🚀 go-pugleaf

I hate to engage *again*, but...

waltrip@capd.jhuapl.edu writes:
>
>	And, of course, we should note, Steve, that porting NeXTSTEP to NT
>	would permit it to run on the DEC ALPHA 8^)
>

Okay, goal:  get NeXTStep on Alpha.
Choices:  port to NT microkernel, reworking the bowels of the whole env. i
                                   OR
          port the Mach microkernel (relateively straightforward)
          (this assumes that NeXT is working to get the Mach 3.0 mk
           integrated in their env...  though even the monolithic
           kernel should be an easy port.)

Now, why would you go and do a boneheaded, expensive, and time-consuming
thing like ripping out the existing microkernel and reworking all the
lower-level interfaces in NeXTStep, rather than go and port a relatively
portable kernel?  If they went to NT, they wouldn't have the benefit of
visibility to kernel internals, and any time they needed information on
same they'd have to interface -w- Microsoft to get it!

Just wanting to be on Alpha is a *bad* reason to put NeXTStep on NT.
In fact, there is *nothing* that NeXT can gain by porting to NT that
can't be gained cheaper/faster/better by remaining with their own
kernel architecture, *EXCEPT* the supposed market advantage of being
able to say they're on NT.  Repeat:  the *only* reason to port NeXTStep
to NT is to *be* on NT.


'nuf from me.  Standard Disclaimers.


jb



--
-- Jeff Bone ------------------------------------------------------------
         NT...  Just say N0T.  (Viva UN*X!)  Standard disclaimer. 
------------------------------------------------------- jbone@dell.com --
                                                                       --