Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.javascript > #14697
| From | Stefan Weiss <krewecherl@gmail.com> |
|---|---|
| Newsgroups | comp.lang.javascript |
| Subject | Re: Detect user graphics card? |
| Date | 2012-06-27 01:05 +0200 |
| Organization | albasani.net |
| Message-ID | <jsdf79$c8m$1@news.albasani.net> (permalink) |
| References | <e297bfd5-2f14-4f87-b6b4-92dc73c8b83d@r3g2000yqh.googlegroups.com> <js719u$7j2$1@dont-email.me> <4454122.7EdHYbU6jS@PointedEars.de> <jsc6bu$o4k$1@news.albasani.net> <2493601.R5VuAm1OKa@PointedEars.de> |
On 2012-06-26 22:38, Thomas 'PointedEars' Lahn wrote: > Stefan Weiss wrote: >> A half-hearted benchmark indicates that the last version could be >> between 2x faster (Chrome, Opera) and 5x faster (Firefox) than the >> original. > > AISB, I am not that interested in benchmark results as they are treacherous. > I am more interested in provably working, efficient algorithms. I only mentioned that because you seemed concerned about the overhead of function calls (in a different thread). I've also underestimated the efficiency of the current compilers before [1]. These days it's hard to predict which algorithms will be more efficient than others, let alone "provably efficient". It depends too much on which optimizations the engines make. Without intimate knowledge of the compiler internals, the best we can do is measure the alternatives; and the best we can derive from that is some rough idea of efficiency in the tested engines. Normally, I wouldn't have bothered, but a function which converts color values could actually be called many times in succession (for example, to manipulate an image-as-pixel-array). - stefan [1] There was a thread in this group a few months ago (I can't remember the exact title), where different alternatives for Math.floor() were mentioned. I would have bet a tidy sum that one of alternatives using binary operators would be fastest, and I would have lost. IIRC, Scott Sauyet did a benchmark and found out that Math.floor() is apparently very well optimized in some engines.
Back to comp.lang.javascript | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Detect user graphics card? justaguy <lichunshen84@gmail.com> - 2012-06-24 04:31 -0700
Re: Detect user graphics card? Gregor Kofler <usenet@gregorkofler.com> - 2012-06-24 14:30 +0200
Re: Detect user graphics card? justaguy <lichunshen84@gmail.com> - 2012-06-24 07:43 -0700
Re: Detect user graphics card? Thomas 'PointedEars' Lahn <PointedEars@web.de> - 2012-06-25 05:14 +0200
Re: Detect user graphics card? justaguy <lichunshen84@gmail.com> - 2012-06-24 21:01 -0700
Re: Detect user graphics card? Stefan Weiss <krewecherl@gmail.com> - 2012-06-26 13:27 +0200
Re: Detect user graphics card? justaguy <lichunshen84@gmail.com> - 2012-06-26 12:54 -0700
Re: Detect user graphics card? Thomas 'PointedEars' Lahn <PointedEars@web.de> - 2012-06-26 22:38 +0200
Re: Detect user graphics card? Stefan Weiss <krewecherl@gmail.com> - 2012-06-27 01:05 +0200
Re: Detect user graphics card? Scott Sauyet <scott.sauyet@gmail.com> - 2012-06-27 07:07 -0700
Re: Detect user graphics card? "Michael Haufe (TNO)" <tno@thenewobjective.com> - 2012-06-29 16:41 -0700
Compute web-safe colors (was: Detect user graphics card?) Thomas 'PointedEars' Lahn <PointedEars@web.de> - 2012-06-30 10:40 +0200
Re: Detect user graphics card? Dr J R Stockton <reply1226@merlyn.demon.co.uk.not.invalid> - 2012-06-27 19:44 +0100
Re: Detect user graphics card? Arno Welzel <usenet@arnowelzel.de> - 2012-06-27 17:57 +0200
csiph-web