I may have a way to make it possible to build a complete, recursive
dependencies list in one shot at the beginning of compilation without
having to prefetch every single recipe/package involved.
I'm experimenting on my home server with an alternative layout of the
recipe store. For each recipe, rather than just having the archive, I
have a folder with the same name as the archive. Within that folder,
I have the archive itself, and also the Resources folder extracted
from it.
A find on the directory for my Sudo recipe looks like this:
./Sudo--1.6.8p12-r2--recipe.tar.bz2
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Sudo--1.6.8p12-r2--recipe.tar.bz2
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Defaults
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Defaults/Settings
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Defaults/Settings/sudoers
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Environment
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Wrappers
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Wrappers/Sudo
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/Dependencies
./Sudo--1.6.8p12-r2--recipe.tar.bz2/Resources/BuildInformation
In the parent directory of all of this, I have an .htaccess file with
mod_rewrite set up to make requests for the folder itself give you
the archive, so scripts don't have to be aware of my directory
structure.
wget "http://localhost/~drmoose/Sudo--1.6.8p12-r2--recipe.tar.bz2";
#gets the archive
wget "http://localhost/~drmoose/Sudo--1.6.8p12-r2--recipe.tar.bz2/
Resources/Dependencies" #gets the dependencies
Building directories like this is extremely easy to do automatically,
and getting a dependencies list based on them should also be pretty
simple.
Just an idea.
_______________________________________________
gobolinux-devel mailing list
[email protected]
http://lists.gobolinux.org/mailman/listinfo/gobolinux-devel
