๐Ÿˆโ€โฌ› today's musings - the sentence burning on the back of my tongue this week is "I don't think you know what pluralkit is".

multi-identity chat is not a technical capabilities problem, it's a social integration problem - and this is rarely articulated.

when multi-identity is natural to the tech, it stops being exceptional to the user. "can" doesn't cut it.

ask the question: "how does your software avoid implicitly ostracising plural systems"

they won't ask it on their own.
<๐Ÿงต>

๐Ÿˆโ€โฌ› we allude to this in the first screenshot, but we've meet endless plural systems who knew they were plural for years on a rational level, but couldn't see how to integrate it into their life socially and personally until they encountered other systems actively using pluralkit and it clicked. you do not know what pluralkit is. you do not understand the gravity of what you're designing. "we need a pluralkit so systems will use our chat platform" is not good enough. ask actual questions.

๐Ÿˆโ€โฌ› the more you've been exposed to the way that the diversity of plural self expression expands to fit the capabilities and technical-social friction of the tools immediately available, the more it becomes eye-burningly obvious that "let's set the bar where it already is" is a completely ridiculous take. neither of us knows who has no way to comfortably speak today but could tomorrow - if we don't accept stagnation on our standards as the default.

give us, give eachother, more thought than the bare minimum.

๐Ÿˆโ€โฌ› chat multi-id implementation pitfalls:

- Impermanence i.e treating using a different ID as a "change" (IRC /nick)
- Scoping i.e admin agency (tupperbox perms, discord bot invites, matrix via getting IP banned after joining 20 times)
- Armbanding i.e. BOT, APP, "masquerade", "x used proxy", anything that de-naturalises it being a message sent by a user
- Opaqueness i.e. missing interactions (moderator, profile view, etc) compared to other messages OR multi-id not being trivially findable (e.g. client mods)
- Serversidedness i.e client can't preview the ID avatar before sending a message
- Complexity i.e scaling anything with identity quantity (matrix account switching, anything that lacks in-message tag parsing)

This is not something you can "tack on" - the ultimate ideal is that ALL messages/accounts are multi-id - most just never set up their second ID.

PluralKit is worlds better than most people have the perspective to understand, but we can do better. don't compromise on us.

@sleepingdragoninn

a conversation about this is currently going on in the sable matrix client space
https://matrix.to/#/#sable:sable.moe

but people are suggesting labeling to prevent phishing and im not really sure how to respond

like is there a better solution then arm banding to be able to prevent against those kind of abuses of the system
i am really excited that its being worked on (am not trying to bike shed) just hope there is a better solution then arm banding

@lavenderdotpet so a big reason that the reason that the APP tag exists on discord webhook messages is that you can't click on them and see where they came from.

if you're in an impl where you can trivially reveal the username of the sender account on click, you don't need a mark (or it can be a user-defined tag, as if PK system tags were clickable)

if you're not, the mark should be the username itself, not something related to the impl/feature.

@sleepingdragoninn
some poeple agreed with the username pill/tag option i think that might be the route the team goes for the implementation