://wiki.whatwg.org/wiki/New_Vocabularies updated
18:18
<Hixie>
lots more on there now
18:23
<Hixie>
how the hell do we solve the custom-data-in-html issue
18:26
<Philip`>
Namespaces!
18:27
<Philip`>
Or make any element name and attribute name conformant
18:31
<Hixie>
"namespaces" would just lead to people taking divitis to the next level
19:13
<Lachy>
hmm. even http://www.dreamhoststatus.com/ is down, that probably means dreamhost is having serious problems right now
19:13
<Lachy>
that sucks, cause I even log into my email right now
19:15
<Lachy>
allowing any element or attribute name to be conforming would make the language less useful and prevent future standardised extensions
19:15
<virtuelv>
Lachy: dreamhoststatus is offsite
19:17
<virtuelv>
should be noted that there is a conflict between Cogent and TeliaSonera that may affect connectivity for you
19:19
<Lachy>
who are Cogent and TeliaSonera?
19:19
<Lachy>
but the problem is apparently with dreamhost, they wrote:
19:19
<Lachy>
"Sorry about the downtime. A lot of you have already been receiving blingy status messages. Blingy is your cluster of machines for web, mysql, and email. The main file server for this cluster is having serious problems."
19:24
<Hixie>
dreamhost is fine for me
19:24
<Hixie>
i'm editing through it right now
19:24
<Hixie>
the status page does seem down though
19:26
<virtuelv>
Lachy: backbone providers in the US and scandinavia, respectively
19:30
<Lachy>
my site is still up, it's just my email having problems
19:31
<Lachy>
I'm probably also on a different server from you
21:27
<Hixie>
othermaciej: that use case is the first one people have presented that actually would require just "raw xml"-equivalent support in text/html
21:29
<othermaciej>
Hixie: is that because XBL2 inherently requires raw XML?
21:29
<Hixie>
not so much raw XML, but it allows you to do things like have <option> elements outside of <select>s and stuff like that
21:29
<Hixie>
since the DOM is built up in parts
21:30
<Hixie>
so the parsing rules for html content end up being different than in normal html
21:31
<Hixie>
wtf, my opera build always starts up in a "wedged" state where it won't access the web for 5 minutes
21:32
<virtuelv>
Hixie: huh?
21:32
<Hixie>
when i launch opera, and tell it to go to a web site, it just sits there spinning
21:32
<virtuelv>
using feeds?
21:32
<virtuelv>
mail?
21:32
<Hixie>
only after several minutes (and closing and reopening the window) does it go there
21:32
<Hixie>
web
21:32
<virtuelv>
9.5 or a prior version?
21:32
<Hixie>
just normal web
21:33
<Hixie>
9.50 Beta 4681
21:33
<virtuelv>
there is a vps folder in your profile directory
21:33
<virtuelv>
try nuking it
21:34
<virtuelv>
but hang on, I'll hear with some of the mac users
21:35
<Hixie>
i nuked Library/Application Data/Opera* and it didn't solve the problem
21:36
<virtuelv>
sounds omnious
21:37
<virtuelv>
I don't use mac, and have never seen similar behavior on other platforms
21:37
<virtuelv>
and short of actually upgrading to a later weekly, I'm afraid I can't offer more advice than "file a bug"
21:41
<Hixie>
is there a later weekly yet?
21:41
<Hixie>
i can never find them
21:42
<Hixie>
i wish opera had something like nightshift, or even better, like firefox's autoupdate
21:43
<virtuelv>
http://my.opera.com/desktopteam/blog/
21:43
<virtuelv>
latest post has newer builds
21:44
<virtuelv>
build 4716
21:44
<Hixie>
k i'll see if i can remember to update next time i try to test opera
21:44
<Hixie>
thanks
21:44
<virtuelv>
latest builds have a bunch of acid3 fixes
21:45
<Hixie>
cool
21:48
<Hixie>
same problem
21:55
<virtuelv>
heh, that's ... odd
21:56
<virtuelv>
noone I spoke to had heard anything about anything like this either
22:02
<hsivonen>
Hixie: yes, I referred to Sam's icons again with reading the irc log first
22:28
<Hixie>
woot, the whatwg was cited in a paper: http://www.cs.uic.edu/~mter/htmi-xss.pdf
22:29
<met_>
Hixie, favicon http://acid3.acidtests.org/favicon.ico is only joke or is it real part of Acid3?
22:31
<Hixie>
there is no http://acid3.acidtests.org/favicon.ico
22:31
<Hixie>
that's a 404
22:32
<Hixie>
if you see that used as a favicon.ico, it's a bug
22:32
<vlad_>
not a 404 for me, unless your 404 serves a red image with a hand-drawn lion-thing
22:32
<annevk>
it does
22:32
<Hixie>
it does
22:32
<annevk>
:p
22:33
<Hixie>
hey anne
22:33
<Hixie>
nice vacation?
22:33
<annevk>
yeah, lots of snow :)
22:33
<Hixie>
cool
22:33
<annevk>
heading to San Diego now
22:33
<met_>
Hixie, thx misses 404 there, so bug in Firefox
22:33
<annevk>
CSS WG meeting starts tomorrow...
22:36
<Hixie>
annevk: enjoy
22:36
<annevk>
oh yes :)
22:37
om_afk
wonders how the exciting charter discussion will go
22:37
<Hixie>
have you created yourself a bingo card yet? with squares like "fantasai brings up an irrelevant point" and "bert says javascript is irrelevant to the web" and so on?
22:37
annevk
too
22:37
<annevk>
lol
22:39
<Hixie>
gsnedders: actually html5 is already under feature freeze :-)
22:39
<Hixie>
has been since late last year
22:39
<Hixie>
it's the main reason i didn't add workers
22:41
<tommorris>
does feature freeze mean that if an existing html4 behaviour has been removed it cannot be reinstated?
22:41
<Hixie>
no
22:42
<Hixie>
e.g. headers="" was added last week
22:47
<gsnedders>
Hixie: how solid feature freeze, though? We don't have SVG or MathML or the like in HTML 4/5, for example
22:53
<Philip`>
Gmail really isn't very good at keeping track of unread message counts when a message has multiple labels
23:03
<Hixie>
gsnedders: feature freeze with the caveat that issues that had been raised before the feature freeze would still get resolved
23:16
tommorris
wonders whether the lack of compatibility with the GRDDL specification counts as a raised issue
23:24
jgraham
guesses phillips problem with gmail is less annoying than the random number that thunderbird occasionally displays over the dock icon on OSX
23:24
<jgraham>
s/phillips/Philip`'s/
23:25
<jgraham>
Presumably they are supposed to be unread message counts but random numbers might work just as well