Luca2 liliana.perossa at
Fri Apr 20 23:33:43 PDT 2007

Hi Matthew and Dan!

The new "HASH" code should have been already pushed in current svn FSF 

Anyway I took it as an example; months ago I used other file-systems 
requiring other bare system packages as example but there's XEN too and 
these are bare system features.

My opinion is that steps towards future are needed or at least LFS 
should give the user the chance to decide.

If we're talking about standards current FHS has invalid standards like 
X11R6 because the new standard should be X11R7 (release 6 is outdated 
and there's out release 7) which I began to use more than one year ago: 
an example: I install Xorg 7.* (now 7.2) in /usr/X11R7 hierarchy.

I mean that being conservative is not always the best thing and does not 
cover the current times anyway, as already said, I think that at least 
the LFS users have the right to decide to use, examples, new HASH or a 
XEN-kernel etc...

If we're taking pure defaults there is no innovation.


----- Original Message ----- 
From: "Matthew Burgess" <matthew at>
To: "LFS Developers Mailinglist" <lfs-dev at>
Sent: Friday, April 20, 2007 9:29 PM
Subject: Re: Thoughts

> On Friday 20 April 2007 17:26, Dan Nicholson wrote:
>> I think --hash-style=gnu is a great innovation and plan to build some
>> systems that way. But LFS has never done anything more than just take
>> the defaults that the toolchain gives it. Why not 
>> enforce --as-needed,
>> too? Or default CFLAGS to -march=whatever?
>> I'd be open to the idea, but I don't know how likely it is to become
>> the default in the book unless it comes that way out of the box.
> LFS has always assumed, pretty safely I might add, that upstream 
> developers
> know what optimizations or package-specific options are safe/suitable 
> for the
> majority of users.  This is particularly the case with the toolchain. 
> I
> personally wouldn't be comfortable with this going into LFS until 
> upstream
> think that it's the right thing to have on by default.  At the moment, 
> it
> doesn't even appear to be in binutils-2.17 (not in --help output or 
> ld(1) or
> in the output of a recursive grep).  Maybe it's just in HJL's 
> binutils?
> Regards,
> Matt.
> -- 
> FAQ:
> Unsubscribe: See the above information page
> -- 
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.5.463 / Virus Database: 269.5.6/770 - Release Date: 
> 20/04/2007 18.43

More information about the lfs-dev mailing list