logs archiveBotHelp.net / Freenode / #2f30 / 2015 / August / 11 / 7
atmc
i drank kefir
i ate tuna and couscous
(not sure if thats russian) :P
stateless
k0ga, the openssl developers disagree :P
k0ga
stateless: and the bourne shell developer xD
stateless
gnu emacs also has its own malloc
lol
yeah him too
k0ga
I think in that moment was wery common
atmc
emacs should have its own kernel
k0ga
if you read very old code of unix then you see that was common
atmc
everyone and his mother redefines malloc somehow
k0ga
vi has also an own malloc
s/has/had/
atmc
http://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Sys/PetscMalloc.html
even my math library has its own malloc
stateless
k0ga, at the end of the day, the standard library gets in the way of you using system calls :P
k0ga, the reason they did it is probably because of emacs lisp, they probably have hundreds of thousands of allocations to keep track of the objects
be back later
k0ga
stateless: I don't understand you
retard
postfix also has its own malloc
atmc
i bet you could write an OS in postfix
stateless
k0ga, functional languages may require elaborage GC or generally many allocations to keep track of all the internal objects
k0ga
yeah
stateless
so that's probably why they did their own malloc
sth like pool allocator I would imagine
k0ga
this was the thing that was missed in my lisp implemantation, and it is the reason why I didn't finished it
Implementing a GC, even a simpler one, it is not easy
stateless
yeah
z3bra
http://i.imgur.com/PlAHPKb.png
sbase compiles fine against pcc+musl
(Makefile require patching though, to remove the -s flag)
Evil_Bob
noice, whats the binary size of a static sbase-box with musl+pcc? just curious
(stripped)
z3bra
hold on
FRIGN
z3bra: ah cool! :)
z3bra
text data bss dec hex filename
1353550 26276 95244 1475070 1681fe sbase-box
ouch
sorry about that
1.4M sbase-box
as per du -sh
odd.
haha
Evil_Bob
compiled with -Os and stripped?
z3bra
not -Os
Evil_Bob
that cant be right though :)
z3bra
this sounds weird indeed
make CC=x86_64-linux-musl-pcc CFLAGS=-Os LDFLAGS=-static sbase-box
kori
hi FRIGN
FRIGN
hey kori
stateless: http://simongrondin.name/?p=394 dude, just statically link it :P
z3bra
same result Evil_Bob :/
FRIGN
kori: what's up?
kori
nm you hl'd me earlier
z3bra
Evil_Bob, that's scary
Evil_Bob
z3bra: its 472272 bytes on OpenBSD 5.7
(static stripped)
z3bra
yeah
it's like 256K with standard gcc on my system
retard
"The good way forward is to package your app as a container and package all dependencies of your app inside the container."
jesus fu*k
fffuuuck me
z3bra
retard, that's static linking
retard
z3bra:
z3bra
but withing a directory instead of a folder :P
instead of a binary
retard
it is hardly equivalent
z3bra
I fu*ked up
retard
containers for this is fuuuucked up
containers in general aren't the genious idea we pretend it is
much like virtualization
z3bra
Evil_Bob, this is annoying.. I was planning on replacing gcc with pcc for my toy distro
gonna be hard withthose results :(
Evil_Bob
yeah, the difference is huge
z3bra
lcc perhaps?
if I can figure out how to compile it
haha, fu*k it
$ du -sh sbase\:0.0.pkg
25M sbase:0.0.pkg
same package with gcc:
644K sbase:0.0.pkg
Evil_Bob
waaaaaaat
z3bra
basically make; make DESTDIR=chroot install
then tar bz
this is...
I have no words
perhaps I fu*ked up when making the cross compiler
FRIGN
http://freshbsd.org/commit/openbsd/234baa39102d1fa8985230463712fcd21de7f317
lool
check this out
retard
quality
z3bra
Evil_Bob, well. It's pcc which is doomed
not my stupidity
1.6M sbase-box
with standard pcc against glibc
dami0
stupidity is never doomed
if its proliferation is anything to go on :P
z3bra
true, but this time, it's not my fault
dami0
oh, i was just making a joke
« prev 1 2 3 4 5 6 7 8 next »