zurn

53 ms·

511.
by zurn 11y ago
The single problem with TCP this rightly points out is head-of-line blocking. But overcoming that still leaves you with the problem of coping with out of order or missing events at app level, with clients and servers seeing events in diffe
512.
by zurn 11y ago
Apart from the functional differences, OpenMP is implemented in compilers and support is currently hit-and-miss even though mainstream compilers are slowly getting there. So eg targeting Mac, Linux and Windows would probably get quite &quot
513.
by zurn 11y ago
Is India's road network really suitable for high speed PRT as-is? If not, the question becomes how SkyTran track compares to new highway, per foot.
514.
by zurn 11y ago
Many programmers have fairly strong preferences about programming languages and definitely many people would consider a Go job more desirable than a Java job. Most of those people aren't currently Java programmers - it's said[1] t
515.
by zurn 11y ago
Note they cite higher numbers as the current status than the Cloudfare article. In the LWN article it says "The kernel, today, can only forward something between 1M and 2M packets per core every second" Also eg. http://
516.
by zurn 11y ago
Those are real-world speed bumps that make sense when you want to do stateful processing w/ large working sets, but not really for this "just get packets through the stack" toy benchmark that fits in cache and the workload is
517.
by zurn 11y ago
When 100M ethernet was introduced in the 90s, PCs had Pentiums, at < 100 M instructions per second. Both x86 CPUs and Ethernet have gotten ~1000x faster in the same time - cf 100G Ethernet and processors with 30x frequency, ~2x more IPC,
518.
by zurn 11y ago
It isn't as slow as they claim, just like was discussed in HN comments to the predecessor article ("How to receive a million packets per second"). Still they repeat the general claim of "Vanilla Linux can do only about 1
519.
by zurn 11y ago
Here's a debunking from one of the economists involved. http://www.theguardian.com/commentisfree/2008/aug/22/climate...
520.
by zurn 11y ago
I don't think it's especially useful to think about emissions by country for the purpouses of thinking who can effect change. Any more than a small town resident should feel discouraged about voting in national elections. People n
521.
by zurn 11y ago
Also those emissions are relatively easy to fix since it's mostly cheap and nasty coal plants. They're already sick of coal plants due to local pollution, and switching electricity generation tech is much easier than say changing
522.
by zurn 11y ago
Depends on how your firmware or OS sets up the memory protection hardware; see https://en.wikipedia.org/wiki/IOMMU
523.
by zurn 11y ago
Do we know that Apple announces (and gets CVEs for) vulnerabilities that they find and fix internally? If not, the comparison is not meaningful.
524.
by zurn 11y ago
The virtual addressing part might be fine now that the GPUs are doing it, and hypervisors support programmable passthrough using the x86 IOMMU (VT-d etc) features. Though I'm convinced custom hardware is doomed here for the usual custo
525.
by zurn 11y ago
In addition to the cyclic references problem, rc has a high constant overhead (memory traffic from constant manipulation of rc fields) and a problem with pauses (cascading deallocations). Also it precludes efficient shared memory concurrent
526.
by zurn 11y ago
Or possibly saves power, if work is being offloaded from the CPU.
527.
by zurn 11y ago
Firefox has a pretty good reputation among users and web developers, despite risks of hitting bad extensions. Their problem amont end users is more the relative obscurity vs the big 3 commercial browsers. IE on the other hand has been a swe
528.
by zurn 11y ago
This is the Racket library in question for context, it provides a framework for doing world simulation in a FP way. http://docs.racket-lang.org/teachpack/2htdpuniverse.html
529.
by zurn 11y ago
Just get the right hardware with good Linux support, it all just works then. You won't find any lasting happiness by digging deeper into the third party drivers swamp. (all-Intel for wifi and graphics is the safest bet)
530.
by zurn 11y ago
Yes, registers and the belt are just ISA features. Register renaming is a behind the scenes implementation tech to mitigate the problems of register semantics. It has unwanted costs that a belt machine doesn't have to pay, and it'
531.
by zurn 11y ago
The article says userspace ASLR is used, and indeed presented a speed bump along the way.
532.
by zurn 11y ago
Can you give references to these experimental results? I'm interested to know what they were measuring, was system they compared against etc.
533.
by zurn 11y ago
ATMs are actually a pretty fruitful point of comparison since people have a good idea of how ATMs work in their heads already. Some differences in the threat model come to mind: - ATMs and their backend systems invest heavily into making co
534.
by zurn 11y ago
Erlang co-evolved with OTP so they have a naturally close relationship. The Go situation (and philosophy) is much different. Also the Go authors sound like they don't want the language to be too closely associated with Google.
535.
by zurn 11y ago
But like ploxiln said you now need a conditional branch to test for the overflow flag after the arithemetic operation, unlike the old overflow traps on x86/MIPS/Alpha etc.
536.
by zurn 11y ago
The naive check-and-branch approach on current x86 only costs a couple of percent in integer-heavy and well tuned C code (see http://danluu.com/integer-overflow/ ). Integer operations in Python and Ruby are so glacially
537.
by zurn 11y ago
There's a performance cost now because processor instruction sets have dropped hardware overflow detection due to disuse. Current processors are largely engineered to just run legacy C code fast. See eg. https://news.ycombin
538.
by zurn 11y ago
Nukes are actually much better at load following than is commonly believed. It's used a lot in countries without significant hydro production, like Germany and France, and load following capability is a regulatory requirement in many&#
539.
by zurn 11y ago
NSA attacks most people. It's just selective about who it targets for special treatment after going through everyone's stuff.
540.
by zurn 11y ago
Good point. I guess in this way C++ has a quality similar to dynamically typed languages.

More ›