This seems like a good fallback given the API limitations 👍

huh. from a product/UX perspective, i have to admit i'm not convinced. it seems like this could be an unwelcome surprise to users:

• we'd only do it for tweets with pictures. we'd continue to ignore person tags when publishing non-photo tweets and to other silos.
• twitter has actual first class person tags, so users would expect we'd use them (not knowing that we can't). instead, we'd do something different and unusual.
• person tags could quickly eat into the valuable 140 char real estate and shorten the text too drastically.

the last is (reasonable) speculation, but the inconsistencies in the first two are real. if we were to do this, i wonder if we should do it for all tweets and silos, for consistency. worth thinking about more.

twitter has actual first class person tags, so users would expect we'd use them (not knowing that we can't). instead, we'd do something different and unusual.

Agree, but also if a lot of people started using 👤, it's a signal to Twitter "hey we'd use this feature if you supported it" (like @kevinmarks's emoji reactions, which seemed ... particularly effective)

It'd also be important not to duplicate the person's tag if they are @-mentioned in the body of the post.

person tags could quickly eat into the valuable 140 char real estate and shorten the text too drastically.

I convinced myself that this wasn't a big deal because the text that accompanies a photo is usually short and/or less important compared to the photo itself. I definitely wouldn't be OK with losing a bunch of characters from note posts

I person tag speakers in noterlive.com html generation. I don't in the tweets I make, just @ mention them

looking at this another way: bridgy is big enough now that it may not the right place to experiment with unusual new designs, patterns, and techniques. it may be more appropriate for individual sites to try out this kind of thing, see how it works, iterate, and then we add it to bridgy once it's stabilized and there's a bit more of a consensus.

That I agree with. Bridgy Publish should be a lagging indicator -- adopt conventions based on what people are doing with their own sites, not set new conventions.

Agreed in general with a lot of this analysis, especially with personal site experimentation first. Planning on it. :)

Some updates:

• For more than 3 person-tags:
  👤 @exampleperson @exampleorg …
• Put original post link at the end as usual in Bridgy (easier, more consistent, makes more sense in the abbreviated person-tag text case).

Continuing to iterate on design thinking here: https://indiewebcamp.com/person-tag#tweet_text

In continuing to iterate on the brainstorming of how to POSSE person-tags to Twitter, especially from in-content UI vs a separate explicit UI, I am discovering even more complexity in "doing it well".

Thus while I'd like to leave this as an open issue for Bridgy (my intuition is still that there exists a simple/elegant/predictable solution that Bridgy could adopt), for now the burden of simplicity / implementability falls on some sort of indieweb site version.

very related to #527. somewhat related to #765, #761, etc.

it's been 3y, and we haven't really started doing this at all in the indieweb community. tentatively closing. feel free to reopen if/when we do!