[rescue] Why is everyone so OSX happy?

[Dave McGuire]

On Friday, April 11, 2003, at 09:45 PM, Patrick Giagnocavo 
+1.717.201.3366 wrote:
>>> Indeed.... many disk performance problems go away when you add more
>>> RAM (up to a point, of course).
>>
>>    ...then it wasn't a disk performance problem in the first place!
>
> Not necessarily, if having more disk caching in RAM allows you to
> re-order the disk writes and thus reduce seek time.

   I have other uses for my RAM than acting as a crutch for a 
poorly-designed disk subsystem.

   Hmm...further, I'd rather just have my drives reorder their writes.  
Oh, I forgot...IDE isn't smart enough to do that yet, is it?

   Shit, I had an SMD disk subsystem on a VAX nearly fifteen years ago 
that could do that.

> Note the caveat - perhaps Apple's IDE drivers suck.  Wouldn't put it
> past them since it gives them market differentiation.

   Ok, I could buy that.  But what about NetBSD...FreeBSD...Linux..on 
Alphas and x86 boxes alike...all of which I've seen perform better with 
a real disk subsystem?  My observations are based on a lot more 
observation than dicking around with OS X for the past few 
months...otherwise I wouldn't take such a strong stand about it.

> Anyone here got a dmesg from an OSX box with IDE drive(s)?

   My TiBook has an IDE drive...I don't see anything in the message 
buffer or in syslog about it though...? :-/

           -Dave

--
Dave McGuire             "I've grown hair again, just
St. Petersburg, FL           for the occasion."       -Doc Shipley