Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
faragon
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
75 ms
·
751.
▲
by
faragon
11y ago
I remember that in 1994-1995 Linux was a shock: you were able to run for free in your personal computer a system that did similar things to SCO Unix or Coherent (Unix clone) without the need of pirating it (students were used to pirate SCO
752.
▲
by
faragon
11y ago
Thank you for the explanation. That is not really a problem, as there is information about drives known to work (you can reach it e.g. via Google). Also, USB pendrives are cheap, so just trying with pendrives you already have it is probable
753.
▲
by
faragon
11y ago
Compare how thin is the C book from Kernighan vs the C++ book from Stroustrup. The result is that most people can fully understand C easily (if willing to), while the C++ case is impossible. C is beautiful.
754.
▲
by
faragon
11y ago
That router uses as CPU a Cavium Octeon processor (MIPS architecture) with two cores. It supports MIPS 64 r2 and aligned data access, and it is a great platform for running big-endian unit tests. It boots from a internal USB drive, that can
755.
▲
by
faragon
11y ago
Wow. That's amazing. Anyone knows how they do a backup/restore of a GFS distributed file system? (in case they do backup of some sort)
756.
▲
by
faragon
11y ago
In my opinion, DRM will "fix" that in the future: browser plugins could not be able to identify those ads. So we could reach "Black Mirror"-like ads sooner o later. Brave new world...
757.
▲
by
faragon
11y ago
Why? Is Oracle "sacred" or something?
758.
▲
by
faragon
11y ago
"Don't bite off more than you can chew", anyone?
759.
▲
by
faragon
11y ago
Thank you! :-)
760.
▲
by
faragon
11y ago
You can consider also set system-independent locale, e.g. "set_turkisk_mode" (I had that problem, too), etc. I thought that the only case conversion exception as the Turkish case. Do you remember which cases are an exception for G
761.
▲
by
faragon
11y ago
Is the Turkish case conversion working? edit: it seems it is: https://bitbucket.org/knight666/utf8rewind/pull-requests/1/t...
762.
▲
by
faragon
11y ago
How vs why I started learning genetic algorithms (random permutations selected/promoted for picking better combinations of Hamiltonian paths for solving NP-C problems obtaining near-optimal solutions quickly) and neural networks (m
763.
▲
by
faragon
11y ago
You can buy services with a credit card that are legal in one country and not in a different one. Theoretically, you can not be punished in your local legislation for operating in other countries. E.g. you can buy/rent servers in Hong
764.
▲
by
faragon
11y ago
No. I mean services for world-wide audience (currently you need lots of datacenters ), not for a local thing. If you don't need high performance, go higher level, of course.
765.
▲
by
faragon
11y ago
> C cannot - there is apparently no way that is both reasonably efficient and reasonable foolproof to write C code; if you have large scale examples written by more than one person[0], I would love to see them. Not easy. Sure. In that re
766.
▲
by
faragon
11y ago
The problem of "Unix Philosophy for everything" is forgetting that involves multiple buffer copy and parsing. So if you're able to reduce the space problem at first step(s), you can reduce the problem, being that method effi
767.
▲
by
faragon
11y ago
> are slow even with native code, in the vast majority of cases, and FastCGIs are fast enough with properly written Python/Perl. (...) Yes, CGIs are slow because of the process spawn. About FastCGIs, if you want 10-100x throughput
768.
▲
by
faragon
11y ago
The dream could come back. In my opinion. Defeating legal threats: - Design server architecture so can be migrated from one country to a different one without censorship. Now is easy, with VPSs, you can move infrastructure in few hours, if
769.
▲
by
faragon
11y ago
"That storage location itself must be allocated." Not really, e.g. you can allocate with one allocation both: header + the buffer for the data. When growing, you grow with: size_of_header + elem_size * num_elems (if allocated in t
770.
▲
by
faragon
11y ago
Because "shrink_to_fit" involves calling a "realloc"-like function ("realloc" with a lower value is the best way for returning unused memory back to the system when you have a buffer with too much memory alloca
771.
▲
by
faragon
11y ago
Thank you very much for taking a look into the code :-D Regarding identity, you can use double pointers. In the case of willing to eliminate leakage risks (or speed-up the operation, avoiding calling the heap allocator), you can use stack a
772.
▲
by
faragon
11y ago
The C++ STL vector, as is, has some limitations: - As far as I know, it requires two calls to the allocator (one for the instance, and another for the resizable buffer). - Even if allocated in the stack, unless you use a different allocator
773.
▲
by
faragon
11y ago
Yes, that way is for C99 and later.
774.
▲
by
faragon
11y ago
Questions: - Does it support unaligned access hardware exceptions for ARM, MIPS, and Sparc? - Documentation and download links are broken. Are you going to fix it?
775.
▲
by
faragon
11y ago
No matter if you use signed or unsigned types, be sure you handle both underflow and overflow . From my experience, signed integers main usage is integer arithmetic (e.g. accounting things that can be negative or positive), pointer arith
776.
▲
by
faragon
11y ago
I think this is not the memristor thing. I remember memristor-related posts in Slashdot (2008 [1], 2012 [2]), but the ones related to Intel + Micron (2014 [3], 2015 [4]) tell about "3D nand", which may be just NAND-flash using mul
777.
▲
by
faragon
11y ago
VLIW processors with dynamic configuration of the opcode instruction group has a lot of potential in the future, in my opinion. E.g. 128 or 256 bit "opcode groups" with variable-size opcodes, using their own temporary registers, e
778.
▲
by
faragon
11y ago
I stopped commenting/voting on YouTube since the mandatory G+ thing. I hope Google think it twice before playing that hard again.
779.
▲
by
faragon
11y ago
So they're going to cut their CPU R&D and become a generic ARM IP integrator, plus their GPU and radio-frequency circuits (Adreno GPU, GSM, WiFi, GPS). Next moves could be discarding their own GPU for the ARM generic ones (Mali) an
780.
▲
by
faragon
11y ago
That's a huge effort, kudos. Certainly, it is harder to get maximum performance from C, as sometimes you have to reorder things in non-trivial loops in order to "help" the C compiler to get it "right" in the perform
More ›