15:49
<zell>
hello
15:51
<zell>
I got an error when I'm importing script from a worker using importScripts function: Script file not found: assets/js/gamecs.min.js using file:// or gh-pages
15:51
<zell>
is possible to use Web worker's importScripts from file:// ?
16:18
<hsivonen>
Hixie_: please let's not make things. let's keep iso-8859-1 as an alias of windows-1252. we already shipped that.
16:19
<hsivonen>
let's not make things complicated, that is
16:27
<jarek>
Hi
16:28
<jarek>
is there any hope that web components / custom elements / shadow specs will never be implemented by all browser vendors?
16:28
<jarek>
it looks like only Mozilla and Google are pushing it
16:29
<gsnedders>
So, what, will it get implemented by MS?
16:29
<zewt>
that doesn't sound like an "only", heh
16:29
<gsnedders>
It sounds like almost all browser engines nowadays.
16:30
<jarek>
I really hope that it will be blocked by MS because those specs are disastrous
16:30
<gsnedders>
How so?
16:31
<jarek>
gsnedders: it's horribly bloated, all we (developers) really need is a way to sandbox DOM subtrees for the rest of the page
16:32
<jarek>
shadow DOM is overkill
16:32
<jarek>
s/for/from
16:33
<gsnedders>
Last I looked it didn't do much more than that?
16:34
<jarek>
gsnedders: it does, it introduces syntax for custom elements and external template
16:34
<jarek>
and the <style scoped> is just ridicolous
16:34
<jarek>
it could make sense 10 years ago
16:36
<jarek>
are web components implemented in vanilla WebKit?
16:36
<gsnedders>
Mostly, yes.
16:37
<jarek>
we are doomed :/
16:37
<gsnedders>
It seems to be a supserset of what you want?
16:38
<gsnedders>
The shadow DOM is the only part that's pretty much stable — and that does everything you want, no?
16:39
<jarek>
gsnedders: let's say that I wanted to create "button" widget
16:40
<jarek>
if I understand it correctly, the fastest way to implement it with shadow dom would be to create new shadow dom root element (document.createShadowRoot())
16:40
<jarek>
and then insert all button elements (e.g. label, icon or tooltip) inside it
16:42
<jarek>
so I end up with something like <x-button><div class="button">...</div></x-button>
16:42
<jarek>
I would rather prefer to just define my button element right away, e.g. <div class="button" sandbox="true"></div>
16:43
<gsnedders>
Or use <shadow>
16:43
<gsnedders>
<shadow><div class="button" sandbox="true"></div></shadow>
16:43
<jarek>
what about styling? Everyone should switch to inline styles now?
16:44
<jarek>
I suspect I could put @import into scoped inline style, but this would be still ugly
16:44
<jarek>
even CSS namespaces (for XML) were more elegant than that
16:44
<gsnedders>
I'm still not convincd there's any clean way to do styles in shadow DOMs.
16:45
<jarek>
gsnedders: <div class="button" scoped-style="some-style.css"></div>
16:46
<jarek>
or just allow "scoped" attribute on <link rel="stylesheet">
20:20
<zewt>
dear chrome: when i change the filename in a save-as dialog to not have a .txt extension, don't ignore me and silently put it back. thanks -me
20:28
<pingveno>
I'm adding HTML 5 support to a Python library. For datalist tags with conflicting/duplicate ids, which one should I choose? The first one? The last one?
20:30
<wilhelm_>
pingveno: Are you sure you're not duplicating the work done here? https://code.google.com/p/html5lib/
20:33
<Ms2ger>
pingveno, what the spec says
20:41
<pingveno>
wilhelm_: No. I'm adding to Mechanize.
20:42
<pingveno>
It's a library that emulates how a browser behaves when it visits web sites, including filling out forms.
20:43
<pingveno>
I have to figure out the best way to handle an input referring to a duplicated datalist id.
20:46
<pingveno>
I can't find anything in the spec about how to resolve the problem.
23:11
<Hixie_>
pingveno: you mean when list="" attributes point to a <datalist> element?