[this mail is CC to three interested people. If any of you get this mail by other channels, for example debian-policy, and don't wish to get CC'es, let us know.]
Hello, Santiago raised the issue of the names of the environment variables. I am open to discussion about this, because the perfect names are probably not proposed yet. In particular the GNU_* names are probably not chosen very well. The current proposal has: SYSTEM: for the whole string. ARCH: for the cpu KERNEL: for the os/kernel (former OS). The INSTALL file of a GNU software package says: TYPE: for the whole string. CPU: for the cpu. SYSTEM: for the kernel/os. (It also has VENDOR for the middle string, which is not used in Debian). Please let me know what you like more. I chose the former set by adopting commonly used nomenclature among Debian developers, which was probably not a good idea after all. The second set is more in sync with the actual use of the variables. Actually, now that someone draw my attention to it, I like it more to be consistent with GNU nomenclature, especially because the strings are used mainly for GNU configure. It has more advantages: the cpu is really labelled CPU, and it does prevent confusing the Debian ARCHitecture with the GNU system TYPE. If nobody has any (strong) opinions about it, I would like to propose to change my own proposal to use the second set. Geek, now you must notice that this is the first time I go through this policy proposal procedure :) Waiting is, Marcus -- "Rhubarb is no Egyptian god." Debian GNU/Linux finger brinkmd@ Marcus Brinkmann http://www.debian.org master.debian.org [EMAIL PROTECTED] for public PGP Key http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ PGP Key ID 36E7CD09
