jhalfs development

M.Canales.es manuel at linuxfromscratch.org
Mon Mar 6 13:20:15 PST 2006


El Lunes, 6 de Marzo de 2006 21:41, George Boudreau escribió:
> Manuel:
>    I have not spent much time on jhalfs as I have been waiting for a
> concensus on a new format. 

If you meant about a new XML parser format, look like now the work in on the 
C++ version.

IMHO, having the jhalfs sources modularized we could to changein the future  
the parser functions files to use that new C++-based parser (or the 
bash-based one, if George M. is working yet on it) instead of current 
xsltproc+XSL, if the generated output is suitable for jhalfs.

> However, the attached file contain some 
> changes I made to my earlier work on a unified system (an ugly beast but
> with potential). It will produce (h,c,l)fs bootable partitions with a
> little hand holding. Note:: No attempt was made to integrate blfs into
> the system.

I agree with the general approach and code structure. What do you think about 
my proposed layout?.

>    The clfs has only been tested on x86,multi,pure. The remainder of the
> systems may work but I have not attempted to verify the output. Any
> review I could give would be like a eunuch giving a review of the Karma
> Sutra: it could only be from a theoretical point of view.

Well, we can't test all archs. That is for the potential users ;-)

>     There are many details that need to be ironed out in the config,
> package and patch entity files (a common format would make life easier)

Yes, in HLFS patches.ent is missing, and in CLFS all packages for all archs 
are mixed in general.ent and patches.ent. I think that we should to try to 
extract the list of actual used packages and patches from 
{packages,patches}.xml instead.


-- 
Manuel Canales Esparcia
Usuario de LFS nº2886:       http://www.linuxfromscratch.org
LFS en castellano: http://www.escomposlinux.org/lfs-es http://www.lfs-es.com
TLDP-ES:                           http://es.tldp.org



More information about the alfs-discuss mailing list