5 ms·

there is a running joke among a part of the community that goes something like "0 days since last persistent workflow orchestration library announcement". I fe

there is a running joke among a part of the community that goes something like "0 days since last persistent workflow orchestration library announcement". I feel like Elixir community is a bit too nice to call out obvious slop (some folks straight out post LLM-generated blog posts every week or so), so instead we could focus on elevating human effort. Recent Oban release for example had a note that all the docs a human-written. We should praise this!

sodapopcan 16d ago
> there is a running joke among a part of the community that goes something like "0 days since last persistent workflow orchestration library announcement". Ha, I missed this! I rarely click on any topics anymore, but I have noticed a trend of similar-sounding libraries being posted. Other than one person whose blog I already read a bit, I don't click on any of the blog posts anymore. It's not a conscious decision, I just open scan the topics, usually nothing looks interesting and a close the tab.
- arodenmaxxing 16d ago
  I feel you! I still try to share/comment on human-written posts even if they're not fully up my alley. Pangram extensions helps a lot with that. But yeah, in general it's just crickets and a buzzing noise of tokens floating around
  - sodapopcan 16d ago
    > Pangram extensions Was not aware of these. Thanks!
cpursley 16d ago
Ha, I feel attacked: https://github.com/agoodway/pgflow https://github.com/agoodway/pgflow ^ this one is pretty cool, pgmq does most of the work.
sorentwo 16d ago
Glad somebody noticed and appreciated that! Human written long form content is extremely important (to us, but I think for the world). Every single email and issue reply is also 100% human authored. People deserve to interact with humans online, especially when they have reached out for help and real interaction.
- solid_fuel 16d ago
  Every single time I interact with Oban, it's a delight. Glad to see you guys continuing to put in the work and sticking with human writing. I've got a production service largely backed by Oban Pro and last month alone it handled running heavy ML tasks against ~200,000 long-form video files. Each analysis has multiple steps, all handled with Oban's dependent jobs feature, so there was something like a million unique jobs, and I don't think we hit a single production issue with Oban. Being able to connect into the cluster and directly manipulate things with Oban is a superpower.

sodapopcan 16d ago

> there is a running joke among a part of the community that goes something like "0 days since last persistent workflow orchestration library announcement". Ha, I missed this! I rarely click on any topics anymore, but I have noticed a trend of similar-sounding libraries being posted. Other than one person whose blog I already read a bit, I don't click on any of the blog posts anymore. It's not a conscious decision, I just open scan the topics, usually nothing looks interesting and a close the tab.

arodenmaxxing 16d ago
I feel you! I still try to share/comment on human-written posts even if they're not fully up my alley. Pangram extensions helps a lot with that. But yeah, in general it's just crickets and a buzzing noise of tokens floating around
- sodapopcan 16d ago
  > Pangram extensions Was not aware of these. Thanks!

cpursley 16d ago

Ha, I feel attacked: https://github.com/agoodway/pgflow https://github.com/agoodway/pgflow ^ this one is pretty cool, pgmq does most of the work.

sorentwo 16d ago

Glad somebody noticed and appreciated that! Human written long form content is extremely important (to us, but I think for the world). Every single email and issue reply is also 100% human authored. People deserve to interact with humans online, especially when they have reached out for help and real interaction.

solid_fuel 16d ago
Every single time I interact with Oban, it's a delight. Glad to see you guys continuing to put in the work and sticking with human writing. I've got a production service largely backed by Oban Pro and last month alone it handled running heavy ML tasks against ~200,000 long-form video files. Each analysis has multiple steps, all handled with Oban's dependent jobs feature, so there was something like a million unique jobs, and I don't think we hit a single production issue with Oban. Being able to connect into the cluster and directly manipulate things with Oban is a superpower.