deraadt@enme3.ucalgary.ca (Theo Deraadt) (08/08/89)
I find that one of the greatest things about using Unix over Amiga is that everything in the entire system is size independent. Generally there are no maximum sizes in the machine anywhere. If there are, they can be changed easily. Look at how easily and invisibly the maximum number of file descriptors for a process changed. See how select() nicely took care of it in SunOS.. in the Amiga to do anything of the kind you have to play with system internals. It's simply disgusting. There are many wonderfull things in Unix, and for some unexplainable reason, a great deal of bias against it. I don't know why.. I don't remember who said it on the net a while ago, but it was something like "Those who ignore Unix are doomed to repeat it, badly" Looks like a classic case. <tdr. Theo de Raadt (403) 289-5894 Calgary, Alberta, Canada
shadow@pawl.rpi.edu (Deven T. Corzine) (08/10/89)
On 8 Aug 89 09:27:00 GMT,
deraadt@enme3.ucalgary.ca (Theo Deraadt) gets everything bass-ackwards:
Theo> I find that one of the greatest things about using Unix over
Theo> Amiga is that everything in the entire system is size
Theo> independent.
WRONG. One of the worst things about Unix is its extensive use of
static arrays, system structures, etc.
Theo> Generally there are no maximum sizes in the machine
Theo> anywhere.
There are maximums all over the place. Maximum number of processes,
maximum number of file descriptors, maximum length of arguments, etc.
Theo> If there are, they can be changed easily.
Yes, many are configurable, but it doesn't change the fact that
they're static. You still need to modify the kernal and restart the
system to change configurable maximums.
Theo> Look at how easily and invisibly the maximum number of file
Theo> descriptors for a process changed.
("no maximums?") Yes, the maximum number of fd's under SunOS 4.0 is
now 64 instead of 32. A static table, wasting space for processes
which need little or no I/O, limiting those which need lots.
Theo> See how select() nicely took care of it in SunOS..
SunOS 4.0 changed the semantics for select(), providing macros to hide
the actual implementation details from the programmer, and to make it
easier to extend the maximum again later. "Nicely?" Well, nicer than
forcing the programmer to deal with it, but still forcing code changes
to use >32 fd's for select(). It was a good move, in any case.
Theo> in the Amiga to do anything of the kind you have to play with
Theo> system internals. It's simply disgusting.
Exec I/O you can only have 32 signal bits but can assign the and group
them as you like. There is NO limit on open files...
Theo> There are many wonderfull things in Unix, and for some
Theo> unexplainable reason, a great deal of bias against it.
Granted. I don'w understand it either, but it is there.
Theo> I don't know why.. I don't remember who said it on the net a
Theo> while ago, but it was something like "Those who ignore Unix are
Theo> doomed to repeat it, badly" Looks like a classic case.
Probably some element of truth there. (or cliche, at least. :-)
Deven
--
Deven T. Corzine Internet: deven@rpi.edu, shadow@pawl.rpi.edu
Snail: 2214 12th Street, Troy, NY 12180 Phone: (518) 271-0750
Bitnet: deven@rpitsmts, userfxb6@rpitsmts UUCP: uunet!rpi!deven
Simple things should be simple and complex things should be possible.