02:26
<erlehmann>
mimesniff now can into text/html http://news.dieweltistgarnichtso.net/bin/mimesniff.html
02:26
<erlehmann>
regarding features, it is only missing the bitmask.
02:26
<erlehmann>
anyone having an idea how to do this in shell?
04:21
<terinjokes>
Domenic: should I bother IDLing any of this?
04:21
<terinjokes>
getting the hang of the document formatting
04:21
<caitp->
console is an interface, so you probably do want to IDL it
04:22
<terinjokes>
wonder if my choice of "any" for most of the arguments is acceptable
04:33
<terinjokes>
and does specifying a type as "array of arrays" even make sense?
04:50
<terinjokes>
(i ask because it doesn't seem possible to inlcude such a thing)
04:56
<caitp->
you might be able to do (type[])[]
04:56
<caitp->
I'm not sure if anyones idl compilers would need updates for that
04:56
<caitp->
well
04:57
<caitp->
it would be a syntax error i think
05:00
<terinjokes>
also thinking about other ways to write out the IDL
05:03
<caitp->
in blink, the idl uses extended attributes that don't exist in the spec, but it's probably the only way you can really do it
05:14
<terinjokes>
yeah, i'd noticed the extended stuff when hunting down docs
07:31
<annevk>
terinjokes: yeah, for the console API I expect most of it to take any
07:31
<annevk>
caitp-: http://url.spec.whatwg.org/#application/x-www-form-urlencoded is the query format
07:32
<terinjokes>
annevk: yep, that's what i've got so far
07:32
<caitp->
that's not "the query format", because it leaves many things unspecified (which is where the whole "jQuery style vs PHP vs ...." nonsense comes from, which is driving me crazy on bug trackers and why am I still awakeohgod)
07:33
<annevk>
terinjokes: what's mostly important about this API spec is that it clarifies what kind of methods are exposed
07:33
<caitp->
rules for interpreting things like repeated query strings, or query[], or query[a], etc
07:33
<caitp->
that stuff is all totally unspecified anywhere
07:33
<annevk>
caitp-: there are no rules for those
07:33
<annevk>
caitp-: that's application-layer stuff
07:33
<caitp->
exactly, and that is what sucks
07:33
<annevk>
meh
07:34
<caitp->
for one thing if there were rules, URLUtils's search object would be a lot more useful
07:43
<Ms2ger>
if (navigator.userAgent.match('Gecko/(\\d+)') && RegExp.$1 == '20060217' && RegExp.$1 != '00000000') {
07:43
<Ms2ger>
Guess the site
07:48
<annevk>
Ms2ger: google.com
07:49
<Ms2ger>
Actually, the live dom viewer
07:49
<annevk>
Now I'm curious :-)
07:50
<annevk>
I don't really see what difference that causes
07:51
<Ms2ger>
Oh, I guess the tree view styles were really broken in Gecko back then
13:51
<erlehmann>
i think i understand the pattern mask
13:51
<erlehmann>
>>> chr(0x41), chr(0x61), chr(0xDF & 0x41), chr(0xDF & 0x61)
13:51
<erlehmann>
('A', 'a', 'A', 'A')
13:51
<erlehmann>
Hixie_ is the mime sniff algorithm normative or descriptive? i may be trying to detect text/tab-separated-values
13:53
<erlehmann>
oh, you are not the current editor
13:53
<erlehmann>
whom of you is Gordon P. Hemsley <me⊙go>?
13:54
<erlehmann>
anyone knows why the mp4 matching is so weird? http://mimesniff.spec.whatwg.org/#signature-for-mp4
14:31
<annevk>
JakeA: dude, https://twitter.com/jaffathecake/status/508258170398343168 is so true
14:31
<annevk>
JakeA: one of the reasons I'm afraid of background updates going wrong
14:32
<JakeA>
G+ did the same until a couple of months ago
14:32
<annevk>
erlehmann: don't do MP4 :-)
14:32
<JakeA>
I guess it's still an area of UX newness
14:32
<annevk>
erlehmann: algorithms are meant to be normative however
14:32
<JakeA>
(my trains demo did it too, just fixing it up now)
14:33
<erlehmann>
annevk i probably will not do mp4, yes. do you have any idea how to do the bitmasking in shell script, btw? http://news.dieweltistgarnichtso.net/bin/mimesniff.html
14:33
<erlehmann>
thx
14:34
<erlehmann>
“Hey I've got some new data, you're probably not reading the stuff currently on screen right?" *replaces everything*” is one of the reasons i seldom allow js for social media sites
14:35
<erlehmann>
whomever came up with that (and infinite scrolling, hey, lets allocate memory forever) needs to be subjected to the savage beatings antipattern
14:36
<annevk>
erlehmann: I can recommend stackoverflow :-)
14:37
<erlehmann>
standard stackoverflow or common stackoverflow? ;----DDDDD
14:37
<erlehmann>
they should rename it to common stack!
14:43
<Ms2ger>
Tsk tsk
14:43
<Ms2ger>
We don't condone violence here
14:53
<erlehmann>
Ms2ger http://cssquirrel.com/blog/2012/10/01/the-savage-beatings-anti-pattern/
14:53
<erlehmann>
;)
17:55
<erlehmann>
mimesniff(1) can now do pattern masking http://news.dieweltistgarnichtso.net/bin/mimesniff.html
17:56
<erlehmann>
i must admit it is a bit slow
18:48
<Ms2ger>
Is that seriously written in shell?
19:14
<Ms2ger>
Nice: http://www.w3.org/mid/E1XPZQR-0002NO-U7⊙di
19:41
<ericandrewlewis>
does the whatwg or w3c define html5 spec?
19:41
<boogyman>
neither, because there is no "HTML5"
19:43
<ericandrewlewis>
there's just HTML as far as whatwg is concerned?
19:44
<boogyman>
HTML is a living document.
19:44
<caitp>
people do call it html5 though, whether they're talking about 2009 or 2015
19:45
<ericandrewlewis>
w3c thinks that there's an html5 tho
19:49
<Ms2ger>
The W3C HTML5 and 5.1 documents are both more and less out-of-date forks of the HTML Living Standard
19:51
<ericandrewlewis>
cool
19:51
<ericandrewlewis>
well, not cool, but thanks for the info :)
19:51
<Ms2ger>
Np
19:57
<smaug____>
caitp: "html5" is a buzzword and under that term there are plenty of other specs than just HTML. All the CSS stuff and XHR etc.
19:57
<caitp>
that's how the public recognizes it
19:57
<smaug____>
(mostly thanks to Google's pr )
19:58
<caitp>
and it's perfectly fair, because you're seldom going to see one without the other
20:03
<smaug____>
yeah, I guess so
20:04
<smaug____>
it is just confusing to have HTML5 spec, HTML spec, HTML 5.1 and then the HTML 5 which contains more than just HTML