LFS Future Braindump (Reflecting)

Robert Daniels rdaniels at linuxfromscratch.org
Tue May 27 12:35:26 PDT 2008

I'm just going to try to summarize some of the feedback from the thread 
so far, with a few more thoughts from me.

First off, the request for more information in the package descriptions.  
The responses seem to be generally in favor, qualified with the 
question of "How much is too much?"  Christian appears to have much the 
same opinion on this as I do, although he communicates it better by 
giving an example of a good description.  Keep the description short, 
but long enough to give a general idea what the package is for.  Randy 
emphasizes a need for users to do outside research, as we cannot do 
everything for them.  I agree with this, adding only that the 
description should make the package interesting enough to do this 
research.  Gerard and Randy (maybe others) suggest linking to outside 
sources of documentation, which I think is a good idea.

Regarding use-case scenarios, I think the general concensus is that this 
type of information is good wiki material, but shouldn't go into the 
main book for reasons of maintainability, clarity, and flexibility.

The only direct response to rearranging the book to cover reference 
implementations for various computing purposes was from Randy.   He is 
against the idea, both because the software to include in 
these "prototypes" would be too arbitrary, and because we would not 
have the manpower to support them.  It would result in a loss of 
flexibility for the book.  I would also have to add that there are 
almost as many end uses for computers as there are computers, and it 
wouldn't do much good to only mention a few.  Somehow, I only just 
realized that.  Due to these problems, I have to reverse my opinion on 
this.  The original idea came from the course modules idea on the LFS 
list, for the LFS book.  I still think that idea could be a good fit 
there, but I realize now it would likely not be good for BLFS.

There is also the perennial issue that we lack manpower.  It is a 
failing of mine that I always forget that we don't have the resources 
that larger projects have access to.

More information about the blfs-dev mailing list