becoming thoroughly pissed off at ISC

[barry hannigan]

can anyone provide answers or bug confirmations for the following:

1. as of the introduction of "POSIX" compliance in 2.2, the include files
	appear to be shit. gcc, which compiled fine under 2.0.2, now requires
	the -traditional flag to eat the getc() and putc() macros. this among
	other things.

2. does the "none" option in setting up host based tcp signify "no card"
	or "no tcp". two socket based programs i tried refused connections, 
	with one passing an invalid address message back from t_bind(). ping
	and telnet work fine.

3.  the quality control in this company is abysmal. i can understand getting
	a bad floppy after a cross country flight maybe, but 12 unreadable
	diskettes in 2.0.2 and 2 identical disks labeled LPI C #1 and LPI C #2???

4.	at what breakpoint does ISC UNIX decide to swap the entire program out to 
	disk. compilation of gcc on a 4 meg machine showed 5 % cpu utilization
	after 1.5M swap allocated. (figures from u386mon, great program)

5. you cannot make a file system on 720k floppies. i dont believe you can even
	use them for tar. ISC support said "thats been a problem for a long time".

6.	when exactly does the new "jobs" code in the c-shell work. try tossing a 
	large compress job into the background. it ignores you.

7.	copies to bad media cannot be killed. 

sorry for the tirade

barry
-- 
~~~~~~~ ..walkin on the beaches, lookin at the peaches