[KLUG Members] Laptop Linux users

[Buist Justin]

Good point... perhaps I should re-think my ACPI implementation.  I was just following suite with the APM implementation.  Maybe I could spend some time this next week to fix -both- of them :)

Thanks, 
Justin Buist


> -----Original Message-----
> From: adam@morrison-ind.com [mailto:adam@morrison-ind.com]
> Sent: Friday, October 18, 2002 5:00 PM
> To: members@kalamazoolinux.org
> Subject: RE: [KLUG Members] Laptop Linux users
> 
> 
> >>>Keeping the fd open though dup()ing it and seek()ing back 
> to the beginning
> >>>would let you re-read the file though w/out that disk hit. 
> Little tricks
> >>>like that scattered about the core gnome apps might do a 
> lot to keep the
> >>>battery running a lot longer.
> >>Fascinating.  Maybe open in glibc is checking xtab or some 
> >>other status file?
> >Nah... it has to figure out how to open up proc first, so 
> you have to open up
> >a handle to / and then loop through it to find the 'proc' 
> entry, which then
> >begins your journey into memory-only space.  / lives on the 
> disk though, so
> >to get that inode open you've got to ask the little magnetic platter.
> 
> Nah... I don't buy it.  Path walking takes place non-stop, 
> opening any file goes
> through the same process.  The inode and contents of "/" are 
> almost certainly
> in-buffer,  and certainly wouldn't go out-buffer every five 
> minutes unless your
> starved for RAM, in which case your drive would be grinding.
> 
> Possibly bdflush could be adjusted to make it lazier.
> 
> _______________________________________________
> Members mailing list
> Members@kalamazoolinux.org
> 
>