peterwaller

16 ms·

121.
by peterwaller 13y ago
Honest question: what field are you in which means your candidates are going to need to write custom allocators on a regular basis? What problems need custom allocators? I'm a fan of lower level details and all, but surely you can get away
122.
by peterwaller 13y ago
For perspective, the ATLAS [1] experiment at the LHC records O(400 MB/sec). To disk. Permanently. (With a duty cycle of something like ~0.5, in principle). So, I guess that's a different ballgame than temporary acquisition for decision maki
123.
by peterwaller 13y ago
Relevant: https://github.com/jteeuwen/mitosis
124.
by peterwaller 13y ago
Another interesting one, byteplay [1] makes bytecode-hacking childs play. The most evil code I ever wrote [2] actually used this to modify the caller's bytecode in order to turn a log message into an exception with the correct stack trace f
125.
by peterwaller 13y ago
I found that earlier - unless I'm missing something, it doesn't actually explain how the closure is packaged. Does the worker need compiled-in knowledge of what the master wants to execute in advance, for example? It's not clear to me.
126.
by peterwaller 13y ago
Can someone explain how `Spawn` ships the closure for execution to a remote host? I'm quite curious to know. I had a grep through the source, but I only found one [1] implementation of `Spawn` accepting a function as a parameter in the code
127.
by peterwaller 13y ago
Re: "Why not?" Generally you should strongly avoid putting generated artefacts into version control. This leads to complete pain if ever you find yourself trying to diff or merge when they inevitably change. The problem is that you end up w
128.
by peterwaller 14y ago
Not for me. http://www.downforeveryoneorjustme.com/duckduckgo.com/
129.
Direnv 2.0 is coming, rewritten in go(direnv.net)
2 points by peterwaller 14y ago|0 comments
130.
by peterwaller 14y ago
I feel this is against HN etiquette but I need to thank you. Be encouraged and go far. This looks awesome. I wish you many a decent pull-request.
131.
by peterwaller 14y ago
I'm on your side, I think this is stupid, but I can't let this stand: > I'm going to point a finger and say that this is clearly untrue, and very easy to disprove just from basic network monitoring over 5 minutes of playing the game. c'
132.
by peterwaller 14y ago
I don't like the use of the word greed here. If you're going to liberate information, you have to be timely. And it would still have been a lot to download even over years, so whilst your proposal reduces the instantaneous bandwidth, the ag
133.
CERN students release DECAY zombie flick under CC license(decayfilm.com)
4 points by peterwaller 14y ago|0 comments
134.
by peterwaller 14y ago
> [...] there's no firm consensus on whether it's literally infinite. If it is finite, there is no "wall" or "center", since it's not a big sphere -- a finite Universe just means that if you go straight long enough, you end up in the sa
135.
cppyy: JIT-friendly C++ calls from PyPy(morepypy.blogspot.co.uk)
1 points by peterwaller 14y ago|0 comments