xtdb

jarohen 2025-05-12T09:56:35.760659Z

It's blog day πŸ“ πŸš€ part 2 of "Building a Bitemporal Index": "Bitemporal Resolution" I've promised for a while to write up how XT's bitemporal magic works - well, here it is ☺️

πŸ”₯ 1
πŸ˜‚ 2
refset 2025-05-12T11:49:11.211209Z

Just in case folks here hadn't seen it advertised anywhere else already, JUXT is running an invite-only 'fintech' conference in London next month (12th June): https://www.juxt.pro/xt25/ Clojure is not the focus of the event but certainly will be discussed in a couple of talks. XTDB will feature lightly in the agenda also. Please fill in the form if you'd be interested to attend and think you might benefit from the content/networking. Recordings will be available online afterwards in any case. Unfortunately I won't be there myself due to the imminent arrival of a new baby(!), however James and others will be there to meet and chat about XTDB πŸ™‚

Stef Coetzee 2025-05-13T09:42:22.387339Z

Congratulations, @taylor.jeremydavid!

❀️ 1
jarohen 2025-05-12T12:32:30.534499Z

🚨 Anyone here using the XTDB Clojure API for a remote (HTTP) XTDB node (`xtdb.client` namespace)? πŸ™‹ In the next release, we're looking to replace the underlying implementation here with one that uses the Postgres-compatible protocol instead. This'd be a breaking change - in your infrastructure, you'd need to expose the pgwire port rather than/in addition to the HTTP port to continue using the Clojure API client. This is because the Postgres server is starting to move ahead of the HTTP API (due to the XT usage that we're aware of). We will likely keep the HTTP API for simpler use cases (e.g. single queries, or putting maps) but the Postgres API (either via next.jdbc or our xtdb.api helpers) will become our recommended default approach. (If you're connecting to HTTP directly - i.e. not using our client - then this won't affect you)

πŸ™‹ 1