01:43
<MikeSmith>
per the URL spec, is scheme data not allowed to contain "<" and ">"?
01:43
<MikeSmith>
since they are not "URL code points"
08:01
<annevk>
Did anyone else notice Web Crypto introduces https://dvcs.w3.org/hg/webcrypto-api/raw-file/tip/spec/Overview.html#big-integer
08:01
<annevk>
Seems odd
08:04
<MikeSmith>
annevk: they should use something else instead?
08:04
<annevk>
MikeSmith: we don't have big integers, but I have the feeling they should coordinate with TC39
08:05
<annevk>
MikeSmith: been reviewing Web Crypto's IDL a bit this morning; wanted to watch a TV series since it's a holiday, but got distracted
08:05
<MikeSmith>
annevk: yeah they don't seem to be doing that
08:05
<MikeSmith>
holiday?
08:06
<annevk>
MikeSmith: https://www.gov.uk/bank-holidays
08:06
<MikeSmith>
ah
08:11
<MikeSmith>
annevk: is "1:1" a valid URL?
08:11
<annevk>
MikeSmith: scheme cannot start with a number iirc
08:13
<MikeSmith>
annevk: yeah but galimatias treats that as a relative URL, if the base URL is not null
08:13
<MikeSmith>
afaict
08:14
<annevk>
MikeSmith: that makes sense, an error should be logged for that somewhere though
08:14
<MikeSmith>
so it sees it as a path
08:14
<annevk>
MikeSmith: oh maybe it's error-free in that case, hmm
08:14
<MikeSmith>
annevk: I think per the current spec it's error-free
08:15
<MikeSmith>
it's only a parse error if the state override is set
08:15
<annevk>
seems fine to me, perhaps a tad confusing, but given that 1 can never be a scheme...
08:21
<MikeSmith>
annevk: for the case of the validator it seems like it would be nice to help users catch cases like "1http://example.com", but reporting an error for those
08:21
<MikeSmith>
annevk: hmm maybe I can report a warning for those cases
08:23
<annevk>
MikeSmith: so the problem is that it's a legitimate path
08:23
<MikeSmith>
right
08:23
<annevk>
MikeSmith: and that has not changed
08:25
<MikeSmith>
annevk: yeah it's just that I'd like to be able to help users catch cases of mistyped URLs
08:25
<MikeSmith>
annevk: but I guess we really can't for this case
08:28
<annevk>
MikeSmith: seems hard without the kind of computing power Google has for detecting typos and such
08:29
<annevk>
MikeSmith: flagging ":" if no scheme and no "/" came before it as potential problem might be enough though
08:30
<annevk>
MikeSmith: you'd have to do some telemetry to see how signal / noise plays out
08:30
<MikeSmith>
yeah
08:31
<MikeSmith>
annevk: I'll also ask zcorpan what he thinks
11:51
<IZh>
Hi.
15:00
<JonathanNeal>
Goodmorning (from PT)!
15:06
<Ms2ger>
Good evening
16:16
<SamB>
okay wtf google
16:17
<SamB>
what's so hard to understand about https://www.google.com/search?q=whatwg+ms-icon
16:17
<Ms2ger>
ms -icon
16:18
<SamB>
yeah I kind of gathered it thinks that but come on I can't be the first to google a hyphenated term
17:29
<Domenic_>
So W3C DOM Parsing & Serialization is just a fork of http://domparsing.spec.whatwg.org/ ? /cc Ms2ger
17:29
<Ms2ger>
Yes
17:29
<Domenic_>
How bad is the damage?
17:30
<Ms2ger>
I haven't had the time to follow the changes
19:06
<IZh>
Hi.
19:09
<TabAtkins>
IZh: http://www.nohello.com/
19:30
<IZh>
TabAtkins: I tried to be polite. ;-)
20:30
<SamB>
hmm, are the Author lines in mimesniff.git actually an accurate representation of authorship?
20:31
<IZh>
Hixie: hi.
20:38
SamB
grumbles about "open in new tab" not working for bookmarklets ...
21:09
<TabAtkins>
Man, I really wasn't kidding about nohello.
23:01
<SamB>
hmm, someone needs to replace the IANA registration for .ico files ... they're currently registered as http://www.iana.org/assignments/media-types/image/vnd.microsoft.icon :-(