Investigating dependency trees
I'm working on a very customized minimal install.
A selection criteria for top level applications will be how many
packages will it pull in that are not used by other packages.
A further optimization may be which packages automatically
installed by Debian Installer are not needed by actual target
audience.
Synaptic and dselect provide this information for *ONE* level
down from a particular package *IF* an actual installation and
repository exist.
All the information exists in "control" of "*.deb". For my
purposes there is a lot of irrelevant information also. What I
envision is a relational database containing the package name,
Version, Depends, Suggests(doubtful), and Recommends(doubtful)
fields.
How the data would be displayed in interactive mode - haven't
decided yet.
In a "batch" mode I envision entering a list of "top level
packages" and getting out a list of lower level packages required
and how many (and which of) the "top level packages" required each.
Any suggestions of applications with similar goals?
TIA
Reply to: