| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9
Comments:
<0> It's not even funny or amusing.
<1> the gnaa 'gay niggers' video is pretty good though
<1> _gay niggers from outer space_ i should say
<2> speaking of gay niggers, i hear they made a Crackheads gone wild dvd
<0> _Aegis_: Ha, cute.
<3> yah thjey make us put up serious banners at work, I thought I'd poke fun at it
<4> what the hell is it about this channel and trolls
<0> _Aegis_: Of course, atheists or monotheists may not like the "god bless you". ;)
<3> trolling is an old sport on #c
<4> Teckla: monotheists should be fine with that
<3> yah. But I don't care if they like it or not. :)
<0> Oops, sorry. I meant, polytheists.
<3> lol, same difference
<4> not quite
<4> oh wait, this is C!
<3> yup
<4> anything that is nonzero is true
<3> you get a cookie
<4> OK you are correct
<0> Whee. My 802.11b xfer is back in the high 300k/sec range.
<1> monothesists for large values of 1 ...
<3> haha
<0> How free is VMware these days? Could I run Ubuntu 6.06 in a free version of VMware?
<2> yes
<0> Sweet.
<2> i use qemu thoo
<0> Those VMware guys, they're all right. :)
<2> much much faster.
<0> I ran Damn Small Linux in qemu for a while. It seemed a little slow.
<2> qemu kills vmware here
<0> Where is here?
<2> my 1.86ghz pentium mobile laptop with 1gb ram
<2> in montreal ;/
<0> Doesn't qemu emulate x86?
<2> it can do alot of archs
<0> I thought qemu strictly emulated all of x86 whereas VMware emulated only a few x86 instructions and ran the rest natively.
<5> teckla yeah
<2> Teckla, i havent really dove into it, but i benched it on my ubuntu 6.0.6
<2> err 6.06
<2> and qemu was way faster
<5> teckla i think vmware/vpc do some tricks like running the OS in ring 1 and running themselves in ring 0
<5> and they trap certain privileged instructions and translate them
<0> How could qemu be faster than vmware/vpc then?
<2> no clue, thats what the benches said ;]
<5> oh, i dont know. didnt see that
<2> i only tested 1 os tho
<2> plan9
<5> plan9 is awesome
<5> it may well be the future, just like unix was
<2> plan9 is special ;]
<3> wait until you see xcpu
<2> xcpu?
<3> it is gonna repolace the mosix kernel
<5> there is a midpoint between unix and plan9, it's microkernel unices like QNX
<5> QNX's QNET/FLEET is pretty nice
<2> qnx wasnt really for my use
<3> it hooks task management via the network using 9p2000
<0> I loved Qnx.
<5> you can look at another node's process list, etc.
<5> mount another nodes devices
<3> no special kernel needed for xcpu, just tke 9p2000 kernel module
<6> god, I can't wait until the Dragon Book I ordered gets here
<0> niv_: Yeah, Qnx even had that, way back in the mid-80s.
<2> finally, openbsd downloaded. brb making ISO
<5> or tunnel through a Sock[l]et server on another node as if youre proxying through it
<3> ls
<5> Microkernels and mountable namespaces will be the future (as in plan9 and qnx type stuff)
<0> Not according to Linus.
<5> i like plan9's dial() api
<4> plan9... from outer space?
<5> int fd = dial("tcp!kremvax!telnet");
<3> dial is as old as dirt
<6> hahaha
<0> Personally, I think microkernels are the way to go, I'm sick of how unreliable software is these days.
<4> I hope it runs on terrestrial hardware platforms.
<6> nice comphobby
<0> Anything to increase reliability is good even if it involves a performance hit imo.
<5> teckla: as soon as microkernels can get around the performance hits with context switching, yeah
<4> yeah, performance is getting pretty crazy these days anyways
<5> teckla: plan9 is a huge leap though, whereas qnx type microkernels are a compromise
<4> in certain cases go for full performance, but mostly I'd take a performance hit for more stability and security as well
<5> teckla: i'd prefer to see something like plan9 become the future more than i would qnx
<0> I don't know anything about plan9.
<2> plan9 is kinda ***y
<7> the thing that bugs me isn't that software today is slow, or that it's unreliable. it's that it's slow *AND* unreliable.
<0> What's special/different about plan9?
