Randy McMurchy wrote:
However, I thought our goal was to provide instructions to build a safe, secure, accurate Linux system. Why should readers not trust the book and do extra steps to ensure what *the Editors* say is the steps to accomplish this.
Except that as you pointed out earlier, a sysadmin that is worth anything won't install a package without first inspecting it. If that's the type of trust that a good sysadmin should have for the developers of a program, how can we expect the readers of the book to trust us any *more* than they would the developers? It's all in the same category.
Also, you mentioned a goal of building a safe, secure and accurate Linux system. You forgot the aspect of education, which to me comes even before that. (Without education LFS becomes rather unwieldly. From a commercial standpoint a distro is almost always a better solution - you have a complete system that is guranteed to work in a specific way; tried and tested and tried and tested, and support to ensure it is working is provided as well. But that's all another topic...) And as for the sake of accuracy, which you did mention, the idea of a fake root and the ability to better audit what a package is installing can only improve the accuracy of LFS in general.
-- JH -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
