Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
andriy_koval
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
3 ms
·
1.
▲
by
andriy_koval
5d ago
Thank you, its useful finding that only transparently compressed btrfs is affected.
2.
▲
by
andriy_koval
5d ago
Its odd to me because it looks trivial to add another compression algorithm implementation
3.
▲
by
andriy_koval
5d ago
> Do you perhaps have 'discard=sync' I have 'discard=async'
4.
▲
by
andriy_koval
6d ago
My conclusion is that after many attempts, you couldn't comprehend index issue. Looks like its time to finish discussion.
5.
▲
by
andriy_koval
6d ago
Yeah, I know that thing, but oddly they support only lz4 compression which is much weaker to zstd used everywhere else.
6.
▲
by
andriy_koval
6d ago
I don't, I doubt any nuclear plants could be brought online that fast. But my point was that type of energy matters in this context.
7.
▲
by
andriy_koval
6d ago
as I mentioned its not just SpaceX but also Anthrhopic and OpenAI is also in the queue, which is up to total 6T extracted from index holders. Do you think this is valid concern or trolling?
8.
▲
by
andriy_koval
6d ago
Do you have any numbers how much fossil energy capacity was installed and how much nuclear?
9.
▲
by
andriy_koval
6d ago
its unrelated to index inclusion. Are you just trolling?
10.
▲
by
andriy_koval
6d ago
is it gas turbines energy?
11.
▲
by
andriy_koval
6d ago
index thing happened this year, and now investors are obviously lining up before openai and anthropic to get a cut too during IPO.
12.
▲
by
andriy_koval
6d ago
To me its not just about AI, but how they utilize all this financial scenarios to push things absolutely over the top, e.g. fast inclusion into indexes and essentially funding bubble from people's 401k.
13.
▲
by
andriy_koval
6d ago
no quota enabled
14.
▲
by
andriy_koval
6d ago
> How large is large? I've deleted files with sizes of tens to hundreds of GBs and not seen that, and can probably whip up a test with a single-digit TB file if motivated. for 1TB compressed (probably 5tb uncompressed) it is reprodu
15.
▲
by
andriy_koval
6d ago
btrfs has one critical issue they don't fix: it blocks access to fs for minutes if you remove large files. I am not sure how this is acceptable for prod grade fs..
16.
▲
by
andriy_koval
6d ago
would want to have compression..
17.
▲
by
andriy_koval
8d ago
He just gad better prompt
18.
▲
by
andriy_koval
8d ago
> There doesn't need to be an admission of guilt to determine that the US acted in violation of international law. Its not about admission of guilt, but as I said multiple time about material evidence, which is missing in this case.
19.
▲
by
andriy_koval
8d ago
> the interviews those interviews are opinions not evidence, to my knowledge there was no public statements of anyone directly involved into the target selection process in that case
20.
▲
by
andriy_koval
8d ago
Or there is no reliable source. We know that we don't know.
21.
▲
by
andriy_koval
8d ago
> I've already summarized the material findings I asked about material evidence(I mentioned word "evidence" multiple times), not "findings", which is loaded term
22.
▲
by
andriy_koval
8d ago
So, you didn't read it yourself since can't summarize material findings in 3 sentences (what I asked instead of your "go read yourself" suggestions which are obviously to be ignored)?
23.
▲
by
andriy_koval
8d ago
> Those professions will likely evolve, but the traditional forms (ie writing code by hand, writing law filings by hand etc) there will still be writing initial and incremental prompts by hands, until and if LLMs surpass humans in all in
24.
▲
by
andriy_koval
8d ago
there are always be next frontier of problems which require creativity, unless AI become supersmart completely make humans redundant in all cognitive functions.
25.
▲
by
andriy_koval
8d ago
burden of proof is on one making statement, especially if he is passively-aggressively toxic. Productive discussion would be if you go read that report before bringing it here, and tell what material evidence they have for that specific inc
26.
▲
by
andriy_koval
8d ago
> has a lifetime of experience and credentials to be able to make these determinations [1] This sounds like extreme hand-waving to me.
27.
▲
by
andriy_koval
9d ago
Why instead of answering specific question you switch on explicit personal insults?
28.
▲
by
andriy_koval
9d ago
> The UN mission concluded why "The UN mission" conclusions are blindly considered as some high quality analytics?
29.
▲
by
andriy_koval
11d ago
> let me know if you figure out a good answer :) I currently explore two routes: - targeted ads - producing social impact for goodwill and marketing which may materialize into leads eventually: use my platform to solve problems and remov
30.
▲
by
andriy_koval
11d ago
Why do you need so much apps, what you will do with them?
More ›