🚀 go-pugleaf

In article <Bvy9vq.6s4@world.std.com> gsk@world.std.com (Geoffrey S  
Knauth) writes:
> In article <BvtoDH.Lzz.2@cs.cmu.edu> beatty+@cs.cmu.edu writes:
> > NXHost over SLIP works so badly that there must have been an  
intentional
> > decision to make it very hard to even try.
> 
> I agreed with everything else you wrote, but not this.
> 
> As far as I understand, NXHost is slow because of all the Display
> PostScript being transmitted from the remote application to your
> screen.  NXHost bears no malice.

Perhaps I wasn't sufficiently clear.  Let me try again.  I wasn't trying  
to say that NXHost over SLIP was made slow intentionally.  I was trying to  
say that it might be deliberately hard to make it work at all.  So far as  
I know, to NXHost over SLIP, you have to have the same IP address for both  
the ethernet and the slip interfaces, and to make that work, you have to  
rather carefully adjust the routing tables, in the case of a standalone  
machine.  Making it hard to set up NXHost over SLIP (not the speed) could  
have been a design decision at NeXT, Inc.  If they purposefully made it  
hard to set up, then people would tend not to do it, and hence not  
complain about how poorly it worked.  It's rather like disabling access to  
the Help files in Icon.app in 2.X: the files are there, you can read them  
if you like, but since you can't get to them just by clicking on 'Help' in  
a menu, you know that they are likely to be an inaccurate description of  
the program, i.e., you shouldn't be doing this.  It's also like making it  
hard to power off without syncing the disk: sure you can pull the plug,  
but you know that's not the preferred way, since there's a power button  
right on the keyboard.

 -- Derek
-- 
Derek_Beatty@cmu.edu   ABD   Comp Sci, CMU, 5000 Forbes, Pgh, PA 15213 USA