02:24
<MikeSmith>
TabAtkins: as far “authors must” being a conformance class you can’t actually test, that’s why it is better for such requirements to be stated as “documents must”, which can be tested (and which the HTML checker and CSS do actually check)
02:27
<MikeSmith>
but I think if a spec uses “authors must” it should also make clear what that means
02:27
<MikeSmith>
for example, the HTML says this:
02:28
<MikeSmith>
“Conforming documents are those that comply with all the conformance criteria for documents. For readability, some of these conformance requirements are phrased as conformance requirements on authors; such requirements are implicitly requirements on documents”… “For example, if a requirement states that "authors must not use the foobar element", it would imply that documents are not allowed to
02:28
<MikeSmith>
contain elements named foobar.”
02:28
<MikeSmith>
https://html.spec.whatwg.org/multipage/infrastructure.html#conforming-documents
02:30
<MikeSmith>
astearns: ⬆
02:40
<MikeSmith>
anyway, the HTML spec also does have a few cases of such “authors must” requirements that are not actually machine-checkable/testable, and that’s OK too
09:58
<hsivonen>
MikeSmith: I went with ハロー・ワールド: https://hsivonen.fi/rs/encoding_rs/#examples
09:59
MikeSmith
looks
09:59
<hsivonen>
the streaming example is hilariously verbose compared to the non-streaming one
09:59
<hsivonen>
streaming is hard
10:00
<MikeSmith>
heh
10:00
<MikeSmith>
I thought Rust made everything better!
10:00
<MikeSmith>
:)
10:01
<hsivonen>
MikeSmith: It does! After this, java.nio.charset feels sad.
10:01
<MikeSmith>
oh well yeah
10:01
<hsivonen>
MikeSmith: though that's partly because this API design is informed by what I perceive as design mistakes in java.nio.charset.
10:01
<MikeSmith>
not sure what else could be much worse
10:01
<MikeSmith>
ok
10:04
<hsivonen>
aside: I'm trying to migrate the backend of Validator.nu to Ubuntu 16.04 and IPv6 connectivity
10:04
<MikeSmith>
oh cool
10:04
<hsivonen>
for some reason, ip6tables doesn't appear to work symmetrically with iptables
10:05
<MikeSmith>
oh? glad I have not had to mess with it too much then
10:05
<hsivonen>
I guess I'll just omit the firewall
10:05
<MikeSmith>
I find that every time I need to use iptables I just have to re-read the docs anyway
10:06
<hsivonen>
I wonder how much the slowness of IPv6 deployment is about stuff being subtly weird compared to IPv4
10:06
<MikeSmith>
a lot I bet
10:07
<MikeSmith>
but there are many things slowing it down I think
10:07
<MikeSmith>
e.g., try accessing anything at *.whatwg.org through IPv6
10:08
<hsivonen>
does dreamhost not offer IPv6 or has whoever owns the whatwg.org server config not gotten around to listening on IPv6?
10:09
<MikeSmith>
Dreamhost apparently claims they have IPv6 working for but all my /etc/hosts on IPv6-environment machines that I want to get to whatwg.org stuff from, I have had to force use of the IPv4 address instead
10:09
<hsivonen>
validator.nu itself has been available over IPv6 for 3 weeks or so now
10:09
<MikeSmith>
ah OK
10:10
<MikeSmith>
the W3C vnu instances are all still running under IPv4 but that is a general issue with lack of support for IPv6 in all MIT CSAIL networks
10:11
<hsivonen>
hmm. I remember reading some IPv6 comments by an MIT network admin
10:11
<MikeSmith>
yeah
10:11
<hsivonen>
it was critital of IPv6, but I thought MIT had deployed it anyway
10:11
<MikeSmith>
they had tried but he had to roll it back
10:13
<MikeSmith>
I think one big problem was with Windows machines, dozens of them end up flooding the network with junk if are running in an IPv6 environment
11:46
<nimai>
I'm front-end developer. How can I contribute to HTML standard? Oh I read the faqs, but I seem to be lost in details
14:52
<annevk>
nimai: did you look at the GitHub repo?
14:52
<annevk>
nimai: the README has a few suggestions
14:53
<annevk>
nimai: if you remain unsure, a good first step might be to take some time and read through it, or parts of it
15:22
<nimai>
thank you annevk. I hope to find something to work upon after looking into the repo
15:27
<annevk>
nimai: cool, succes
21:35
<caitp>
what do web developers feel is missing from html?