Which is why a good query plan is so important, so that as much disk I/O can be avoided as possible (predicate push down, index elimination, join ordering, partition/scan pruning). Like the old CTE optimization fence problem.
4,354 karma · joined August 11, 2014
In practice even NVMe SSDs at the time were maybe 10x faster for sequential R/W and maybe only 200-500x for random I/O. Most databases in 2015 were limited by serial dependency chains like walking a B-tree to figure out what nodes to read next and couldn't max out what an NVMe was capable of. Not to mention we were stuck with libaio in 2015 that made O_DIRECT ops synchronous.
So the OPs claim of reading multiple terabytes of data in seconds on AWS in 2015 literally wasn't technically possible.
SSDs simply were not as ubiquitous in 2015 as they are today, and I didn't know anyone using them outside of mission critical systems were the cost made sense.
That sounds unlikely unless those machines had access to crazy disk I/O. A local RAID 5/10 with 8 drives would still take 30-50 minutes just to read that much data. Even with a mid-range SAN you still would have spent 15-25 minutes just reading data. This assumes 7.2K SAS/SATA since SSD/NVMe were not ubiquitous in 2015, but even with 2015-era SSDs you're still looking at half that much time spent reading.
Maybe 15 years ago multi-touch attribution was the big push but even that only adds minimal insights into how ad spend affects the behavior you're trying to measure. It helps with high-level channel attribution but it's still at best directional.
So once you reach a point when increasing spend at Meta/Google results in superlinear increases in CAC vs. spend, you eventually end up having to move those ad dollars elsewhere and over time (and a lot of money) you get an intuitive understanding of how these other mediums trickle down to the metrics that matter.
You're totally right about agencies and their perverse incentives and how that leads to sub-optimal spend allocation.
Baically protobuf, Cap’n Proto, msgpack, Avro, or every other serialization approach that versions schemas?
Postgres doesn't do "tiny writes" - it writes whole pages multiple times on every update even if you're only changing 4 bytes, combined with even more writes later on when vacuuming tables. This is one of the reasons why the historical advice was to not build high-volume queues on top of Postgres and why "oh look another post about queues on Postgres" keeps soliciting comments like this.
This ignores the fine print. It scales a very long way under specific circumstances with specific workloads. The more write-heavy your workload the less eloquently Postgres scales.
For OpenAI's use case you could swap Postgres with MySQL and it would scale just as well.
So in other words you're just vibing conspiracy theories and will believe whatever makes you feel better.
The only example of thin skin I see is the character assassination attempts and petty attacks by the Zig team.
This attitude is so precious to me - "Sorry guys, you can't criticize my criticism."
> What Bun did here is basically a sabotage of Zig's reputation
If Zig's reputation was so fragile that Bun migrating to Rust caused this kerfuffle, then Bun isn't the issue.
> unrequested fork
No such thing exists.
Can you point to even one such example?
> not having any morals
You don't agree with someone so you claim they have no morals. I don't have words for how stupid this is that anyone can so casually dehumanize someone over something so trivial.
And yet you follow this up by making assumptions about the motivations behind TS. Anders has mentioned that TS was purely motivated by internal MS teams struggling with huge JS codebases, nothing to do with C# but clearly his work on other languages would have influenced how he approached designing the language.
> our argument is basically "the law says so".
That's the only argument that matters, because monopoly is a legal matter.
Both iOS and Android both generate more in gaming revenue than Sony. The PC gaming market is only slightly smaller than the entirety of the console market. It's going to be very difficult to make an argument that customers don't have reasonable choice, and the recent case against Apple set precedent that owning the single channel of distribution to a product and platform you own isn't monopolistic unless you do shady things like Google did.
Sony's exclusive rights to distributing games on its owned platform and console will not give it pricing power (won't pass SNNIP) or exclude competition.
Sony can be called a dick for what they're doing without it also having to be monopolistic. People are right to be angry and calling them a monopoly over it is being cathartic.
They've always had control by licensing who can publish games on their platform. Now that contract with publishers has changed who can distribute. Publishers are still free to distribute on other platforms in whatever formats those platforms allow.
> that market being Playstation games
This is arbitrarily narrowing the definition of market as far as possible while ignoring reasonable (legal) thresholds. Case in point how Apple's App Store was not ruled a monopoly while Google's Play Store was - there are reasonable tests to pass, and Sony's case will land closer to Apple than Google. And a monopoly is not the only form of anti-competitive abuse.
Might as well claim Domino's has a monopoly on Domino's pizza.