User based package management (was: inetutils fails in compilation)
t.rice at ms.unimelb.edu.au
Tue Nov 3 13:46:48 PST 2009
I can see why you might say that, but being new to LFS myself, I actually
found the user-based package-management regime invaluable. This is because
it makes it easy to undo the mistakes that invariably happen, even when
trying to follow the book closely (which I have been).
In fact, my first couple of attempts to work through the book resulted in
complete messes that required starting all over again. These experiences
suggested that the costs of using UBPM would be more than outweighed by
the gains of having a safer (from me!) system.
As I have now finished building everything, and just need to figure out
how to configure things correctly for booting, this estimation appears
correct. I might have gotten where I am now quicker without UBPM, but now
I will be able to take more pride in my new system (once it boots!) and
have more assurance that I will be able to depend on it for the long haul,
eg when working on BLFS.
In fact, I think my attempts to get the boot configuration set up properly
would certainly have destroyed all my work already if I *didn't* use UBPM,
since I have had to recompile the kernel a few times, for instance to
compare the difference between using modules and not using them. Doing
such "playing around" as the root user - rather than, say, the
linux-22.214.171.124 user - seems like something of a risky proposition,
suitable maybe for LFS pros but certainly *not* us noobs.
> Trying to use package users (or really any package
> management system) is generally not a good idea unless you are very
> familiar and experienced building LFS. Package Users in particular is
> tough for anyone relatively new.
More information about the lfs-support