sirfz

I've plugged this several times here (just a fan) but chdb's DataStore is a "lazy" drop-in replacement for pandas dataframe. Pandas API with chdb performance. I don't use it myself (I do everything mostly in sql with either duckdb or chdb) but always found pandas more convinient than polars for quick and dirty data crunching (less typing and frankly more pythonic if you're used to slicing).

Firefox was amazing even when Chrome was released

Understood (fyi I was genuinely asking as I only used pyreqwest, which I like, and was looking for comparison if any)

Why not pyreqwest?

Yep, and chdb as well. Said this before, chdb's DatStore is a pretty neat pandas replacement too.

They have a very strong DIY culture which I'm sure helps with making people explore and discover knowledge and develop skill

Cython is a no-brainer really. You write the same language with immense speedup (matching what the "other languages" can achieve at much less effort).

Also tools like numba can beat them all at way less effort.

Imho, dropping into other languages should be the last resort in any project.

My comment is to say the US has proven how brutal they are consistently through all the wars of aggression they have waged in the past several decades. They do not see their "enemies" as human. I can't fix anything unfortunately.

It's incredible, after all the grotesque stories about rape, torture and murder of children, men and women during the Iraq war, active support of genocide (and 10s of thousands of children murdered by Israel, on purpose), prisoners rape and child imprisonment, a "secretary of war" and president publicly admiting to war crimes and saying things like "negotiate with bombs" you still "refuse to believe" that anyone in decision chain wouldn't do anything like this.

Nowadays I'd just use clickhouse-local / chdb / duckdb to query json files (and pretty much any standard format files)

Fwiw, I thought the article is full of great information and well researched. I think your reflex is holding you back.

You can even chain them if you have deep ssh connections (i.e. ssh from one instance to another). I think it would be ~~. to terminate the 2nd hop.

Edit: it's already explained in the OP

You don't need a vector db, you just need np.dot

You being Iranian-american bares no weight on your opinion

No he's saying that israel intentionally bomb schools and hospitals

chdb's new DataStore API looks really neat (drop in pandas replacement) and exactly how I envisioned a faster pandas could be without sacrificing its ergonomics

I think that's a sane take. Indeed, I think most data analysts find it much easier to use pandas over polars when playing with data (mainly the bracket syntax is faster and mostly sensible)

It'll use less storage space

That doesn't work well with conflicts tho iirc

I just ordered this monitor earlier this week (still waiting for it to be delivered), did you need to tweak it or use some thing like MacType like others are suggesting in this thread?

love the casual "solely", truly unhinged

Indeed, I was delighted when I read the part about wutang's time capsule and obviously OP is a wu-tang and general hip hop fan. The intro you shared is dope!

It's a lunatic society.

She is wonderful and her work on investigating the complicity of countries in the Gaza genocide and Palestinian rights / Israeli agression is incredibly thorough and extremely important. More power to her!

That doesn't cover OP's point, some people's brains just work differently and they can achieve something in 1000x less time than others. You can have all the time in the world and you'll never reach their level. That's essentially what talent is.

As already mentioned in other comments, it doesn't really make sense to compare to json parsers since lite3 parses, well, lite3 and not json. It serves a different use case and I think focusing on performance vs json (especially json parsers) is not the best thing about this project

I think the problem with your take is your assumption that the US topples regimes because of dictatorship or to support "democracy".

Good to know, will be interesting when we run our tests. Thanks for the info and for your work on pyreqwest, looks very promising

I've recently updated an internal tool which basically acts as a configuration and dependency/context manager for performing hundreds of api calls. I added an httpx backend (to test vs the current urllib3 backend) and also introduced an async API (httpx as well). However, from your benchmarks it seems like I should've went with aiohttp for faster async? I will work on integrating pyreqwest as well

I find GCP frustrating (coming from AWS) but luckily asking Gemini how to do things makes it much easier.