[RFC] Package Rationale

Matthew Burgess matthew at linuxfromscratch.org
Tue Jul 6 12:23:43 PDT 2004


On Tue, 6 Jul 2004 13:17:09 -0600
Archaic <archaic at linuxfromscratch.org> wrote:

> On Tue, Jul 06, 2004 at 06:34:23PM +0100, Matthew Burgess wrote:
> > 
> > As do I.  So, if anyone has any problems with the content of those 2
> > pages then speak up.  I'll add that content to the book within the
> > next few days if there are no major concerns.
> 
> The only thing on it that still seems up in the air is readline.
> 
> 1) The BLFS build of it is not broken,
> 2) bash doesn't need it,
> 3) very few need history in debugfs
> 4) most will install a *good* ftp client in BLFS.
> 
> And for those who want to argue (especially the last 2 points), then
> I'll restate, again, that a link to readline is quite apropos, just
> like for filesystems and editors. The links serves all points of
> everyone's argument, so why fight against it so vehemently?

Agreed (plus debugfs actually does a dlopen() call to it at runtime, so
it's not a compile-time dependency.  Let this then be the end of it, at
least for 6.0 anyway!  A link to BLFS will be put into the book for
those packages that can utilise readline.  I still don't agree that the
readline patches aren't required/useful for BLFS but I'll leave that up
to the blfs-dev crowd to argue over.  If they're not useful and/or
required then why have we got similar patches to bash's in-tree readline
functionality?  Do we need to rethink these?

Cheers,

Matt.



More information about the lfs-dev mailing list