00:00
<dekiss>
yeah I know but here I am not sure what 'newline character' is : https://developer.mozilla.org/en-US/docs/Web/CSS/white-space
00:00
<dekiss>
thank you for help
00:00
<dekiss>
Hixie nice live DOM tree viewer if you made it ^^ :) http://software.hixie.ch/utilities/js/live-dom-viewer/#
00:00
<dekiss>
good job
00:01
<Hixie>
thanks
00:02
<dekiss>
I am not sure why but I have set white-space: pre-wrap; on textarea and when I press enter and when I put more white spaces with spacebar nothing hapens
00:02
<slowhands>
it would be really cool if you integrated it with jsfiddle or gists
00:02
<Hixie>
not gonna happen :-)
00:02
<dekiss>
in paragraph sorry not textarea
00:04
<slowhands>
it would be cool. but thats ok
00:05
<dekiss>
I am sorry I missed the selector -_-
00:06
<slowhands>
i didn't realize you were referring to form inputs
00:07
<dekiss>
yaa word-space ftw!
00:07
<dekiss>
white* haha
00:07
<dekiss>
omg I need one episode of star trek to make little pause ^^
00:11
<dekiss>
which is best element for making structure of webpage? I use section element as generic element to divide my webpage into sections I have a lot of section elements is this bad? any idea what to use if not section? div will be better? or it doesn't matter so much?
00:14
<slowhands>
no not bad
00:14
<slowhands>
but older browsers will bug out
00:14
<slowhands>
you will need the shiv/shim
00:14
<dekiss>
ok thanks
00:14
<slowhands>
the typical hierarchy is sections then articles
00:15
<slowhands>
divs for more unconventional things
00:15
<slowhands>
try to keep it semantic
00:15
<slowhands>
with nav and footer
00:15
<slowhands>
a good resource is CodeSchool. I really like their html5 course. its really interactive and pretty fast
01:59
<dekiss>
which is newline character in html?On deciaml html unicode code unit of enter key I don't get new line
02:56
<dekiss>
YEEEEEEEEE
16:40
<marcdm>
hey all
16:41
<marcdm>
I have some questions regarding html5lib-python, anyone want to help me out?
16:46
<Ms2ger>
You can always try
17:20
<SimonSapin>
marcdm: just ask, someone might answer later
17:21
<marcdm>
it's ok. I think I posted it to the mailing list. If it doesn't go thru (I don't see it yet), I'll post an issue on github
17:22
<marcdm>
it's about the treewalker/_base.py. it does isinstance(name, six.text_type) where I think it should be isinstance(name, six.string_types)
17:48
<GPHemsley>
Ms2ger: Did you see all my babbling from yesterday after you left?
17:48
<GPHemsley>
about html5lib and anolis
17:49
<Ms2ger>
Attribute order? I don't quite remember what the story was there
17:49
Ms2ger
looks
17:51
<Ms2ger>
http://code.google.com/p/html5lib/issues/detail?id=153
17:52
<GPHemsley>
https://github.com/html5lib/html5lib-python/issues/37
17:53
<GPHemsley>
but AFAICT, attributes are not actually maintained in their original order
17:54
<GPHemsley>
(and there appear to be 4 different commits associated with that bug that presume to do the same thing)
17:54
<GPHemsley>
oh, I should say that I installed the git html5lib using the instructions on the wiki
17:55
<GPHemsley>
and I didn't get the dependencies I needed
17:55
<GPHemsley>
so it's possible that I missed others
17:56
<GPHemsley>
"sudo python setup.py install" apparently doesn't use the requirements.txt file(s)
17:58
<gsnedders>
GPHemsley: Basically no Python document tree API preserves attribute order
17:58
<gsnedders>
GPHemsley: So what html5lib does is basically irrelevant
17:59
<GPHemsley>
gsnedders: Are you sure that your fix made things better and not worse? At least sorting them alphabetically maintains a predictable order.
17:59
<gsnedders>
GPHemsley: There's a filter that gives the old behaviour
18:00
<gsnedders>
GPHemsley: Of giving alphabetical order.
18:00
<GPHemsley>
gsnedders: How does that work? (Or does Ms2ger already know?)
18:00
<Ms2ger>
I don't
18:00
<Ms2ger>
I don't
18:01
<GPHemsley>
because this breaks the tests for anolis, and will likely cause unnecessary churn in specs (IIUC)
18:01
<gsnedders>
GPHemsley: The point is given a tree API that does preserve order, html5lib should roundtrip correctly. This means we can't blindingly reorder attributes.
