05:59
<sideshowbarker>
jgraham: Sam Sneddon [:gsnedders] ↑ (above html5lib question)
07:12
<sideshowbarker>
Is anybody familiar with “provisional navigation”? Is it maybe a synonym for “provisional load”? I had thought “provisional navigation” might be some WebKit-specific thing — or maybe iOS-specific — but I now see mentions of it in the Blink sources also (in code comments).
09:39
<Noam Rosenthal>
zcorpan: after speaking with Domenic I'm leaning towards keep the current proposal using href.... it's not more confusing wrt the base element so consistent with the notion of an "internal link"
09:39
<Noam Rosenthal>
I think it's OK if people who use the base element need to take extra care with this new feature
09:40
<Noam Rosenthal>
.... also we'll support <a name> just as we do for other internal links
09:40
<zcorpan>
Noam Rosenthal:
09:40
<zcorpan>
Noam Rosenthal: OK
09:41
<Noam Rosenthal>
and thanks everyone for re-raising doubts about this when they come, it takes longer but I believe we'll end up with something that's thought through.
09:42
<zcorpan>
Language design for HTML is non-trivial
09:48
<Noam Rosenthal>
zcorpan: Yup.. learning a lot about it every day
15:00
<Sam Sneddon [:gsnedders]>

for typeshed it probably doesn't matter because the input stream should very much be an internal implementation detail.

but: it does seem like we no longer require it to be anything but Iterable[str], based on the implementation, albeit where each str must be len(s) == 1 (but AFAIK that's not expressible in the type system).