🚀 go-pugleaf

In article <3vcn47$a2o@park.uvsc.edu>, Terry Lambert <terry@cs.weber.edu> wrote:

> kanefsky@datamagic.com (Steve Kanefsky) wrote:
> ] The vast majority of the expense of going to preemptive multitasking on    
> ] the Mac will be in figuring out a way to do it without having to
re-write    
> ] all the existing apps.  It's trivial by comparison to design it in from    
> ] the start.
> 
[snip]
>
> When will vendors learn that people running their old apps on
> a new OS aren't buying replacements for their old apps?
> 
> Consider that Novell makes their money selling ervers, but
> their client API libraries are capable of talking with their
> legacy systems.
> 
> As long as these APIs exist, third party programmers will use
> them for portability reasons, which will in turn keep the NetWare
> server code the hulking-behemoth-without-a-logically-complete-
> set-of-APIs that it is today.
> 
[snip]
> 
> Grandfathering is OK up to a point, but eventually you have to
> bury your dead and move on.

And Apple does do that. Copland allows support for PM, Gershwin will
require it. Apple is bringing out a fully native OS for Powermac only.
They are burying 15-18 million machines with that move. It is not
reasonable to force people to buy new apps overnight - it is reasonable to
give developers at least 12 months to go through a normal revision cycle
and expect that people move at that point. 

Apple has routinely told developers not to expect certain things to work
at given points. Lots of guidelines were provided when the PMacs came out
- many API's have been buried over the years.

-- 
Bob Cassidy
UC Irvine