turtles3

16 ms·

1.
It's 2026, Just Use Postgres(tigerdata.com)
532 points by turtles3 7mo ago|333 comments
2.
by turtles3 9mo ago
To be fair, postgres still suffers from a poor choice of MVCC implementation (copy on write rather than an undo log). This one small choice has a huge number of negative knock on effects once your load becomes non-trivial
3.
by turtles3 9mo ago
I'm also curious about this, especially if anyone has operated postgres at any kind of scale. At low scale, all databases are fine (assuming you understand what the particular database you're using does and doesn't guarantee)
4.
by turtles3 1y ago
I'd say that the 'poor man's closures' aspect of OOP - that is, being able to package some context along with behaviour is the most useful part for day to day code. Only occasionally is inheritance of anything other than
5.
by turtles3 2y ago
I agree, but this > Your program must be really small and scoped for this to make sense. to me suggests that it's not really global state if your program has to stay small and scoped. In a sense, your program has just become the con
6.
by turtles3 2y ago
Indeed, we also have technology that can read the written language. Seems like that would be the way to go, and you get to keep backwards compatibility with eyeballs
7.
by turtles3 2y ago
I appreciate this survey for how thought-provoking it is. Ironically, I'd say that the survey is itself art. And not a piece of art that AI in it's current state could ever pull off. Maybe that's when the AI art turing test w
8.
by turtles3 2y ago
To be fair, newer research is demonstrating that smaller more power efficient models with the same performance are possible, so the hope is that these giant LLMs are just a stepping stone to a less energy hungry place. In contrast, proof of
9.
by turtles3 2y ago
This is sort of the role that L3 cache plays already. Your proposal would be effectively an upgradable L4 cache. No idea if the economics on that are worth it vs bigger DRAM so you have less pressure on the nvme disk.
10.
by turtles3 2y ago
I'd say there is a fourth category too - things that would be perfectly fine as a simple, local program purchased once that grow over-complicated cloud features to justify a subscription model. Examples of this would be Lens, Postman a
11.
by turtles3 2y ago
Having spent an unhealthy amount of time thinking about this, I think it's even worse than an abstraction _level_. I suspect that the fundamental problem with visual languages is that you have to reify _something_ as the objects/s
12.
by turtles3 2y ago
My understanding of these prefixed IDs is that translation happens at the API boundary - what you store in the database is the unprefixed ULID/UUID in an efficient binary column. Then, whenever the API includes an id in a response, it
13.
by turtles3 2y ago
As a random thought, this seems to be about the same order of magnitude compute as Karpathy's recent GPT-2 work: https://github.com/karpathy/llm.c/discussions/677 You could take the final checkpoint from
14.
by turtles3 2y ago
The hope though is that by splitting indexing that puts search providers on an equal footing in terms of results quality (at least initially). Advertisers go to Google because users go to Google. But users go to Google because despite recen
15.
by turtles3 2y ago
Honestly I think Google needs to be broken up. It's not a novel idea but the more I think about it the more I like it. So, Google becomes two orgs: Google indexing and Google search. Google indexing must offer its services to all searc
16.
by turtles3 2y ago
Replication can be for HA, not just for scale. All depends on your business requirements. Also replication can be good for other operational reasons, such as zero downtime major version upgrades. Again depends on the business need/expe
17.
by turtles3 3y ago
Same, as it stands you the user are legally liable for the full bill unless netlify graciously forgive it. Even in op's case, they didn't (still charging 5k!). If there was an option to cap billing, or at least some legally bindin
18.
by turtles3 3y ago
I'm also interested to know this. I have a couple of static sites running on the free tier for friends/family and now I'm planning on moving them all to a VPS as soon as I can. It is beyond ridiculous that serverless provider
19.
by turtles3 3y ago
Perhaps it is _anemoia_ - nostalgia for a time you've never known https://www.dictionaryofobscuresorrows.com/post/105778238455... In this case, it's for the harmless charm of an imagined past, but the same fo
20.
by turtles3 3y ago
> Another thing that radically changed my mind about ocean plastics is that most all of it comes from a handful of rivers. The article you linked specifically states that this was previously thought to be true, but is not. It does howeve
21.
by turtles3 3y ago
The worst thing that I've found with teams is the latency in a video call is _just slightly_ too high (and noticeably higher than Google meet and zoom). I find this latency to be absolutely critical in keeping the flow of conversation
22.
by turtles3 3y ago
Just letting you know it's not you - this is a well known long standing issue with keycloak. Typically users see a significant performance cliff at around 300-400 realms. While one realm is not necessarily the same as one tenant in key
23.
by turtles3 3y ago
Cnpg's docs articulate this better than I could: https://cloudnative-pg.io/documentation/1.16/controller/ Statefulsets have their place but are surprisingly inconvenient for database workloads.
24.
by turtles3 3y ago
A subtle advantage of cnpg is that it doesn't use statefulsets, instead the operator handles things like mapping storage volumes and stable identities. Regular kubernetes statefulsets have some tricky sharp edges for failure recovery.
25.
by turtles3 3y ago
I believe what you're looking for is Hypernetworks https://arxiv.org/abs/1609.09106
26.
by turtles3 3y ago
Oh sorry I didn't mean to imply I'd used it in production or it was production ready, I simply consider it to be a really promising project. It's tackling what I see as one of the foundational weaknesses of postgres, which is
27.
by turtles3 3y ago
This is a great question, and if the choice is between mysql and postgres, I would like to make the case that despite the current popular momentum behind postgres, mysql is a better default. Please note I'm not saying mysql is better,
28.
by turtles3 3y ago
It all depends on whether the present king of France is bald https://en.m.wikipedia.org/wiki/Definite_description
29.
by turtles3 3y ago
Can recommend the cnpg project https://github.com/cloudnative-pg/cloudnative-pg It's a kubernetes operator for postgres with a pretty good range of features Honestly I wish something of this standard existed for
30.
by turtles3 3y ago
Agreed on not becoming too dependent on Stripe. Billing (or in this case, pricing) is the last thing you want under another entity's control. There is surprisingly little in the OSS space around this, but I can recommend KillBill (no a

More ›