LiveCD or No LiveCD?

[Jeremy Huntwork]

Alexander E. Patrakov wrote:
> Jeremy Huntwork wrote:
> 
>> The LiveCD exists as standing proof that the LFS book is 
>> sound and produces a working system.
> 
> Here I disagree. Because of numerous deviations and wagons of extras, it proves 
> nothing. Here is a counterexample: 
> http://www.linuxfromscratch.org/pipermail/lfs-support/2007-September/033586.html 
> . In this case, the network card worked only because the LiveCD included 
> pcmcia-utils. So the choice is, in fact, between deviations/extras and a 
> non-working CD for a non-negligible amount of users. I think this also has 
> something to do with the fundamental conflict between LFS not intending to be a 
> distro, and the LiveCD being a binary distro in some sense.

I think your definition of a 'working system' is different from what I 
intended. I did not mean that LFS or the LiveCD by extension should be 
able to run on any hardware imaginable. I meant that the binaries 
produced by LFS are stable, usable, technically sound and inter-operable 
as a system on x86 processors. What _drivers_ you use to access hardware 
is custom and not truly covered in LFS, therefore that aspect of the CD 
does not correlate to LFS.

So my argument stands. The base system is sound and, generally speaking, 
this is proven by the LiveCD. I will concede that the existing 
deviations of the CD from LFS does not make this argument 100% accurate, 
but again, most of those deviations do not, in my opinion, reflect on 
the whole of the system to a great degree.

--
JH