You are writing a kernel from scratch. And now you have the opportunity to forge a brand new sandard library for your operating system, since you hated for years all that nasty programming stuff out there.
You may use C for your library, but you'd like to use an object oriented language. C++? Cool story, if you want to waste time being overwhelmed by compiling errors and getting a huge shit.
>>8
I'm sick of all the "GC considered harmful" bullshit going around. Yes, it's easy to implement a retarded GC that murders your pr0n starperformance with a spork. It's also easy, if you actually know what you're doing, to implement a real time GC that manages memory at predictable intervals.
Name:
Anonymous2009-04-29 15:17
>>9
...FOR THE KERNEL?? Kernels don't use that much memory, a GC would be overkill. The kernel is so time-critical, anyways, that I don't care how good your GC is, it's gonna slow it down.
Name:
Anonymous2009-04-29 15:39
>>10
You could say EXACTLY THE SAME THING about function calls
I want a kernel that does nothing. It only needs to exist. Theoretically, my performance should be much superior to the petty alternatives with their system calls and modules.
Name:
Anonymous2009-04-29 16:18
For you programmers who can't read, >>1 is talking about what the standard library would be written in, not the kernel.
Note that the C standard library is included in "that nasty programming stuff out there."
Name:
Anonymous2009-04-29 16:30
>>13
No one will care as long as it uses the C standard calling convention. And it has to, or every programmer that wants to do something with your system will have to write their own wrapper to your toy language.
Name:
Anonymous2009-04-29 16:32
it's sad that there isn't much work being done with languages that build to robust, reliable machine code. besides C and C++, you've got Pascal sorta waddling along, D trying to act relevant, and perhaps LLVM figuring out a cross-platform solution. meanwhile you've got an assload of toy languages popping up every few minutes.
anytime someone pipes in that there should be something better than C/C++, all the lobsters piss all over the independent thinker with comments such as "nothing really provides a worthwhile solution to C/C++".
>>15 anytime someone pipes in that there should be something better than C/C++, all the lobsters piss all over the independent thinker with comments such as "nothing really provides a worthwhile solution to C/C++".
Yeah, don't remind me. I once said something on a C IRC channel about how there should be a "C with typesafe macros" and this dude went ballistic on me and spent like half an hour spamming me with messages like "STOP BEING A WHINY BITCH JUST BECAUSE THE PREPROCESSOR DOESN'T IMPLEMENT YOUR POINTLESS TOY FEATURES! YOU DON'T UNDERSTAND THAT IT'S A PREPROCESSOR! PRE!! PRE!!!! PREPROCESSOR!!!!!!"