Le Mon, Feb 11, 2002, à 04:28:47PM +0100, Jose Carlos Garcia Sogo a écrit: > > If that happens (ie, if the system is so broken that the keyboard layout is > > US-ASCII), couldn't the user run APT with LANG=C for the purpose of bringing > > the system into a more usable state ? > > > > Yes, he can do it. The question is: does he know that he can do it? APT demands this long confirmation message, because the user is about to do bad things to essential packages (IOW, APT thinks that if the user doesn't know precisely what he's doing, he's going to make the system unusable). I don't think it's too much to ask from a sysadmin about to do extremely dangerous things to his system to know about working around lacking keymaps using LANG=C tricks. IOW: if it was required to type this long message in order to *recover* the system, from so-crippled-the-keyboard-is-US-ASCII state to "normal", in a foolproof and safe way, then it would have been Wrong to demand that the user is able to interact at least a little with his system in LANG=C mode (actually, it would have been Wrong to demand that the user be required to use LANG=C at all). However, we're talking about the exact opposite: we're asking the user a confirmation that he's a power user knowing exactly what he's going to do, and knowing he takes extreme risks with his system. Demanding knowledge of LANG=C to work around a potential mis-configuration, in that precise case, is probably not too much IMO. -- Cyrille -- Grumpf.
Attachment:
pgpg2FPSKCB7g.pgp
Description: PGP signature