2009/2/19 Huan Truong <huantnh at gmail.com>:
> On Tue, Feb 17, 2009 at 8:14 PM, Jean Christophe Andr?
> <jean-christophe.andre at auf.org> wrote:
>>> I don't think that is the Microsoft way of thinking,
>> I was talking about giving no other choice but buying new hardware.
>> That's the Microsoft way, for sure. With GNU/Linux you have multiple
>> options, and some cheap/light ones among them. Is there a light
>> version of Vista somewhere?
>
> The problem is not Vista demanding more and more, the problem is that
> MS don't give a shit about people wanting to have old stuff, e.g. they
> don't sell XP anymore, they don't provide updates to XP anymore. So
> they have no choices but upgrading the hardware. But if you find KDE4
> demanding too much, you can stay with 3.5 and the KDE guys are cool
> with that I think - they allows you to keep your old PC - So what's
> the problem? Almost everything requires more and more hardware
> resources and there's obviously a natural thing. Speaking of software
> for everyone, why *do* I have to design my *new* OS/Wm/whatever for PC
> to work with a 128 MB of RAM, while only say 1% of people out there
> are having < 128 MB of RAM on their PC?
>
>>> I also think that in general, accelerate stuff by using powerful hardware
>>> is a 'more efficient' idea than trying to twist your code to bits to speed
>>> things up.
>> I strongly disagree on that: here you are encouraging bad code quality
>> to get more "productivity" (in the bad, commercial, sense of this term).
>
> I don't know how can you come up with "being bad" with "getting things
> done without having to care much about performance." Python and Java
> and Mono for example, are obviously slower and takes a hell lot of
> space compared to C++ programs, but it doesn't directly implies that
> programs written in Java or Python and Mono are bad code. It just get
> things done faster, and if you don't care much about performance, it's
> the way to do it. In other words, I don't think there are any direct
> argument to prove that programs that requires more powerful hardware
> to run are badly coded.
>
>> Modern machines are more and more power consuming, which means
>> using more and more energy and, at the end, natural resources.
>> So, in the end, you get higher and higher cost?
>
> Actually modern machines are more and more 'green'. They consume less
> power. You can try by comparing a 33Mhz workstation 15 yrs ago with a
> pentium 4 5 years ago and with an Atom machine today. A typical 19"
> CRT would consume like 95 ~ 120 watts, a typical 22" LCD consumes less
> than 60watts. Don't tell me that is higher cost because it isn't.
>
>
> --
> Huan Truong
Hmm, I think we are missing the point here. It is not about using
tools that are easier to have greater creativity, it is about
something else:
- First, while computer resource is pretty cheap today, a program
still should not assumed that the system it is running on is strong
enough to use all its 'feature'. Ok, I have 2GBs of RAM, and even if
Firefox consume 500MBs of it, my system still run fine, but that is
wrong, eveyone can see that. The programs that I run the most is
firefox, mpd and gvim, but does that mean I feel okay if MPD eat
400MBs of those, and 1GB for firefox, 200MBs for gvim, even if my
computer is still strong enough to run all of them at the same time?
BTW, I have seen MPD ate 1.5GBs RAM before, it is a memory leak due to
bad handling of MP3 IDv2 tag. We all have standard for how much
resources a program should use, like no more than 50MBs for a media
player that has no features other than just decoding and output to the
soundcard.
- It is okay for a programs to have a lot of feature, but please let
me be able to turn it off if I don't need it. Everything should be
made as modular as possible (the Unix way?) and be able to use alone.
But too bad, Akonandi is one part of KDE API, you have to turn it on,
and that also mean you have to run an instance of mysqld.
Back KDE 4 case, it consumes about 480MBs of RAM, which is even more
than Vista (Vista only consume about 330MBs of RAM, 'superfetch'
turned off of course), .
--
:(){ :|:& };:
