kornel c wrote:
>
> Dear Nathan,
>
> >Be careful of getting too hung up on the numbers. Memory management is a
> >bit of an art, and different JDKs take different approaches to fitting
>
> It wasn't me who got hung up on the numbers, it was my OS. Hey, I wouldn't
> have even noticed the memory requirements for long had it not been for
> my machine getting disabled.
>
> In my contracts my clients would usually not accept such waste of memory.
> Why should we accept it from Sun?
Ah, well there you're getting into a really big question. What we accept
from Sun - Java's memory and performance problems, its iron-fisted
control of the "standards" aspect, its treatment of partners, yadda,
yadda, yadda - is the price we pay for working in this really
interesting area. We can always choose to vote with our feet.
But, such considerations aside, it is useful to understand what problems
are inherent versus what are implementation details. Some of Java's
annoyances, like this need to "tune" your memory footprint with the -Xmx
option, are implementation details. Granted, they're ugly details and
they've been around too long and they give us all a lot of headaches.
But they're fixable, if not by Sun then by someone porting their code or
implementing independently. Either they'll get fixed before users get
fed up, or Java will go down the drain.
Nathan
----------------------------------------------------------------------
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
