Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Computers




OK.. this is not Physics.. if you want to discuss Physics.. delete
this message now... and if you are getting tired of this thread
I apologize for one more message *not* about Physics. :->

Now.. with that out of the way let me just say:
"Paul" == Paul Camp <pjcamp@coastal.edu> writes:

Paul> NeXT does have some rather serious technical drawbacks.

Paul> An interesting perspective on the NeXT OpenStep OS comes
Paul> from a company that used to make heavy use of it, Trilobyte
Paul> software, makers of multimedia games. They used the OS due
Paul> to its heavy emphasis on object-orientation so complex
Paul> multimedia projects could be easily assembled out of large
Paul> groups of self-contained multimedia-handling subsystems --
Paul> divide and conquer. That and Display PostScript mader NeXT
Paul> desirable. However, they have scrapped most of that,
Paul> retaining only a few high-end HP RISC-stations for data
Paul> compression tasks. The reason is that OpenStep is SLOW. It
Paul> requires some pretty hefty hardware to get effective use out
Paul> of the system and the company was unwilling to put $25,000
Paul> workstations on every desktop. Furthermore, due to lack of
Paul> resources, NeXT was apparently never able to advance the
Paul> object-oriented development environment as far as the
Paul> developers needed it to go.

OK this is almost surely a response to the MacWeek article:

http://www.macweek.com/mw_1108/ga_openstep.html

Where they say among other things:


"One of the big advantages to NeXT, and the primary reason we used it
originally for our development tools and assembly, were the large
groups of self-contained multimedia handling subsystems, such things
as the image-handling library," said Roy Eyman II, a Trilobyte
software engineer.

"It made object-oriented subassembly of large products a snap," Eyman
said. "You could slap things together quickly and easily." Code
objects were written rapidly and reused to handle typical operations,
such as loading and decoding objects within a multimedia program.
That streamlined the grunt work in the development process, letting
Trilobyte engineers meet tight schedules and still have time for
cutting-edge experimentation.

Eyman also lauded the networking capabilities in OpenStep. "It's an
almost invisible networking system," he said. "You could put a bunch
of NeXT machines on the same network and have them talking to each
other right away."


The presence in the OS of preassembled networking modules for programs
is useful to companies developing content for online use, Eyman said.

Display PostScript also made multimedia projects easier to put
together. "I know we used Display PostScript for a lot of our
on-the-fly text generation," Eyman said. "We would 'snapshot' the text
combined with graphics to produce a final frame that was then
compressed with our proprietary compression software."


Then there was the bad part...


"There were a couple of drawbacks," Eyman said. "The bigger one in my
mind is that the OS is slow. It requires some pretty strong hardware
to really get effective use out of the OS."

NeXT's own hardware was somewhat expensive, he said, and to run the
Intel version of OpenStep on 486-based machines, "you had to have such
a beefed-up 486 that it wasn't really worth it."


A beefed up 486? He's talking about a bygone era (two years ago
:->)... when memory was expensive.. and folks used 486s. With proper
memory, and a fast processor I can't tell the difference between
Openstep and 95/NT. With (only) 24MB my 040 NeXTStation even competes
with my 133MHz Pentium pretty well (and it has 32MB).

I've been watching this thread in a group about Physics, thinking to
myself.. 'hello?' and trying to keep my mouth shut. I don't really want
to fuss about it, but how about we just wait and see rather than
posting vague quotes about how slow this stuff ran on old hardware
with too little memory.

sorry again,
(now, hopefully, back to some real Physics. :->)
-steve