heijmans
Joined 16 years ago
Karma
9
Submissions
18
Member since
Aug 27
Recent submissions
Comment · 5 months ago
Comment · 5 months ago
We are talking about the benchmark game here. The code for different languages was written by different programmers. For some languages, there were multiple implementations and many iterations. In this case, the researchers used the fastest one.
Some of the implementations are extremely optimized and took a lot of effort. Some implementations are not. So you might be comparing highly optimized JavaScript code with naive, below-average TypeScript code. You cannot compare those.
It would be much better if they used the same level of optimization for each language, but they didn't. Furthermore it is called a "Game" (it used to be called a "Shoot-out") because you shouldn't take it seriously. So it shouldn't be the basis of serious research.
Comment · 7 months ago
Here is small example of ChatGPT giving a wrong answer without expressing any doubt (aka bullshitting):
https://chatgpt.com/share/69d78ec7-67b0-8395-9fd1-522b760ab5...
GPT-5.4 Thinking, Pro account.
Comment · 4 years ago
Unfortunately, this will not be open source:
https://github.com/swc-project/swc/issues/571#issuecomment-7...
Comment · 4 years ago
For more info, look at:
https://github.com/chalk/chalk/releases/tag/v5.0.0
https://gist.github.com/sindresorhus/a39789f98801d908bbc7ff3...
And for the problems with esm and TypeScript:
Comment · 4 years ago
See also https://github.com/npm/cli/issues/2701
I think this is quite a serious issue which has been open for almost a year. I don't understand why there hasn't been a reaction from the npm developers on the issue (as far as I can see). npm 8.3.1 (the current version) is still vulnerable.
It might not be directly exploitable but it can leave you open for all kinds of security and/or stability issues. It is also a regression from npm 6.
Comment · 4 years ago
It was actually reported on (or before) Jan 3rd.
Comment · 13 years ago
It seems that they will release their software as open source. I couldn't find what license famo.us will use.
Comment · 13 years ago
They are using CSS3 3d-transforms to create visual interfaces with good performance on modern browsers (including mobile). See:
(BTW: It took me a while to figure out that if you close the dialog window, you can click the buttons at the bottom for more demo's.)
Unfortunately they are only using the webkit-prefixed properties, so their demo doesn't work in Firefox or IE. I'm guessing they wrote their demo a couple of years ago and never updated it.
There are some problems with CSS3 3d-transforms: 1) Text looks blurry, just look at the dialog window in the demo. 2) Some mobile devices (e.g. old iPads) have limited GPU memory. If you use too much (by setting 3d-transforms on large surfaces), the browser (or your app) will crash.
By the way, the answer to the question in the title is: No. And if they don't open source their code, they won't get a lot of traction.
Comment · 13 years ago
The documentation is scattered, but the best information can be found at:
https://github.com/cocaine/cocaine-docs-en/wiki
https://github.com/cocaine/cocaine-core#readme
https://github.com/cocaine/cocaine-core/wiki
The docs state that they refer to the 0.9 version, while development is going on in the 0.11 branch, so the docs might be more of less out of date.
It seems that they have already implemented a large part of the Heroku infrastructure, including service discovery, auto balancing, cross-language events (using ZeroMQ) and that they are using it in production. All in all, an impressive feat!
It would be interesting to hear if anyone has experience running/using this outside of Yandex. The source code seems well written, though comments are quite sparse.
As Lazare remarked, it seems to directly compete with flynn.io. It is good to see that different high-level platforms are created based on docker. But it would be nice if these platforms would consist of modules (for example service discovery or messaging) that can be used without using the whole kitchen-sink. I am not sure if this is possible with cocaine.
BTW, the repo cocaine-core is quite a bit older (since 2011) than docker (since early 2013?). The docker-core readme states that docker support is "on it's way" so it is not clear how mature this is.
P.S. As adults and hackers, can we look beyond the name at the technology presented here?
Comment · 13 years ago
This is very impressive. The code is very readable, even the asm.js based code (which is probably written by hand). You can find the CPU code in js/worker/cpu.
On Chrome I get about 30 MIPS while running scummvm. On Firefox I get about 6 MIPS on the standard core and about 40 MIPS on the asm.js core.
I must say that I am impressed by Chrome performance on the standard core. However, sometime Chrome drops to about 6 MIPS for no discernible reason.
For people like me, who didn't know what OpenRISC is: http://en.wikipedia.org/wiki/OpenRISC
But of course, a TypeScript program could be much slower than a different JavaScript program, as is probably happening in this case.
I agree with others that the difference in results between JavaScript and TypeScript makes the quality of this paper highly suspect.