Aha – since you’ve got Webmentions up, let’s do this!
Hey Jacob! I’m familiar with your website – I covered your linkroll in
Directory Uprising.
It’s really comforting to see you interested in projects like directories and
whostyles that aren’t necessarily protocols – which the Indieweb can get very
focused on. Whostyles are definitely a tough one to turn into a protocol – since
CSS evolves over time and it’s tough to know how to restrict the styling. (But
it’s also important bc perhaps you don’t want to load a bunch of whostyles that
blow up your site.)
Your introduction of all: revert is exciting – didn’t know about that!

My larger plan for this site involves a full comment moderation system, so I
already intend to read everything that people send me. Given the scale of my
site, and the relative geekiness of whostyles as a concept, I’m not too
worried about how many CSS rules I’ll have to manually review day-to-day. When
my webmention endpoint receives a webmention, it will sniff the source site
for a whostyle. If one is detected, it will be downloaded and presented to me
as a part of the comment moderation process. I will review the rules within
it, making sure that it a) doesn’t do anything naughty and b) doesn’t
completely break my site. Perhaps if this becomes a burden, I’ll invest more
time writing a script to do the editing for me.

So this is exactly what I do as well – just manually create the whostyles and
apply them once I get into a longer dialogue with someone. This gives me (and
hopefully you now) plenty of time to mess with whostyles in the field.

Over years of reimagining ourselves online, it would be very complex to create
styles that properly support everything we’ve written.

This is another thing I think about as well – and I guess I was going to take it
on a case-by-case basis. If h0p3 has a new style, I might make a new ‘h0p3_2’
style for him – or might just update the old stuff if it makes sense.
Ok – as far as your proposals, they look good! My original plan was pretty
shaky – so am glad to see improvements. Just feeling a lot of gratitude that you
took the time and have energy to put into it.

I guess, as a bit of additional response, I should also mention that I’ve
thought about doing this as a JSON format rather than as CSS.
Here’s a look at the JSON format we’ve been using for
Multiverse box styles.
{
"header": {
"color": "#6B1173FF",
"back": "#B6B5A8A5"
},
"main": {
"fill": {
"type": "Solid",
"color": "#FAE9FF00",
"back": "#FFFFFFF2",
"direction": "vertical"
},
"border": {
"color": "#000000",
"style": "none",
"radius": 0
},
"shadow": {
"type": "None",
"color": "#B6B5A8A5",
"style": "plain"
},
"highlight": {
"type": "None",
"style": "plain"
},
"text": {
"font": {
"family": "Roboto"
},
"fill": {
"type": "Solid",
"color": "#6B1173FF"
}
}
},
"title": {
"fill": {
"type": "Solid",
"color": "#FAFAFA00"
},
"border": {
"color": "#2DC0A6FF",
"style": "dotted_1px",
"radius": 0
},
"shadow": {
"type": "None",
"style": "plain"
},
"highlight": {
"type": "None",
"style": "plain"
},
"text": {
"font": {
"family": "Red Rose"
},
"fill": {
"type": "Solid",
"color": "#17C27FFF"
}
}
}
}

For fonts, we could keep an expanded list of font names that are supported – or
at least a kind of registry – just like browsers already understand Verdana, Arial,
Courier, etc.
So perhaps this paired with a font registry format would do the trick. I don’t
have a strong preference tho – and am just throwing this out there.

Syndicated copies: