matt42

5 ms·

1.
by matt42 6y ago
To me there is 3 main points: - Even if it's JS, just-js make almost no use of dynamic memory allocations. - The authors rewrote himself a postgresql driver that support batching request. - it wraps high performance c++ librari
2.
by matt42 6y ago
I don't understand why when you say "I have covid" people want the proof that you have it and ask you "did you do the test ?". I personally think I got it twice and never did any test. Just had symptoms I never ha
3.
by matt42 6y ago
or just run npm test / make test locally without setting up anything. But I feel like writting the tests takes usually more time than setting up CI/CD.
4.
by matt42 6y ago
Very interesting, I really think that http APIs are just a user management layer that could be easily abstracted away now that we can do so many things client-side. How do you handle permissions ? or how do you pervent a clients to erase t
5.
Intel defends against AMD Ryzen 5000 with tease of 11th Gen Rocket Lake info(techradar.com)
2 points by matt42 6y ago|1 comments
6.
Lithium high performance C++ HTTP server documentation(matt-42.github.io)
2 points by matt42 6y ago|0 comments
7.
by matt42 6y ago
This is really depressing. How do we stop this cynical game ? is they example of successful companies promoting the right people ?
8.
by matt42 6y ago
Projects using templates and heavy metaprogramming often see most of the compile time spent in template instantiation, not in link time. For this kind of projects I think this would help.
9.
by matt42 6y ago
Great that C++ compiler teams are working on reducing C++ compile time. The next great feature I think would be a game changer is caching the result of compilation (templates and binary code). A fine grained cache acting at the level of a l
10.
by matt42 6y ago
What http server do you use ? Are they static images ?
11.
by matt42 6y ago
And 2.5 per minute
12.
by matt42 6y ago
Is there any doc about how this caching is implemented ? I still don't understand why we need to feed this dependency array to have an up to date version of the count variable.
13.
by matt42 6y ago
I love the way crank is using generators to get rid of the react complexity.... Did not know about crank, thanks for the link.
14.
Pattern matching C++ proposal [pdf](open-std.org)
2 points by matt42 6y ago|0 comments
15.
by matt42 6y ago
This is the official continuous techempower benchmark, run by techempower on the exact same environment than the techempower "official rounds". The only diff I'm aware of is that there is more manual review and check for roun
16.
by matt42 6y ago
My motivation in writing yet another web framework was to get pure performance while providing maintainability of more dynamic languages. Check the implementation of lithium, it is one of the most concise implementation of TFB: https:&#x2F
17.
by matt42 6y ago
The frameworks are downloaded by dockerfiles provided by the framework maintainers, and since there is hundreds of frameworks, this is a bit tedious to manually get all URLs... But anyway, I'm pretty sure all framework maintainers will
18.
by matt42 6y ago
yes, it has been discused already to increase the number of connections. But still, several thousands of simultaneous connections is still ok for the best performing frameworks...
19.
by matt42 6y ago
Sadly seastar has no implementation of this benchmark. I had a quick look at the website and it seams that seastar is much more low level (it does not provide a http parser) and more generic that lithium, which is focused on simplifying th
20.
by matt42 6y ago
I'll write a post later explaining how I reached those performances, several people already asked for it.
21.
by matt42 6y ago
a kqueue port should be ok. I'll add an issue for it.
22.
by matt42 6y ago
The epoll code use is very small, few hundreds of LOC (but is at the very center of the framework since it runs the event loop), actually writing it from scratch was better to understand the mechanism, to optimize it and to adapt it for non
23.
by matt42 6y ago
I would say the profiler output is the best source of information. Lithium is not using black magic at all. Actually on large framework like lithium, performance problems come more from obvious errors that get hiden in the quantity of code
24.
by matt42 6y ago
Yes, this is only physical. The cloud env (Azure) only run once every 1 or 2 months I think. On the official rounds they do both.
25.
by matt42 6y ago
Hi Brian, can't wait for round 20 then :D.
26.
by matt42 6y ago
arf. ok. It's probably the mods who unflagged it.
27.
by matt42 6y ago
omg...
28.
by matt42 6y ago
Yes zero cost is subjective. It has a cost at least at compile time. But you can see metamap is a like loop unrolling for objects (if you think loop unrolling is zero cost then metamap too).
29.
by matt42 6y ago
I think I need to write a blog post to do a retrospective on what I learned writing iod, silicon, lithium (and vpp, a small image processing lib). There is so much to tell. Just to be short, here is my phylosophy: - don't think t
30.
by matt42 6y ago
And back on the frontpage.. I don't understand HN ranking algorithm.

More ›