Actually, I did. I guess I missed that one (or seen it and didn't pay much attention to it).
Deducing anything about the kernel from the history is pretty hard. Actually, I haven't found any source of 'design explanation' about the kernel - I read a description about the linux kernel design a while back (from the kernel's developer guide), but it wasn't updated back then, and it wasn't as detailed as I had wished.
Didn't anyone keep some kind of organised document from the developer discussion or something?
I found it really hard to believe that the linux kernel, being so big and complex as it is, has actually no guide to it's code. No official guide, anyway. And the code comments don't help
(and they're not supposed to. Linus is right to say in/documentation/Coding that comments are not the place to explain design. But he doesn't say where design IS. Or did he say and I missed it?)
Anyway, I'm glad to hear it:)
I've been meaning to read up on the VM change. Went looking for documentation about the new VM's design. Naturally the Changelog didn't say where that could be found, and other kernel guides were not updated as to the new VM. I'll keep looking later today when I have time. I guess it's under the new kernel's documentation dir (haven't downloaded it yet)
I didn't mean copy and recompile the code, but build a driver based on the linux one. The basic problem with driver building, in my very very humble opinion (never built a driver myself, just read about driver writing) is testing the hardware (trial and error) and knowing what to write and what errors to overcome. This would be the same for FreeBSD and linux, so most of the work is done.
Since the code is moslty re-written and not copied, I don't think its a based work (which means you'd be forced to use the GPL license - impossible) but a different work altogether - not limited by license.
I'm just a newbie at kernel hacking, but I've always wanted to ask this:
If FreeBSD really has a much more stable TCP/IP stack and other 'cool' things, why can't linux build something with the same architecture (that's not even code theft, just idea borrowing).
Before anyone flames me on this: I just heard from some experienced hackers, on both the FreeBSD and the linux side, that some parts of the FreeBSD kernel are designed better.
It's all open source. Why can't we all just get along?
(And the same thing works the other way: FreeBSD users claim the lack of comptability with every piece of hardware is a small price to pay to use FreeBSD. Why can't linux drivers be ported to FreeBSD?)
I was actually thinking of trying to move the TCP/IP stack from FreeBSD to Linux myself, or at least get a close enough look at the differences between them, but looking at the code I realised I'm lacking the skills or mind stamina to comprehend linux's kernel code. Maybe some day...
Slashdot Mirror
Actually, I did. I guess I missed that one (or seen it and didn't pay much attention to it). Deducing anything about the kernel from the history is pretty hard. Actually, I haven't found any source of 'design explanation' about the kernel - I read a description about the linux kernel design a while back (from the kernel's developer guide), but it wasn't updated back then, and it wasn't as detailed as I had wished. Didn't anyone keep some kind of organised document from the developer discussion or something? I found it really hard to believe that the linux kernel, being so big and complex as it is, has actually no guide to it's code. No official guide, anyway. And the code comments don't help (and they're not supposed to. Linus is right to say in /documentation/Coding that comments are not the place to explain design. But he doesn't say where design IS. Or did he say and I missed it?)
Anyway, I'm glad to hear it :)
I've been meaning to read up on the VM change. Went looking for documentation about the new VM's design. Naturally the Changelog didn't say where that could be found, and other kernel guides were not updated as to the new VM. I'll keep looking later today when I have time. I guess it's under the new kernel's documentation dir (haven't downloaded it yet)
I didn't mean copy and recompile the code, but build a driver based on the linux one. The basic problem with driver building, in my very very humble opinion (never built a driver myself, just read about driver writing) is testing the hardware (trial and error) and knowing what to write and what errors to overcome. This would be the same for FreeBSD and linux, so most of the work is done. Since the code is moslty re-written and not copied, I don't think its a based work (which means you'd be forced to use the GPL license - impossible) but a different work altogether - not limited by license.
I'm just a newbie at kernel hacking, but I've always wanted to ask this:
If FreeBSD really has a much more stable TCP/IP stack and other 'cool' things, why can't linux build something with the same architecture (that's not even code theft, just idea borrowing).
Before anyone flames me on this: I just heard from some experienced hackers, on both the FreeBSD and the linux side, that some parts of the FreeBSD kernel are designed better.
It's all open source. Why can't we all just get along?
(And the same thing works the other way: FreeBSD users claim the lack of comptability with every piece of hardware is a small price to pay to use FreeBSD. Why can't linux drivers be ported to FreeBSD?)
I was actually thinking of trying to move the TCP/IP stack from FreeBSD to Linux myself, or at least get a close enough look at the differences between them, but looking at the code I realised I'm lacking the skills or mind stamina to comprehend linux's kernel code. Maybe some day...