18:02
<GPHemsley>
gsnedders: Well, I don't know the particulars, but it doesn't seem like whatever anolis is using is preserving order, either alphabetical or given
18:02
<gsnedders>
Probably we should make the filter have a native API in the serializer much as inject_meta_charset, etc., is
18:03
<gsnedders>
GPHemsley: Right, given a lot of tree APIs in Python you will get no defined order (in practice, it's the hash of a dict key)
18:03
<gsnedders>
GPHemsley: Right, given a lot of tree APIs in Python you will get no defined order (in practice, it's the hash of a dict key)
18:04
<GPHemsley>
gsnedders: So are you saying that you removed the alphabetical sorting from html5lib without replacing it with something that preserves order? (I don't really know how html5lib works.)
18:05
<gsnedders>
GPHemsley: By default, yes. (You have to realize html5lib doesn't have any tree format of its own — it just uses various existing XML toolchains, and they're what don't preserve order.)
18:06
<gsnedders>
GPHemsley: Optionally, you can preserve order. We do this for the html5lib test suite, for obvious reason (i.e., determinism).
18:06
<GPHemsley>
gsnedders: So you have to tell html5lib which tree format to us?
18:06
<gsnedders>
GPHemsley: Optionally, you can preserve order. We do this for the html5lib test suite, for obvious reason (i.e., determinism).
18:06
<gsnedders>
GPHemsley: Yes.
18:06
<GPHemsley>
Ms2ger: So I guess we need to do this?
18:07
<gsnedders>
GPHemsley, Ms2ger: can one of you file a bug on the AlphabeticizeAttributes filter not having a native API like inject_meta_charset?
18:07
<gsnedders>
Because that is an oversight.
18:07
<GPHemsley>
Ms2ger: I'll leave that to you, as I assume you have a better knowledge of the inner workings of html5lib.
18:07
<gsnedders>
(html5lib.serialize(my_document, alphabeticize_attributes=True) should work as an API basically)
18:08
<gsnedders>
Rather than having to use the long-hand API.
18:08
<gsnedders>
marcdm: Oh, someone mentioned that on IRC before.
18:08
<gsnedders>
marcdm: File a bug, as that probably shouldn't be the behaviour on Python 2.
18:08
<GPHemsley>
gsnedders: As an aside, do you have any idea why I wouldn't have gotten the right dependencies installed when I installed html5lib per the WHATWG wiki article?
18:09
<gsnedders>
What wiki article?
18:09
<GPHemsley>
sudo python setup.py install
18:09
<GPHemsley>
err
18:09
<GPHemsley>
ugh
18:09
<Ms2ger>
GPHemsley, that's an assumption I'm not sure is warranted :)
18:09
<GPHemsley>
http://wiki.whatwg.org/wiki/Anolis
18:09
<GPHemsley>
Ms2ger: Well, I defer to you anyway ;)
18:10
<Ms2ger>
I don't want to think about attribute order tonight, file an issue? :)
18:10
<gsnedders>
Why does that wiki article not just say to use pip for all of them?
18:10
<GPHemsley>
gsnedders: Probably for historical reasons
18:10
<GPHemsley>
gsnedders: (Which is not to say it shouldn't now.)
18:10
<Ms2ger>
pip? What's that? ;)
18:11
<Ms2ger>
And in particular, that'd require me to upload anolis to pypy rather more often, I'd think
18:11
<gsnedders>
Ms2ger: I meant for the dependencies.
18:11
<Ms2ger>
Python dependencies make me cry
18:12
<gsnedders>
Ms2ger: Do what the hell you want for Anolis, that's your project. ;P
18:12
<Ms2ger>
If you mean for lxml/cssselect, I blame... tantek?
18:12
<GPHemsley>
gsnedders, Ms2ger: https://github.com/html5lib/html5lib-python/issues/62
18:12
<gsnedders>
For html5lib it says you have to use it from git. I wonder why.
18:13
<gsnedders>
Did Anolis not work with 0.95?
18:13
<Ms2ger>
It did
18:13
<gsnedders>
GPHemsley: Cheers.
18:13
<Ms2ger>
(It used to refer to google code)
18:13
<gsnedders>
Now really I'm going away.
18:14
Ms2ger
waves
18:16
<GPHemsley>
Ms2ger: So now what do we do with this order filter thing?
18:17
<Ms2ger>
File an Anolis issue? :)
18:17
<marcdm>
gsnedders: issue #63 submitted on github
18:31
<GPHemsley>
Ms2ger: https://bitbucket.org/ms2ger/anolis/issue/16/force-sorted-attributes
18:41
<Ms2ger>
Thanks
22:43
<gsnedders>
jgraham: https://critic.hoppipolla.co.uk/r/103 has stopped tracking GitHub when I pushed a rebase (that I told Critic about) to GH.
22:50
<gsnedders>
jgraham: "Refusing to perform a review rebase via an automatic update"
22:50
<gsnedders>
jgraham: But I can't manually update it.
23:59
<marcdm>
gsnedders: not only did I post a bug report, I submitted a pull-request as well