An Authors View on C

[Joona I Palaste]

If it doesn't *guarantee* a memory leak, how can you prove that the code
in question generates one?

Are you unable to tell the difference between mandated behaviour and
allowed behaviour?

BTW, any language lawyer could trivially point out that even matching
any malloc call by a corresponding free call doesn't guarantee that
no memory leaks occur. The standard simply does not address the issue in
an unequivocal manner.

Here is the part in my post that you snipped:
"(I had a similar discussion with Dan Pop once. I don't think I won that
argument. I'm not going to try to win this one. Just pointing this out.)"

Keeping that in mind, I won't argue further. Those interested in the
arguments in the previous debate can look it up on Google. AFAICR it
involved the fclose() function.

 

[Mike Silva]

Take a look at

http://www.ganssle.com/bio.htm

Evidently, Jack Ganssle makes his living by lecturing and writing
articles and books on embedded computing.
He doesn't appear to have any other credentials
in embedded computing or in C programming.

You mean other than designing, building, programming and consulting on
embedded systems since the days of the 8008? You mean other than
(from the bio link you posted):

"He started, developed and sold three electronics companies; including
one of the world's leading producers of embedded development tools.

He developed or managed the development of over 100 embedded products,
including in-circuit emulators, underwater navigation equipment (some
used to recover Challenger's wreckage), steel thickness gauges,
near-IR protein measurement instruments, the White House security
system, compilers, colorimeters, numerous classified Government
systems... and even one device that analyzed cow poop!"