It may be quite off topic and all, but i have a question for you daeken : How is it to code such a project in boo, considering an OS project is at the same time quite large , and quite low level in nature ?
I will look at the codebase someday, i'm very intrigued
It's fantastic. Downright fantastic. Even in portions that deal heavily with memory directly (e.g. our AMD PCnet driver), we get the ease of use that Boo and .NET give us. After working on this iteration of Renraku, I don't see myself ever using a non-managed language for OS work again.
Do you use Boo's macro facilities ? I also saw that Boo has got pointer manipulation facilities. Do you have to use that in the low level (-> drivers) parts of your code ?
EDIT : I'm following your project since the beginning. Many kudos on your work :)
We use Boo's macros now and we'll be using them even more soon. We don't use Boo's pointers, however. They didn't exist when Renraku began, and our pointer intrinsics are considerably more powerful. We use direct memory access in a few drivers and system initialization, but that's about it.
My only experience with Boo is Renraku and I was able to quickly figure out the existing codebase, despite its current lack of documentation and comments, and contribute shortly after. I have used Python a lot, so I felt immediately at home with Boo's syntax, so maybe that had something to do with it, but the choice of using a really high level language to develop an OS is a good one, despite the obvious challenges to be overcome.
Renraku has come a long way in a very short time (now with the new design decisions, I expect development to progress even faster still, once the base infrastructure is in place) and I attribute this to Daekens decision to use a really high level managed language.
I will look at the codebase someday, i'm very intrigued