<6> Haha, I just looked up plan9 on wikipedia, and I love the bunny
<2> the bunny rules
<5> plan9 has a /net directory, dial() does that open("/net/tcp!kremvax!telnet") in the line i just typed
<0> evilgeek: ****ing A, you said it.
<5> that gets you an open connection to kremvax, port 23
<5> i like plan9's webfs too
<7> i just don't understand how you can **** up so badly that your code is both.
<6> So, you don't need big iron to run this, would you?
<5> basically you can browse the web using a filesystem
<5> i.e. cd www.google.com ; cat index.html
<4> evilgeek, I blame too many bad developers and not enough good ones
<0> evilgeek: There's a lot of really, really poor programmers these days.
<5> comphobby: after seeing the quality of code of 90% of my coworkers, i do too.
<8> niv_: Mounting another machine's /net in your own process can also be cool. :)
<7> Teckla: but, like, why do they get paid?
<5> hund: yep
<5> hund: that'd be like using another node's Sock[l]et server on qnx
<0> evilgeek: I...don't know.
<4> I'm going to have to find some of my project group member's code now.
<7> if they're not writing software that works, why pay them?
<5> hund: basically using their IP stack =)
<5> evilgeek: a few of us would love for them to be gone and get new people, but management doesn't see everything we do
<5> evilgeek: also, i'm not sure if anyone would apply that is good..
<8> niv_: I don't know how much work it would be for plan9 or QNX, but in Inferno you can make a socksify-like program in very few lines.
<4> niv_, Teckla: http://users.wpi.edu/~eastein/hownottocode.cpp
<6> niv_: hire me :D
<5> i can't hire anyone ;), i just started 7 months ago
<4> that's some code from a project team I was in... one of the other members (eastein) found that gem in a commit by one of the other guys
<5> i was amazed to see people twice my age that really ****
<4> the commit comment was about thirty lines of build errors. no explanation at all.
<4> he KNOWINGLY committed drastically broken code.
<4> that didn't do anything it was supposed to.
<5> comphobby: that's retarded
<4> yeah
<7> um, that's a lot of infinite loops.
<4> evilgeek, no kidding.
<5> comphobby: at my work we have people who check broken code into SCCS, but they are too stupid to know it's broken
<6> you should have comitted some serious workplace vengeance on him comphobby
<4> Humanoid, we did
<7> that shouldn't even compile.
<4> we all talked trash about him during meetings with the professor/manager
<7> no, you should just fire him.
<4> notice that it's basically just to initialize a chessboard...
<6> CompHobby: good move, anyone who would purposly commit crappy code is just an ***
<4> for some reason he decided he would use infinite loops for that?
<7> why is there a constant "Chess::BOARD_RANKS"? it's 8, yo.
<4> evilgeek: we wanted to be able to adjust it to be, say, 16x16
<7> but then the initialisation code wouldn't work.
<4> and it was a monster project for full credit in a C++ cl***, so we object oriented the living **** out of it
<4> evilgeek, that's because the idiot who wrote thte initialization code didn't know what the hell he was doing and should have stayed in the cl*** instead of joining the project ;)
<7> oh, is that one of those programming cl***es where you're not actually supposed to write code that works, but rather adhere to a bunch of random rules about how your code should look when viewed from a distance? what kind of curves it should have, and so forth?
<4> no, we were supposed to make it work fully
<4> and we did, despite that idiot holding us back and ****ing up the codebase left and right
<4> it was amazing, the day before it was due, we had most of the cl***es working, and then that night we finally actually wrote a test app using them
<7> but that flies in the face of the philosophy of object-orientation. OO code isn't supposed to work. it's supposed to exist so that professors who "research" software engineering can whack off at night.
<4> which just instantiated a game object and a connection object (it was an IRC chess bot), and then threw messages back and forth between them.
<4> and it worked THE FIRST TIME.
<4> we were shocked
<3> that is great. :)
<7> yeah, that's what happens when you write code that works.
<4> I mostly spent the whole term slamming on C++ and saying we could do it with 500 lines in C, but since it was for a C++ cl*** we had to use C++
<4> since we OOP'd the **** out of it we ended up with something like five thousand lines of code
Return to
#c
or
Go to some related
logs:
#computers
#delphi
#openbsd
#linux-noob
#gentoo
#politics
#windowsxp
#beginner
connie rice/lesbian
xedaps
|
|