h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

Installing on Linux does take practice... with any luck it will become less frustrating after a while, and simply be annoyingly time-consuming

On Gentoo, most things are a breeze... yes you have to wait for them to compile, but you don't have to sit there watching it all the time.

I was wondering what people's opinions were on the ? operator.

As a side-effect of some recent unexpected work for an interactive testing company, it transpires that ANSI C regards the value of the ? operator as a definite lvalue, whereas C++ (ANSI C++, I think) considers the result to be an lvalue if the two colon-seperated parts have similar type. Apparently, the GNU C compiler follows the C++ model by default, allowing

( a ? a : b ) = c ;
as an equivalent to
if (a) a=c; else b=c;

while ANSI C would barf on the same code.


On an unrelated and vaguely ironically humorous issue, I recently got a book from the local library which was an introduction to network and information security. In the section about the different ways that unauthorised people can access information (social engineering, dumpster diving, etc), I found a piece of paper with the notes which a previous reader had made, and which they had forgotten to remove from the book before returning it to the library. No passwords, unfortunately.

Hmm, that makes sense... C++ has a much stricter typing system than C, so that kind of thing is compileable, as the compiler can check if it's okay at compile time, rather than allowing it to crash spectacularly at runtime.

Whoops - just spotted a typo in my post. It should have been that ANSI C considers the value of a whole ? expression as a definite *rvalue*.

I can see some vague utility in allowing
(a ? b : c) = d ;
for reasons of compactness of source code, especially where d is a large expression, but it wouldn't be hard to make an optimising compiler realise that
if (a)
b = (some long and complicated expression) ;
else
c = (same long and complicated expression) ;
could be optimised to be the same as the equivalent conditional expression.

Reading around yesterday, I found that where lvalue-returning conditional expressions are allowed in C/C++ compilers, there is some variation about the meaning of
a ? b : c = d ;
some treat it as
a ? b : (c = d) ;
and some as
(a ? b : c) = d ;

at least the less elegant:
*(a ? &b : &c) = d ;
makes it clearer that something odd is happening, and works consistently across all C and C++ compilers.

This post has been removed.

Hmm

Hidden posting? I hope I didn't start some kind of Holy War here...

Love the sound affects!!!!

no there was no hidden posting. my "friend" aaron waited for me to log on and then overpowered me out of the chair, said some bad stuff, and then REPORTED it. I hope they don't kick me off
>8^B

oh and as a response to about a page ago, i am using RH 8
didn't have time to read anything, i was too busy freaking out over the dirty things aaron wrote. i do hope the moderators cover up that stuff at my space
>8^B

i never use the ? operator or the switch() function. i am not familiar with the syntax and more comfortable with an ugly line of for()s and if()s
what sound effects?
>8^B

Technically, switch isn't a function... it's a statement.

see? i told u i dun understand switch()
>8^B

today is no fun. i just found out i didn't get in to the oklahoma school for science and math
>8^B

I finished my degree

good for you, maybe i can go to college some day. i was really relying on getting into OSSM for high school, though. the one i am in the district for isn't very prestigious if u know what i mean
>8^B

i decided that i want to make a 4D RPG.
anyone wanna make me a 4D map?
>8^B