Is Gnome becoming a Microsoft project?

[Ian Malone]

On 21 July 2012 03:43, Chris Tyler <chris at tylers.info> wrote:
> On Fri, 2012-07-20 at 11:07 -0700, Joe Zeff wrote:
>> On 07/20/2012 05:25 AM, Tim wrote:
>> > Whether that be Java, or Flash, or anything else.  It seems far too
>> > common that a programmer would not care that their program is
>> > ridiculously inefficient to the point that multitasking becomes nearly
>> > impossible.
>>
>> Very few programmers, any more, are taught to be frugal with memory.
>> They have no idea how to control memory use and no understanding of why.
>>   Most developers use computers with lots and lots of memory for their
>> work, so they never see the results of their careless attitudes.  As
>> long as *they're* not running short of RAM, it's not a problem as far as
>> they're concerned.  And, if mere users complain, they're response is,
>> "It works for me."
>
> This is why I'm hoping a lot of devs get a Raspberry Pi (256MB split
> with the GPU, 700MHz) and run their own code on it :-)

It's not quite this straightforward. I do a reasonable amount of
scientific computing, which is pretty simple from an architecture
point of view, but even then you don't have some sliding scale which
says this program will use X amount of memory vs take this many CPU
cycles. You might be able to guess where the tradeoff is, and make
estimates for how much memory is going to be required, but you may
still find the amount of memory required is too much for the machine.
Or that it's acceptable but as a result the program is going to take a
month to run. Or that there's no acceptable compromise and you have to
adapt the underlying toolkit to allow a better way of handling the
problem. Or that it's just going to take those resources to run and
there's no significant improvement that can be made. And at each of
those steps you have to refactor what you've written, there are no
sliders to drag around to do it for you.

-- 
imalone
http://ibmalone.blogspot.co.uk