[Clfs-dev] CBLFS - query about pages Liba52 / DJBFFT

db m myheadblewoff at hotmail.com
Sun Sep 14 05:30:22 PDT 2008




> From: wwh04660
> Date: Sat, 13 Sep 2008 20:56:33 -0500
> Subject: Re: [Clfs-dev] CBLFS - query about pages Liba52 / DJBFFT
> 
  << snip >>
> 
> I got a comment,
> 
>    Would you like to be our proof reader and super duper link checker?  
> I mean this in most sincere regard.
> 
> -William
> > 

Sure, I'm glad to help out where-ever I can.

For the record, I started doing the LFS thang many years ago,
can't recall exactly when, but I do remember being number 600
-and-something of people who bothered to get counted (;  I had
always hoped that (one day) I might find the time to give something
back to this community...guess what?....I found some. (and it only
took nearly a decade! ;)

I've just worked my way through CBLFS ...some 396* packages worth
anyhow...using this system instance, which is ostensibly a CLFS-1.1.0
makeup in X86_64_multilib guise. The new CLFS-1.1.0 release will
break some things in the CBLFS wiki. Some (ffmpeg at least) are
marked as broken already... (*some not currently in CBLFS)

ffmpeg -- IIRC, my work-around involved grabbing latest&greatest
from their svn/cvs repo...

 mythtv -- 0.21 won't build anymore, 'release-0-21-fixes' of mythtv
from their svn/cvs repo does...(instructions on wiki would need
changes for this, besides, there should be more here for mythtv
overall..ie; the myth{themes,plugins,backend})

librsvg -- 64bit build breaks


Links....yes, well, as I find them I usually post here lately. If I've
got it right, these are the wiki page parts which Joe has locked
due to people inadvertantly editting 'the wrong thing'.  Normally
speaking, I'd just fix these things, on the fly as it were, equally I
sometimes feel to fill in missing package 'homepage' links and
short description of the package itself, but and alas the wiki lives
in a realm where the word 'normal' doesn't always mean the same
thing for all people....<grin>....and the whole thing would, I feel
sure, quickly loose the plot if no locks existed at all. I'll keep posting (;

Proofing....CLFS-1.1.0 is at least 98% or more 'good' IMHO. The
CBLFS probably doesn't score that high...yet. As I mentioned in
a previous posting, I've been loath to change too much until
1.1.0 was actually 'out there' - this time is nigh. I've already made
a few minor edits to CBLFS in this regard... 

I imagine most folks try to get X up and running as soon as they
can after completing CLFS. I've kept the pristine 1.1.0-rc2 build
on another partition. As I mentioned in another posting, the final
'Xorg -configure' part failed to work for me here, I'm going to redo
all of this with that rc2 build and be a tad more pedantic. I notice
a few Xorg packages have been bumped a version lately on their 
repo...this would leave the strangeness with Xmd.h  I've encountered
with a number of CBLFS targets, Qt-3 (32bit build) being one.

In all cases, the 32bit builds were failing, because it couldn't get
this part right - [from Xmd.h];

#ifdef LONG64                                                                   
typedef unsigned long CARD64;                                                   
typedef unsigned int CARD32;                                                    
#else                                                                           
typedef unsigned long CARD32;                                                   
#endif 

The 32bit builds of some packages like Qt-3 seem to fail as a result
of concluding the #else statement when in fact the code was actually
expecting  'typedef unsigned int CARD32; ' instead. Believe me, 
what I know about c/c++ is rudimentary at best, but I know enough
to sometimes find part of the cause, without any idea of how to fix it
that's correct for the book. Given the build environment at the time,
I figure the conclusion is actually correct, just not what we want, and
this only seems to affect one or two subsections of the Qt-3 package
tree overall. Same goes for other packages so betangled...

As said though, I'll re-travel all of this over the coming week, and see 
if I'm to blame. Be assured that I will proof over the existing texts, 
and check the accuracy of download/website links therein. 

Regards,

Don  

_________________________________________________________________
Net yourself a bargain. Find great deals on eBay.
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Frover%2Eebay%2Ecom%2Frover%2F1%2F705%2D10129%2D5668%2D323%2F4%3Fid%3D10&_t=763807330&_r=hotmailTAGLINES&_m=EXT
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clfs.org/pipermail/clfs-dev-clfs.org/attachments/20080914/161e92f6/attachment-0001.htm>


More information about the Clfs-dev mailing list