Re: Code reviews [Was: FWTK and smap/smapd]

[Carson Gaspar <carson () taltos org>]

--On Friday, July 19, 2002 1:07 PM -0400 "Marcus J. Ranum" <mjr () ranum com> 
wrote:

> My experience with the firewall toolkit is somewhat dated, but if
> I were to guess, 95% of the emails I got to fwtk-bugs were from
> idiots who were too lazy to read the README or who didn't know
> how "make" worked. Of the remaining 5%, there was a breakdown of
> a smattering of genuine bugs or reasonable feature requests, as
> well as a few design queries. But, of the "10,000 eyeballs" effect,
> only a few dozen of those eyeballs were any good. And Carson was
> so darned insufferable about the patches he sent in that even though
> he was right I wished he'd go away. ;)

Ah, those were the days ;-)

Back then, I still had this bizarre belief that buggy code was the 
exception, and that fixes would be happily accepted and quickly integrated 
into the code base. Oh, how naive I was... and probably more than slightly 
obnoxious ;-)

Nowadays, I have far lower expectations. I have now interviewed dozens of 
"UNIX network programmers", and have had _one_ person ever get this 
interview question correct (taken from a fwtk bug, btw):

Given a buffer char *buf, a buffer length unsigned int buflen, and a 
connected TCP session on file descriptor fd, write the buffer to the file 
descriptor correctly, including error checking.

And no, I didn't count syntax errors against them. All I wanted was 2 
pieces of symantic understanding to be present. Only one person got both of 
them. Two people got one of the 2. The rest failed miserably.

(Oh, and if you want the answer, I suggest you read the late Stevens' UNIX 
Network Programming, and look at net_write().)

--
Carson

_______________________________________________
firewall-wizards mailing list
firewall-wizards () honor icsalabs com
http://honor.icsalabs.com/mailman/listinfo/firewall-wizards