On Mon, Nov 24, 2008 at 10:53:00AM +0500, Alexander E. Patrakov wrote: > > Make an empty branch, let editors copy pages (and, where possible, > upgrade versions) there that contain instructions verified against > LFS-6.4rc1 or later and all dependent libraries with versions in this > branch. The text should also be checked, and dubious places commented > out and converted to Trac tickets with a note "no response in 1 month = > permanent loss of this text". This way, we'll lose packages that no > editor cares about (users don't matter if they don't report bugs!), and > get a slim high-quality book. >
I can see you think "less is more". I'm reluctant to say that we should drop every package unless an editor has both the inclination and the time to test and fix it in a month or so. It also doesn't solve all of my issues - I'm happy with xorg-7.4, but I'm building a lot less than some people, and using a different form of automation. I'd prefer to leave xorg to people who follow the book's method. I could throw in the toolkits (gtk+ and so forth), but they won't be a lot of use without xorg. And until the branch has enough to be useful, we can't expect users to test it. Also, branching seems unnecessarily complex and I don't see what it gains (apart from potentially identifying "nobody cares and has time to fix this" packages). I think that creating an almost empty branch (chapters with very little in them - but with enough to get it to render) will be a lot of work, and adding package details back in will take more work than just editing what is in trunk. ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
