From the Computer Errors challenge. See all 409 entries (closed)

( Puromycin will be scraping his "special" underpants on, Sun 24 Jun 2007, 23:46, archived)

haha too true
too true
(Spongeblunt Indeed sire., Sun 24 Jun 2007, 23:48, archived)

arf!
someone who occasionally calls me for advice when stuff crashes insists on reading out all those sodding numbers assuming they'll mean something to me
( eclectech twitter.com/eclectech, Sun 24 Jun 2007, 23:49, archived)

Memory dump
Sitting on the toilet trying to remember what the fuck those numbers mean.
/20+years of programming blog
( Puromycin will be scraping his "special" underpants on, Sun 24 Jun 2007, 23:54, archived)

I always thought that it
was the most pointless debugging tool - I mean, how do you know what address the program started at?
( Damocles Ceci n'est pas une signature, Sun 24 Jun 2007, 23:58, archived)

I usually ask
the program's parents
( Puromycin will be scraping his "special" underpants on, Mon 25 Jun 2007, 0:02, archived)

my program killed it's
parents when it set fire to the family home.
( Damocles Ceci n'est pas une signature, Mon 25 Jun 2007, 0:05, archived)

Serious answer?
It's actually very useful. Programs usually load at the same address, and you can then use either the map file or load the program into memory with a debugger (on the same OS version) to find out where the address is.

In that case the error can be narrowed down considerably, although a process dump is much more useful than just an address.

The problem is moot, though; various Unixes/Vista support Address Space Layout Randomization (ASLR) to make stack overflow/code injection attacks etc harder. This changes the base load address of the process, and thus probably renders the memory address of a crash useless.
(EssY, Tue 26 Jun 2007, 11:50, archived)

very true
*clicks*
(imnotplayinganymore Like a rock. Like a planet., Sun 24 Jun 2007, 23:54, archived)

i go for
Hug!
(chippy time travel agents, Mon 25 Jun 2007, 0:19, archived)