System Tree essentials UNDERSTANDING the system

Sandro Noel. snoel at gestosoft.com
Thu Sep 15 05:49:35 PDT 2005


>That is the usual approach but again, not easy. You need to track
>modules and versions so that you can know if things will fit together
>at the end. I would hold this part to the very end, after the
>architecture of the modules is well understood.

>Good, but let's keep that for later. The approaches taken in nano and
>pico bsd of using the build system itself are a good start.

Yes that is the idee, from source, just like the port tree.

>Doxygen can help us here, but it will take a bit of horsepower for the
>whole kernel. Take a look at the doxygen docs and you'll see that it
>will generate this. We just have to be careful not to do it from root
>and then to annotate the code to fix errors in the automatic
>generation phase.

i'm looking into it... but i must admit it's a mistery to me right now.

>So, that's fine. I think the documentation/cutting up step has to
>come first, which is often frustrating to people who like to mostly
>code, but I just don't see a way around making the map.

I dont have a problem with that, I just need a good starting point.
and something to look for.

>How much time do you think you have to put into this?
A couple of hours a day, mostly after 7:00pm EST

You seem to have more knowlage than I do here, would you like to 
take lead in this project?

Something i want to calrify, it is not just about the kernel and it's drivers"modules"
tought it is the starting point of the system, i also want to map the system features 
like networking ppp, bluetooth, mostly everything that ships with the system sources 
that is not already into a port tree or a package.








More information about the freebsd-small mailing list