Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
zurn
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
23 ms
·
211.
▲
by
zurn
9y ago
Where does even the 2x come from? Currently it looks very much like 1x vs integer hinted JS code like generated by asm.js.
212.
▲
by
zurn
9y ago
The article compares cloud costs to putting a little x86 server on a colo-provided 100 Mbps connection. I think you are not talking about the same kind of service, and also this reference price does include amortized cost of network engin
213.
▲
by
zurn
9y ago
Will this translate to better performance for languages that compile to Javascript and have so far used pre-ES6 constructs?
214.
▲
by
zurn
9y ago
It would be cool if such "GPU resistant" coins became popular, because it would encourage people to figure out new ways to program GPUs.
215.
▲
by
zurn
9y ago
What does the author mean by ASIC resistence here?
216.
▲
by
zurn
9y ago
Ironically Google's Yubikey implementation requires you to register a phone number to your Google account, which many people don't want to do.
217.
▲
by
zurn
9y ago
Jonathan Blow's Jai language lets you do this by annotating structs, along with some other assisted memory optimizations (pointer compression is another).
218.
▲
by
zurn
9y ago
The as-if rule allows compilers to perform this optimization, provided they elide the optimization for those structs that have addresses of members taken/used, though. This requires whole program analysis of course (as in clang /
219.
▲
by
zurn
9y ago
A bit of trivia: Computation is frequently pipelined so that part of the work is done during previous frame(s). This of course means that the game logic latency is longer than 1 frametime.
220.
▲
by
zurn
9y ago
The research side may not be the main culprit, hence I said "compiler engineering" - meaning usable tools delivered to the hands of developers. Though I think the research side has not focused on this enough either. Regarding the
221.
▲
by
zurn
9y ago
Maybe OpenMP and CUDA do more optimizations than I thought they did. Do they perform AoS->SoA reorganizations, conversions from pointer members to inline members, or range analysis to choose smaller data types? Admittedly foremost I was
222.
▲
by
zurn
9y ago
Compilers and languages of course co-evolve. And we already have fairly widespread production use of invasive language extensions/dialects designed to enable use of parallelism, such as OpenMP and CUDA. And older language extensions th
223.
▲
by
zurn
9y ago
TL;DR "Optimizing compilers are still good to have because they are cheaper than programmer labour needed for hand optimization" The original DJB presentation, which this is a response to, is very good and interesting. It would re
224.
▲
by
zurn
9y ago
Say hello to memory-unsafe languages and resulting vulnerabilities in web apps?
225.
▲
by
zurn
9y ago
This might also motivate better support for running ChromeOS apps under standard Linux.
226.
▲
by
zurn
9y ago
Also in most business contexts employees have no right to sign contracts on behalf of their employer.
227.
▲
by
zurn
9y ago
It's only so because the global warming externalities are not priced in. And it's a very bad thing.
228.
▲
by
zurn
9y ago
I'm sure the problem is tractable statistically, by normalizing some variables related to immigration.
229.
▲
by
zurn
9y ago
They can of course see the individual packets and their exact timings. This can be trivially correlated with egress traffic from your VPN gateway to other ISP affiliated networks or web properties. And since it's your private gateway,
230.
▲
by
zurn
9y ago
It's easy for anyone who can do traffic analysis on your traffic, eg your ISP and mass surveillance perpetrators. And whoever your ISP decides to sell or give this data to.
231.
▲
by
zurn
9y ago
This is based on Clang. No doubt they want to upstream the improvements.
232.
▲
by
zurn
9y ago
Note to people who aren't running Windows: this is unrelated to the article, which covers some interesting iommu virtualization and memory controller improvements in the latest firmware updates.
233.
▲
by
zurn
9y ago
An effective population-level way to address type 2 diabetes could be dietary feedback from glucose levels, for pre-diabetic people.
234.
▲
by
zurn
9y ago
Does Amazon or GCE support nested hypervisors?
235.
▲
by
zurn
9y ago
Most wired ethernet is probably deployed in business settings now, and business desktops are stagnant too. iPhones with choppy 3G/4G killed fast wired consumer broadband.
236.
▲
by
zurn
9y ago
I'm pretty sure I referred to open systems and CUDA in their commonly understood meanings. Here are some links that may help to clarify the concepts: http://www.pcmag.com/encyclopedia/term/48478/open-syst
237.
▲
by
zurn
9y ago
Apple is in dominant market position.
238.
▲
by
zurn
9y ago
Successful open systems tend to have standards that "just happened", not prescriptive standards. Like x86 PCs or Linux or most programming languages. Cuda seems to be clearly winning over OpenCL in the real world so other vendors
239.
▲
by
zurn
9y ago
If your team of programmers tried to build something like Servo in assembly, you would never complete the task because the task is so laborious and error prone. Finding and correcting all the bugs in your hand-written assembly isn't fe
240.
▲
by
zurn
10y ago
This kind of NAS should definitely automatically update and reboot rather than become wormable. A high-availability file server is a different beast from a home NAS.
More ›