comex

12 ms·

151.
by comex 1y ago
You're the one who said "you literally can't tell the difference". I agree to a point. It seems plausible that they were experiencing some other effect such as hitching or slowdown, rather than just a constant 100ms de
152.
by comex 1y ago
Let's see. https://web-backend.simula.no/sites/default/files/publicatio... > At the most sensitive, our findings reveal that some perceive delays below 40 ms. However, the median threshold suggests th
153.
by comex 1y ago
Another option is to use the "processor trace" functionality available in Intel and Apple CPUs. This can give you a history of every single instruction executed and timing information every few instructions, with very little obse
154.
by comex 1y ago
Clang is a bit smarter than GCC here (for some definition of 'smart') and does optimize the original version: https://gcc.godbolt.org/z/3Y4aheG6x
155.
by comex 1y ago
Only if you have ECC RAM, and even then it’s not perfect.
156.
by comex 1y ago
Reinforcement learning by definition operates on reward functions.
157.
by comex 1y ago
This is a parody but the phenomenon is real. My uninformed suspicion is that this kind of defensive programming somehow improves performance during RLVR. Perhaps the model sometimes comes up with programs that are buggy enough to emit exce
158.
by comex 1y ago
I've usually seen compilers handle large constants with MOV/MOVK sequences (encoding 16 bits of data per 32-bit instruction) instead of loading them from memory. Loading from memory was more common on 32-bit ARM.
159.
by comex 1y ago
Discord has lots of Easter eggs, presumably because of its gaming origins.
160.
by comex 1y ago
Perhaps people won't use it. But I for one want it to be used, and I will certainly be using it in my own code if the PEP is accepted. Startup time is very important to me, more important than the cost of making the code noisier.
161.
by comex 1y ago
> Even if it has side effects the caller may want to defer those as well. But that's rare, and could be handled with existing workarounds. Normally, a module needs to be eagerly imported if and only if it has side effects.
162.
by comex 1y ago
I don't hate it but I don't love it. It sounds like everyone will start writing `lazy` before essentially every single import, with rare exceptions where eager importing is actually needed. That makes Python code visually noisie
163.
by comex 1y ago
I don't think this is actually memory-safe. It sounds like the JavaScript-specific parts of the VM are untrusted, i.e. bugs in it won't violate memory safety. But the core GraalVM compiler and its optimizations would still have
164.
by comex 1y ago
They already did introduce replays: https://en.wikipedia.org/wiki/Instant_replay_in_Major_League... But replays weren’t allowed for balls and strikes, the subject of the new system.
165.
by comex 1y ago
What prevents an attacker from using nested maps to make the server spend exponential amounts of CPU and memory on the response? Is there some kind of limit on the total number of response items?
166.
by comex 1y ago
Not especially relevant for Git which has never provided a shared library interface.
167.
by comex 1y ago
Sounds interesting as a format, but the implementation is a big supply-chain attack risk if you're not already in the JVM ecosystem. This is because the only implementation is written in Kotlin. There are Python and Rust packages, but
168.
by comex 1y ago
They can be, but the app itself has to opt in. You can see whether an app is sandboxed in Activity Monitor (it’s a hidden-by-default column named Sandbox).
169.
by comex 1y ago
There's pthread_cond_timedwait_relative_np. Looking at the implementation, pthread_cond_timedwait starts by converting the deadline to a relative timeout [1], but this is skipped if pthread_cond_timedwait_relative_np is used. Then, i
170.
by comex 1y ago
Yes for networking. You set your sockets into O_NONBLOCK mode and use poll() or select(). These APIs are in POSIX and also have direct equivalents in Winsock. There is also POSIX AIO for async I/O on any file descriptor, but at least
171.
by comex 1y ago
pthread_cancel is not a good design because it operates entirely separately from normal mechanisms of error handling and unwinding. (That is, if you’re using C. If you’re using C++ it can integrate with exception handling.) A better appro
172.
by comex 1y ago
The comment you link to says that "we only performed a quick preliminary search" and "We do not have a method for automatically checking existing trajectories." In other words, it can't confirm that the issue only
173.
by comex 1y ago
Nope. The court is applying an old law that specifically applies to carriers providing "telecommunications services", no one else. (Incidentally, even the term "telecommunications service" only encompasses voice call se
174.
by comex 1y ago
I found the illustrations confusing; I had to go over them and the accompanying text over and over before I understood what was being depicted.
175.
by comex 1y ago
AI slop aside, just for the record, - SPU is not a processor, it's a generic term that encompasses multiple coprocessors. - The log lines don't even mention the Secure Enclave Processor (SEP). - Each line of log output is its own
176.
by comex 1y ago
For e-books, there will usually be a license agreement that prohibits any kind of nonstandard use. That's why Anthropic had to scan physical books.
177.
by comex 1y ago
This post has a few examples: https://daringfireball.net/2020/07/return_and_enter
178.
by comex 1y ago
Good luck! I look forward to seeing the results.
179.
by comex 1y ago
I don’t understand what you mean. Training an LLM requires orders of magnitude more tokens than any one human will ever read. Perhaps an AI company can amortize across all their users, but it would still represent a substantial cost. And
180.
by comex 1y ago
But how would you prevent use-after-free? In other words, how do you prevent a child object's method from reaching out to its parent object and, say, clearing the container that contains the child object? Specifically, how do you do

More ›