Re: 64 vs 32 lenny

[Mark Allums <mark@allums.com>]

Stefan Monnier wrote:
> > Well, I guess that is a matter of philosophy.  I did not think a bigmem
> > kernel was "stock".
>
> AFAIK "stock kernel" in this context means that I did not build my own.
> I just "aptitude install linux-image-2.6.28-1-686-bigmem" and go on with
> my life.
>
> > Also, we might have a discussion of the meaning of the phrase "in
> > practice".  Personally, my "practice" is to use a 64-bit system.
>
> Well, you said "but in practice, 64-bit is required", so if your
> "practice" is to use a 64bit system, then indeed your statement is true,
> tho vacuously so.
>
> > But I was wrong about it.  How does a bigmem kernel address the full
> > 4 GB?
>
> Via PAE, AFAIK,

In practice, I think most people would agree that the PAE method is 
unsatisfactory.  Okay, change "required" to "cleaner".  64-bit is a much 
"cleaner" implementation.  In my opinion.  It's only important if it 
touches many things, affects how you use it.  If you "just 'aptitude 
install linux-image-2.6.28-1-686-bigmem' and go on with my life", then 
it doesn't affect you in practice, so it's not important.  So I was wrong.

Mark Allums