[comp.windows.misc] My Opinion: NeWS vs X

gaynor@topaz.rutgers.edu (Silver) (02/14/88)

My qualifications in the windowing system/graphics areas, so you know
where I'm coming from, are not staggering.  I like to program, I like
to play with environments to make them better.  Graphics are fun.  I
think in terms of not only what is good for today, but what will be
good in the future.  When I'm not being paid to work on computers, I
do so for pleasure.

- I like the NeWS paradigm.  (What's a pixel? :-)  Nicely extensible,
and PostScript is a neat language.  The data structure of the physical
display is nicely abstracted away from the programmer/user.  I wonder
if Sun has considered adding a byte-compiling feature to NeWS?  Then,
after that (after picking up the pieces :-), an optimizer for the
byte-compiled code?  Also, I understand that there are some `features'
of the lite process manager which some consider undesireable.  Not
being familiar with this, I won't attempt to offer an opinion in its
regard.

- I like the ease of programming.  Routines defining other routines is
a standard concept/practice in computer science...

- For the above two points, I think the NeWS idea is a win for the
same reason GNU Emacs (and other programs which implement a language
to do the dirty work) won.

- For what you get in display quality and extensibility, I am happy
with the speed of the implementation.  Besides, machines are getting
faster...

- Although I am aware that NeWS is much less of a burden on a network
(orders of magnitude, I'm told), I don't think this is a point to do
any bashing over.  Networks are getting stronger, as was mentioned by
a previous poster.  Is there any merit to the following proposal: For
X, write a buffering routine which buffers calls to X routines into
chunks, and `write out' the buffer via a call to a buffer-outputting
routine or by meeting a condition.  (Hmm...  Does X allow more than
one message per packet?  If it doesn't, surely it wouldn't be
difficult to extend it to do so...  Haven't followed up on this one,
and don't intend to - it's just a thought.)

Oh well, I guess I'd better pull on my asbestos underwear, and protect
the important stuff.  (Indescriminant flamage to /dev/null, all else
to /usr/spool/mail/gaynor.)	 ___
				 \o/
			  Cheers, V [Ag]
				 _|_

 Andy Gaynor   201-545-0458   81 Hassart St, New Brunswick, NJ 08901
   gaynor@topaz.rutgers.edu   ...!rutgers!topaz.rutgers.edu!gaynor

gaynor@topaz.rutgers.edu (Silver) (02/14/88)

I wrote, somewhat uninformed as to the internals of X:
> Is there any merit to the following proposal: For X, write a
> buffering routine which buffers calls to X routines into chunks, and
> `write out' the buffer via a call to a buffer-outputting routine or
> by meeting a condition.

Apparently, this *is* the way things are done (or something real
similar).  So, I guess there was merit to the proposal.  :-)

[Ag]