logs archiveBotHelp.net / Freenode / #2f30 / 2015 / July / 17 / 5
atmc
http://marc.info/?l=openbsd-cvs&m=143707950008112&w=2
DID I PROMISE YOU?
k0ga
00:11 <@stateless> unnamed structures and unions <- It was an extension in kencc adopted by gcc
I think the problem with gcc extension is they added too much, and they added in incorrect ways
an example is 'typeof', that is using a namespace that should not be used, so correct C programs cannot use it if they want to be compiled with gcc
about other topic,
tm512
are you working on a compiler or something?
k0ga
yes
well, at this point it is a suckless project
kori
scc?
k0ga
yes
kori
I fu*king called it
I knew it was going to happen one day
this intrigues me.
k0ga
the problem of writing a compiler, is that if you want to be able to use it in general, then in some point you have to add the gcc extensions
because in other case you cannot use glibc, for example
kori
not a problem
k0ga
I don't know if musl or dietc have gcc extensions
kori
I meant that mostly as "just don't support glibc" even though it sounds ridiculous (and it is)
k0ga
but you need a libc, or to write new one
in fact, pcc has its own library
kori
if I HAD to program in C I'd program with musl in mind
k0ga
I'm going to take a look to musl
kori
but THEN AGAIN
I am not a C programmer
so I may be speaking out of my ass here
k0ga
#ifdef __GNUC__
__attribute__((const))
#endif
at least they protect the extensions
FRIGN
henriqueleng: biniar: retard: Holy sh*t, that was some deep discussion there. The feels.
:D
Evil_Bob
:)
stateless
morning
k0ga, yeah it was in kencc first
in any case, I was arguing about extensions as all compilers have them
k0ga
yes, it is true
for example DOS compilers had far and near pointers
stateless
k0ga, musl does not rely on gcc extensions at all
yes
k0ga
but I think gcc moved too far wit the extensions
s/wit/with/
stateless
in some ways yes
k0ga
stateless: the respect the namespaces. In the case they have to use gcc extensions they put ifdef guards, so it's ok
stateless
I am still annoyed at the clang guys for not supporting __builtin_va_arg_pack()
:P
k0ga
s/the/they/
stateless
k0ga, yup
musl can be used with pcc too
some people did it
and found bugs in pcc
haha
k0ga
xD
http://www.etalabs.net/compare_libcs.html
very good comparision work
stateless
k0ga, I will show you a piece of code I wrote that will never build with kcc
#define _FORTIFY_STR(s) #s
#define _FORTIFY_ORIG(p,fn) __typeof__(fn) __orig_##fn __asm__(_FORTIFY_STR(p) #fn)
#define _FORTIFY_FN(fn) _FORTIFY_ORIG(__USER_LABEL_PREFIX__,fn); \ extern __inline__ __attribute__((__always_inline__,__gnu_inline__,__artificial__))
tm512
well got into a C++ argument
stateless
lol
k0ga
xD
stateless
tm512, :P
Evil_Bob
morning stateless !
stateless
ISC dhcpv6 solved all my problems
fu*k wide dhcp6c
:P
Evil_Bob
what issue(s) did it have? :P
or just hard to use?
stateless
Evil_Bob, lease would expire after some time
but the client would not renew it until it thought it was time to do so
possibly several hours
tm512
maybe I should send them here so that you guys can properly educate them
stateless
and there was no option to set renew time
tm512
I'm tired and stuff
stateless
wide dhcp has been unmaintained for some time now, the release in ports is from 2008 so go figure
:P
Evil_Bob
aww yea that sucks :)
isnt it in base? or is that ISC/
*?
stateless
no it is not
base doesn't have an ipv6 capable dhcp client/server
only ipv4
the one in base is not isc, it is written from scratch
well the client side
dhcpd is ISC
Evil_Bob
aw hmm, hopefully a IPV6 will come into base, it should be imho
stateless
:)
we can write it
:P
just the client
not the server
Evil_Bob
lol, should be much harder than sdhcp though :P
stateless
yes :P
Evil_Bob
at home i just use sdhcp (static dhcp lease for desktop and laptop)
stateless
nice :)
I use dhclient in base
and dhcpd on the router
atmc
i should do everything in emacs
Evil_Bob
stateless: nice, does your router run OpenBSD or dragonfly or so?
« prev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 next »