Just a thought that popped into my head

Geert Poels Geert.Poels at skynet.be
Tue May 7 10:32:47 PDT 2002


>Umm....dude....
>
>"LFS -- Your Distro, Your Rules, YOUR PROBLEM!"
>
>Don't you get it?  YOU decide what you want!  YOU decide whether or not to
>use gcc3.x on YOUR system.  So what if the book goes to gcc3? I'm not going
>over to gcc3, or gawk, and I'm even gonna go to using uClibc!  That's MY
>decision! NOT Gerard's!  And if you can't comprehend something THAT simple,
>you should switch back to Mandrake/RedHat/Debian/whatever and get off the
>list.
Thy are not addressing Lord Sauron. ;)

That's an too easy remark.
The compiler is close to the quintessence of the creation of the distro.

For those interested :
Sun's Java SDK features an XML-parsing library (Xerces/Xalan) which got
a major upgrade some time before the 1.4 release (but
in time to make it into 1.4.)
These library features a way to query XML documents (XPath).
Several developers, including myself noticed really HUGE slowdowns
between the old and new versions.
Up till today, it still hasn't been fixed and because of the rewrite, it's
very unlikely to be.

What if gcc has a similar issue(s).  This might take a really long while but
we may only hope for the best.
I agree on an earlier remark about LFS not intended to be a production system
but it should at least compile. And not only a basic LFS, at least BLFS and
a lot more other software.
If it's only an Athlon issue but a patch can fix it, no reasons to not go 
with GCC3.

Why doesn't a list gets compiled of contra's (and pro's) and we'll see how 
badly
the LFS-creation malfuntions because of it ?

(In the end contacting somebody from the gcc team to double-check ?)

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list