Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment: Re:Back to the future .. (Score 1) 126 126

No thanks. I don't want a VM that has mandatory built-in GC and a bunch of other nonsensical limitations. Provide me one which I can allocate and commit virtualized memory and a safe method for system access with all the facilities/opcodes of an ideal modern multi-core processor.

Comment: Re:Not relevant? (Score 1) 89 89

Do people not also see the connection between increases in privacy breeches and global warming? Exactly. Correlation != causation, otherwise we better stop going to see the doctors.

I think it's the other way around -- people are moving to the cloud because they see themselves as bigger targets and can't adequately protect their own systems compared to another company who spends billions hardening theirs. If the hackers can spend 2 seconds breaking into your system vs taking 2 years to penetrate Google's, I'm pretty sure they're going to infiltrate 500k servers like yours instead of bothering with Google's. This is probably also why CaaS is the big new thing now.

Comment: Re:Not bad parenting (Score 1) 446 446

Nowhere did Google or its representative say it's "bad parenting". Google is being asked one question "why are there much less females in tech?", and they're trying to answer that very question. Agreeing values or not has nothing to do with it.

Comment: Re:Power savings (Score 1) 98 98

I don't know why you're attributing this. One of the leading firms for 3D stacking is Tezzaron, and they've been around for more than a decade. You can give credit to AMD for taking on the risk of 3D memory, but not for innovating it -- other companies did most of that work.

Comment: Re:Not much of a debate... (Score 1) 161 161

Yes yes, I know your type. The ones that say "don't use anything the system was designed to be used, and only do things as if you're working in C". I.e. don't allocate all the time, don't modify the structure of the object after construction, make types easier to infer for JIT, etc.... That, or maybe you have 0 clue about what performance really means.

Comment: Re:Not much of a debate... (Score 1) 161 161

So your example of a "well done app" is a game that hardly uses any HTML5/CSS, and is all computation (which, given FFOS, I'm pretty sure is asm.js under the hood). I don't see how your example shows that web standards are actually uniformly implemented across browsers.

Comment: Re:Yes, Please!!! (Score 1) 161 161

Except:
1) Mobile CPU/GPU is still 50X slower than desktop (i.e. your CSS transitions are 50X slower...hardware acceleration included).
2) Mobile RAM allowance is still 10X lower than desktop (i.e. less resources loaded, prefetching, more thrashing, less JITed code, less unboxing, etc...).
3) No native look and feel, so everything seems out of place.
4) Still takes 3 years to get anything through the JS committee, and god knows how long before browser adoption.
5) Users still can feel latency even if backend is optimized (queue post about X% drop in latency equates to Y% increase in app usage/engagement).
6) Did you know that FB was originally all HTML5 on mobile? Did you know solely because of that decision, they were at most two quarters away from going belly up -- until they shipped their mobile app which turned everything around?

The fact that you brought up desktop being your primary client goes to show how little regard you have for the problems on the mobile platform.

The generation of random numbers is too important to be left to chance.

Working...