[RFC] Moving the BLFS and LFS profiles to using XIncludes

Thomas Pegg thomasp at linuxfromscratch.org
Sun May 23 15:01:30 PDT 2004

On Sun, 2004-05-23 at 11:58, Kevin P. Fleming wrote:
> Thomas Pegg wrote:
> > How do using XIncludes make for more complex parsing, I've never noticed
> > anything when I used them in my profiles.
> The entities (common ones) are being parsed repeatedly, and the extra 
> {alfs} elements cause the node tree in nALFS to be larger. These are not 
> large concerns, so I wouldn't use them as a reason not to move forward, 
>   I just wanted to make sure they got mentioned.
> > There could be a potential problem with that approach, say if some
> > things get rearranged by way the order there in the book, then all the
> > files have to cvs removed, renamed and then added again to suit the new
> > order.


> You are right; while we are stuck with CVS that is a big problem since 
> it doesn't handle renames. If this idea is to be pursued then the 
> information would have to be embedded in an XML comment inside the 
> profile file itself, which would complicate the top-level builder 
> somewhat (it would end up being a Perl script, probably).
> I still think this is a useful method, but not worth a lot of 
> implementation time. If you want to move the profiles over to using the 
> XInclude method go ahead... when I start working on nALFS2 it's all 
> going to change anyway :-)
If I do I will probably do it on a branch, so all the kinks can get
worked out before it's merged into a production profile. Although I
would like to hear at least a few more people comment on this before I
move forward with it.

LFS User : 4729 
Linux User : 298329 
warpcore - 17:00:02 up 2 days, 18:10, 1 user, load average: 0.09, 0.09,
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.linuxfromscratch.org/pipermail/alfs-discuss/attachments/20040523/9a3043c6/attachment.sig>

More information about the alfs-discuss mailing list