jeremyjh

14 ms·

391.
by jeremyjh 6mo ago
Those are custom software or heavily customized implementations of ERP and similar systems for very large organizations. I’m talking more about the SMB market where today it’s possible for a small team to carve out a niche and make a nice l
392.
by jeremyjh 6mo ago
Frankly this is what everyone is counting on whether they know it or not. The question though is not “will the models get good enough?”. The question is does the repo even contain enough accurate information content to determine what the sy
393.
by jeremyjh 6mo ago
What concerns me about this is that as these stories multiply and circulate people will just completely stop buying software/SAAS from startups, because 90% or more will be this same thing. It will completely kill the market.
394.
by jeremyjh 6mo ago
20 years is rookie numbers in these systems. I guarantee it’s been at least 40 years since a single fuck was given.
395.
by jeremyjh 6mo ago
Once you are over a certain threshold it’s more about the average quality of training data than the quantity.
396.
by jeremyjh 6mo ago
It certainly makes sense to use python for ML or data science.
397.
by jeremyjh 6mo ago
Perhaps one day, all network services will be provided by LLMs natively. Truly, that would be a day in the future.
398.
by jeremyjh 6mo ago
In the blog article[1] that linked to, it says "Unified transactional and analytical workloads: Lakebase integrates seamlessly with the Lakehouse, sharing the same storage layer across OLTP and OLAP. This makes it possible to run real-
399.
by jeremyjh 6mo ago
Its more like old-man-yelling-at-billboard. Its just marketing. Its like complaining about the font they chose.
400.
by jeremyjh 6mo ago
Lakebase is for transactional use cases - this is more comparable to AWS Aurora.
401.
by jeremyjh 6mo ago
I don't think it should be surprising that vendors are not going to lead with "disaggregated storage". I don't see that taking off either. This isn't a paper in a journal. Aurora doesn't call it that either. Bu
402.
by jeremyjh 6mo ago
I really like that article, thank you for sharing it again. I wish I had read it a decade ago, or even the first time you submitted it - I had to learn some of these things the hard way. I agree with you - I don't think Stripe has made
403.
by jeremyjh 6mo ago
The client is part of a distributed transaction. It can't be oblivious to this. Clear semantics and accurate adherence to them is the only answer that doesn't make the overall system unsound. Client bugs are expected and so the si
404.
by jeremyjh 6mo ago
The client is already participating in a transaction in a distributed system. There is no way to change the reality of that. Suggestions about masking this only make the composite system unsound and will not improve net service reliability
405.
by jeremyjh 6mo ago
> So you have to serialize the requests, Not necessarily - there are different transaction isolation and conflict resolution methods provided by every database built for this purpose. You just have to ensure that only one request actuall
406.
by jeremyjh 6mo ago
They pay for a lot more open source work than that as well, but they also don't get to make any special claims for doing that. None of it is charity - it is simply in the collective interest of a lot of tech companies to commoditize an
407.
by jeremyjh 6mo ago
Its ok that you have no knowledge or skill of software development. You don't have to keep creating new accounts to troll people with. Why not just go post on X?
408.
by jeremyjh 6mo ago
How do you explain how more than half of that work was done in 2021? Or did you not look at the repo either? Even in 2023 AI could not really vibe code much of anything. Tell me about the AI coding tools available in 2021 though. It was al
409.
by jeremyjh 6mo ago
This is incredibly poor advice. The bug was clearly in the client code, which did not understand the purpose or usage of the idempotency key. The API itself probably had a design flaw as well - it sounds like it needed a session or transact
410.
by jeremyjh 6mo ago
They knew what they were doing. They released OSS to build traction and a community. In some cases, the community contributed quite a lot to the quality of the software - even if not a lot of code. It never would have gained any traction or
411.
by jeremyjh 6mo ago
So...you didn't ? https://github.com/nooga/let-go/tree/d9dc094822b2983ebf44604... In 2023 he had a working Clojure compiler with: Macros with syntax quote, Reader conditionals, Destructuring, Multi-arity
412.
by jeremyjh 6mo ago
Did you even look at the repo history? Clearly it blasted off this year that way but that isn't how it started. Probably he got way more into it once he could make faster progress on all the yak shaving required to make it more useful.
413.
by jeremyjh 6mo ago
You are right, I was mistaken about the version. I evaluated it in general chat assistant prompts plucked from my history across a range of topics but did not use it for coding - there was never a time when I thought 4o was “good enough” fo
414.
by jeremyjh 6mo ago
Qwen3.6 9B is as good as GPT-4o and runs on my M2 MacBook Air. Models are getting stronger and less costly at the same time, but these are somewhat separate branches of research. Frontier labs are spending more because they are still getti
415.
by jeremyjh 6mo ago
This discussion is not about superintelligence, it is about continued progress. Fully general human intelligence at much lower cost than humans is all that is required to profoundly reshape society, but it is not clear even that will happen
416.
by jeremyjh 6mo ago
What the fuck does that have to do with “soon”?
417.
by jeremyjh 6mo ago
What’s the feedback loop that leads to a total financial collapse? This looks much more like dotcom bubble. Everyone knows where the exposure is.
418.
by jeremyjh 6mo ago
Have you actually used them ? They are really good now if you configure them correctly. Code Rabbit catches more bugs than anyone in our main platform - mostly because it gets the first crack but that is still significant time and churn sav
419.
by jeremyjh 6mo ago
Sure, but these are small departmental apps, 20 users or less in most cases. It’s not like everyone is using every app. The alternatives at this scale are far worse.
420.
by jeremyjh 6mo ago
It would, but that is not what is happening here so far, these are less than 5% of our workload.

More ›