endatabas

SQL document database with full history.

[endatabas @endatabas]

We've been eagerly awaiting this paper since it was announced. A fun Canada Day read. 🍁

[Andy Pavlo (@andypavlo.bsky.social) @andy_pavlo]

It took three years to finish, but our follow-up to the 2006 "What Goes Around Comes Around" is finally out! Stonebraker and I examine the last 20 years in databases and discuss why relational databases + SQL will continue to remain on top.

📄PDF: https://t.co/ZwTWSxXLWb

[endatabas @endatabas]

@dejohn_paul @sogrady @gather_town This is what the owner user sees, but I just signed in with a new email user and they can't see the 'endatabas' room (and so can't see the test msg). And area '6' says "Chats here are not saved", which is a bit of a pain.

Folks in our Gather don't use DMs at all.

[endatabas @endatabas]

@dejohn_paul @sogrady @gather_town [2/2] If 2 users are online and send chat msgs, both can see them later. Offline users cannot.

If only 1 user is online and sends a msg, no one else can see it, ever.

Cognizant this could be a security thing, but we'd prefer Simplicity over Protection.

[endatabas @endatabas]

@dejohn_paul @sogrady @gather_town [1/2] We would benefit massively from persistent chat. Part of what we like about Gather is the simplicity... but the persistent portions of chat are currently very confusing.

[endatabas @endatabas]

@MichaelDrogalis In which direction? Closed source projects with collective ownership? Sure, that's uncommon (though not unheard of).

Open source projects with singular ownership? Very common — arguably, increasingly common.

[endatabas @endatabas]

@sogrady @zotero ...sharks? (+1 for @zotero)

[endatabas @endatabas]

@sogrady Avoiding Google with stuff like @ProtonPrivacy (email, calendar, file share) and @PlausibleHQ (polite website analytics) has been nice.

[endatabas @endatabas]

@sogrady We love @excalidraw for meetings / whiteboards / doodles. No login required. Great export options. Easy to share and collaborate.

[endatabas @endatabas]

@localfirstconf @jessmartin @dxos_org @EffectTS_ If these folks are still giving away DXOS t-shirts, grab one. Tasteful and well-made ... wishing we'd grabbed more from the post-StrangeLoop unconf.

(Also, yay frameworks.)

[endatabas @endatabas]

@localfirstconf @martinkl That's a heartwarmingly optimistic slide, if we've ever seen one. Wish we were there.

[endatabas @endatabas]

@sarahcat21 @cedar_db We've been making the same argument for years... though, in our minds, Cedar is actually quite conservative.

Cedar is exciting to see, as it's nearer to the existing Postgres market with calculated — rather than radical — departures. Historical data would be nice, though.

[endatabas @endatabas]

@gwenshap @jer_s @EmreSevinc A fun historical read: https://t.co/Y8X0l4SU7x

[endatabas @endatabas]

@arpit_bhayani Can someone review our code, please.

[endatabas @endatabas]

@jeanqasaur HTML, CSS, BibTex, bash, and M4:

https://t.co/uvEFRYSVb7

[endatabas @endatabas]

@byrnemluke Nobody wants a table without space for your legs. They all look pretty uncomfortable.

[endatabas @endatabas]

@medriscoll @pinecone @qdrant_engine @trychroma @weaviate_io @milvusio The MongoDB analogy is funny in another way: MongoDB itself is arguably a (potential) feature of the relational model — and one that complements vector search nicely.

SQL document databases ... maybe their moment in the sun is coming? 🙃

[endatabas @endatabas]

@isb @medriscoll @pinecone @qdrant_engine @trychroma @weaviate_io @milvusio 😉

[endatabas @endatabas]

@apprxlinear @medriscoll @pinecone @qdrant_engine @trychroma @weaviate_io @milvusio Exactly. That said, most of the primary data sources we have today don't serve vector indexes well. For a lot of middle-ground (where teams don't want to run half a dozen data stores), natively document/embedding-friendly databases seem like a good idea.