vidarh

joined 1 year ago
[–] vidarh 4 points 11 months ago (2 children)

Sadly, while you can export everything, you can't import posts in Mastodon without having direct access to the database and poking about.

You can import following lists, bookmarks, blocks and mutes, though.

[–] vidarh 1 points 11 months ago

"Four score and seven years ago"‍ from the Gettysburg Address... Many languages have or had words for counting in 20's. They've just mostly gone out of fashion.

[–] vidarh 1 points 11 months ago

Wiktionary suggests the common proto-Germanic root of eleven/twelve, elf/zwölf are likely to have been "ainalif" and "twalif" - "one left over" and "two left over".

[–] vidarh 4 points 11 months ago

They're both Germanic languages, just like Dutch, German, Norwegian, Danish and a few others. Same origin. All of them have variations of tre/dre/drei/thir/þre/þrēo (say them with sounds halfway between t and d as the first sound, and you'll see how similar they are) followed by variations of ten/teen/tin/tan/ton/tien/zehn as a suffix for ten (again, pick a halfway point between t and z and it's easier to see how similar they are).

In Old English it was þrēotīene ( þ is "th"), and in Old Norse it was þrettán, same as modern Icelandic, so the first common root is even further back, but you can see the similarity. The *hypothesized proto-Germanic root is þritehun. (þriz + tehun.

But, it goes back even further than that. The Romance languages (tres, trois etc) shares the same proto-Indo-European root (hypothesized to be tréyes) for three with proto-Germanic.

The names for numbers are ancient, and though not always recognisable, sometimes recognizable variants pop up even further away than you'd expect. E.g. Pashto (Southeastern Iran) has dre for three, Sanskrit has trí, Indonesian has tri, all of them descendants of the same proto-Indo-European root.

[–] vidarh 1 points 1 year ago

I'm running my own Mastodon instance and might run a Lemmy instance and others in the future, and there are a few immediate issues I'm already concerned about in that I'd love to be able to avoid having a whole bunch of unrelated usernames.

But webfinger responses only distinguish between mime types, and provide no additional hints, so while I now forward my webfinger queries from my main domain to my Mastodon instance, I'm not sure if I can replace that with a response that lists both Mastodon and Lemmy without breaking both (I also don't know whether or not Lemmy handles aliasing the same way Mastodon does, but that's me being unfamiliar with Lemmy). It'd be really nice to be able to have an alias point at multiple different fediverse accounts if you want to.

But that also raises additional usability issues with respect allowing picking which account to follow if there are multiple compatible options in a webfinger result...

There are many additional usability challenges there. E.g. if my Mastodon followers come across a Lemmy post I make on Mastodon, it'd be nice to have a mechanism to resolve back to my preferred alias (and vice versa) so they knew I'm me even if it's coming from another platform. It'd be really hard to get all of this flow smoothly without confusing users.

But it'd also be enormously powerful if this is addressed even if only partially, because if admins can optionally be able to offer users unified identities across platforms they provide (or even aliasing to other instances entirely), it instantly turns the Fediverse from a collection of different platforms to something much more comprehensive that none of the centralised Twitter, Reddit, Instagram etc. "alternatives" can even hope to try to compete with.

view more: ‹ prev next ›