dom0

27 ms·

241.
by dom0 9y ago
> 10,000 MPGe ... 12,600 MPGe (0.024 l/100 km, 0.019 l/100 km)
242.
by dom0 9y ago
> 600-1000mpge (0.24 - 0.4 l/100 km)
243.
by dom0 9y ago
Suggestion: "Tech firm fighting federal order for data on visitors to an anti-Trump website" Or "Tech firm fighting federal order for visitor's data of anti-Trump website" v "Tech firm fights federal order for
244.
by dom0 9y ago
It is possible to do more after a fork (cf. async-signal-safe), but it's hairy enough to just say — don't, always exec (similar to how doing actual work in a signal handler is generally a very bad idea).
245.
by dom0 9y ago
With Python that's a lot of stuff; I suggest running strace python some_small_script.py to see just how much data Python loads on every single startup.
246.
by dom0 9y ago
Things that happen when you employ folks full-time to work on something.
247.
by dom0 9y ago
Considering who funds a lot of FS development on Linux, this isn't that surprising — using O_DIRECT is really really complex and also not in the kernel, so a neat-o way to keep others from doing the same.
248.
by dom0 9y ago
There are no QA processes per se for all file systems. The kernel is pretty much a bundling of many projects and the big file systems are their own projects, managed largely indepently from the Linux masters.
249.
by dom0 9y ago
Differing 64 bit data models don't really matter when software depending on exact widths uses exact width types. Unfortunately, on MSVC this hasn't been supported for many years out of the box, so each project had to include its o
250.
by dom0 9y ago
You're both wrong. P is the union of all TIME complexity classes defined by n^k for every natural k. NP is the union of all NTIME complexity classes defined by n^k for every natural k.
251.
by dom0 9y ago
NP completeness kinda works around the uncertainty of PxNP (with x ∈ {⊆, ⊊}), because it defines some sort of "weak subset" of NP comprised of "pretty sure these problems are not in P[, because no one yet thought of a polynom
252.
by dom0 9y ago
There are a bunch of major assumptions which are not rigorously proven, but generally assumed (hence assumption) to be correct, frequently backed up by a large body of heuristic evidence. P!=NP is one of them.
253.
by dom0 9y ago
> breaking This preprint "implies P not equal NP".
254.
by dom0 9y ago
> There are a lot of people in that boat, and they need better options. Land isn't coming to you, folks, you must start rowing if you want to get there. Rewrite bit for bit. Module for module. Package for package.
255.
by dom0 9y ago
No, it must be explicitly released. The GIL must be held to invoke almost all Python runtimes (main exceptions: acquiring the GIL, low-level allocator).
256.
by dom0 9y ago
Depending on where you want to go and the application, such things may be good idea for a low number of workers but can become a major bottleneck.
257.
by dom0 9y ago
Just like you can share memory between processes, you can also share OS-level locks and semaphores between them. A distributed lock manager is not required for the single-node case.
258.
by dom0 9y ago
Unless your model actually consists of a large number of Python objects (and not a handful of PyObjects referencing something like a np array), there isn't really anything blocking you from doing so. You can have a master process map t
259.
by dom0 9y ago
Removing the GIL (in a non-braindead way) likely entails breaking all existing code using the C API. PyPy could do so without breaking cpyext, by maintaining the illusion of a GIL whenever control passes to cpyext.
260.
by dom0 9y ago
> (see Amdahl's law) Amdahl's law bears little relevance to throughput computing (i.e. most servers). > (It's also buggy, has internal race conditions, and easily leaks resources.) There is also at least one memory cor
261.
by dom0 9y ago
> It's just an implementation detail of the underlying VM. Python itself is just an implementation detail of the underlying VM.
262.
by dom0 9y ago
The most general layer in Linux (VFS) doesn't really support asynchronicity, so changing this would require a bunch of changes to every tree and out-of-tree FS - not viable. However, the usual suspects (ext, xfs) actually use a bunch o
263.
by dom0 9y ago
Cease your investigation
264.
by dom0 9y ago
No, this is about the EM field. The magnetic field of CRTs is very weak (so weak that it is dwarfed by nearby rail power lines which have a rather weak M field as well; so weak that you have to recalibrate for monitor movement relative to e
265.
by dom0 9y ago
https://en.wikipedia.org/wiki/Utility_%28patent%29
266.
by dom0 9y ago
Is such a thing even possible?
267.
by dom0 9y ago
The way things turned out the work from this era defined what computers can and can't do, not the other way around, which would likely not satisfy many mathematicians.
268.
by dom0 9y ago
> The GIF method is probably constantly leaking memory. If the browser only keeps compressed frames, 16 GB will last for more than a year.
269.
by dom0 9y ago
Interesting. Firefox' developer console does not understand this kind of request properly and shows its termination after the first frame. Chromium on the other hand handles it correctly.
270.
by dom0 9y ago
Google and Facebook are bigger than a single country now.

More ›