Re: Extended Volume Format (HFS+)

[Jim Murphy <murph@apple.com>]

In article <69c748$h2h$1@darla.visi.com> Nathan Mates,
nathan@visi.com writes:
>I'm speaking facts here as to what the GS can do, what good code
>design is, and you're just coming up with bullshit after bullshit.
>When you've actually learned with good code design is, and how it
>works in the real world, then you might be able to comment on it.

What *I* think here (and make no mistake - this is what I think,
not what Apple says, etc.), is that Nathan hasn't much (any?)
practical experience with actually designing and implementing bits
of an OS for commercial release. When we (OS designers) are asked
to build something, we usually have a lot of factors that influence
what we ultimately get to build. From schedules, to resources, to
what the customer actually wants, all these go into the decision-
making process that gets used to build a product. We physically
*can't* do everything everyone wants.

In the case of device drivers and FSTs (as have been mentioned), I
would be apt to believe that a certain number of resources were
allocated, deadlines set, and then priorities scheduled. Priorities
at the time likely said that device drivers were more important
because there are bazillions of Apple II cards, and therefore must
have a clean public API (API == application programming interface).
FSTs were less-so because there was a relatively small number of file
systems in the world, and Apple believed rightly or wrongly that it
could supply the demand. This is simply my slightly-educated guess.

This doesn't mean the FST design is bad like Nathan keeps ranting
on about, it's just not entirely suitable as an API. From what I
know about it, it's more of an SPI (system programming interface).
That is, you usually need to source to the components that FSTs
talk to to fully understand how they work and interact with the
OS. If Apple had infinite resources, it could certainly have made
the interface a bit higher-level so that the FST interface could
have been published. But that didn't happen for likely a variety of
reasons.

Nathan can disagree until the cows start flying, but I'm more apt to
believe that the FST design was due to constraints on time and
resources rather than incompetence.

>worship of Apple as evidenced by your trying to take insults to the
>design of GS/OS personally shows you couldn't comment rationally on
>anything out of Apple without shit coming back out your mouth from
>your constant asskissing of Apple.

Nathan? How old are you? I'm serious. You spew the requirements of
logic and facts, but then crap like this constantly comes out. csa2
would be a heck of a lot nicer if *you* went without. Pot, kettle,
black and all that.

(Back to working on OSes and lurking for a few more years.)


Jim Murphy                                   Blue Box Runtime/MacsBug
murph@apple.com                                  Apple Computer, Inc.