#microformats 2018-05-10

2018-05-10 UTC
[jgmac1106], [eddie], webchat75, [miklb] and chrisaldrich joined the channel; webchat75 left the channel
[eddie] and KartikPrabhu joined the channel
tglobe, [kevinmarks], KartikPrabhu, wakest, pniedzielski[m] and schmarty joined the channel
#
@jgmac1106
@protohedgehog Keeping HTML template simple. I am going to add a #microformats2 and other #indieweb support so one day we can all hook into an #OpenScience reader or just connect our #OpenScience feeds to some of the cool readers I know will be emerging soon.
(twitter.com/_/status/994551149234503680)
[jgmac1106], [kevinmarks], nitot and tglobe joined the channel
barpthewire, webchat254, iwaim, KartikPrabhu, [schmarty] and [kevinmarks] joined the channel
#
[kevinmarks]
That is actually about microformats, mf1 but still.
j12t, j12t_ and [manton] joined the channel
#
Zegnat
So ... anyone saved manton’s before and after HTML? Or did he not tweak it yet?
#
aaronpk
but paste that into XRay and it's wrong
#
gRegorLove
I just manually removed the <span class=p-name>
#
Zegnat
That’s not wrong though
#
aaronpk
it's not?
#
aaronpk
why would the src attribute ever make sense in the "name" value?
#
gRegorLove
No alt=""
#
Zegnat
Exactly
#
Zegnat
images are replaced by their alt, or their href, in that order, when plain text is extracted
#
sknebel
yeah, 0.4.4 is what I expected per the rules on parsing-specification, so when you said newer parsers fixed it I couldn't work out where that was supposed to be fixed
#
sknebel
the one on python.microformats.io doesn't
#
aaronpk
yeah i got it backwards apparently. newer parsers introduced that behavior. i forgot xray was running bleeding edge
#
Zegnat
So all is fine, [manton] should probably add alt="" to that image so it isn’t read out (and ignored in mf2)
#
aaronpk
but still, why would the src value ever make sense for the value of the name?
#
KartikPrabhu
what do browsers do?
#
KartikPrabhu
or screen readers?
#
Loqi
[Zegnat] #17 Define removal of SCRIPT and STYLE elements everywhere textContent is requested.
#
Zegnat
It doesn’t make sense for a name. But the parser doesn’t know if it is parsing p-name or, say, p-content. It is just doing the p- (plaintext) rules.
#
aaronpk
but this is the implied name
#
aaronpk
so it does know it's the name
#
Zegnat
And for content it may make sense not to have all the image data disappear
#
Zegnat
Yeah, the issue sknebel linked changed it for implied name
#
Zegnat
Basically implied name parsing was made the same as p-name parsing.
#
Zegnat
It may be worth opening a new issue specifically to not replace <img> elements in the case of implied names? Or maybe do replace them with alt but never with href?
[jgmac1106], [chrisaldrich], [schmarty], KartikPrabhu and [tantek] joined the channel
#
[tantek]
Or src?
vivus and [kevinmarks] joined the channel
#
Zegnat
Yes, good catch, where I wrote href I meant src
[schmarty], KartikPrabhu, [chrisaldrich], [kimberlyhirsh], hober, [kevinmarks] and [tantek] joined the channel; webchat254 left the channel