This one time, at band camp, Roberto C. Sanchez said: > On Sat, Oct 07, 2006 at 12:39:39PM -0500, John Goerzen wrote: > > On Sat, Oct 07, 2006 at 01:34:35PM -0400, Roberto C. Sanchez wrote: > > > I'm thinking of this slightly obscure service called email. It might > > > possibly be affected by such brain damage. Why can't the bug submitter > > > make to with +, ., and other such characters? > > > > According to my very brief examination of RFC2822, a quoted-string is > > permitted as the local prt of an address, so it should be possible to > > address email to, for example, "username@withatsign"@example.com. > > > Then, I guess the relevant question has to do with whether or not > adduser (and the rest of the components that touch or use the username) > are RFC2822 compliant. BTW, I was not aware of RFC2822, but I would > think that since the @ is used as the standard username and hostname > seperator, you would not want to use it as part of the user name. The real thing to my mind, is that there are two questions: what breaks when you do this sill username scheme? should adduser allow it or always bail out with an error? I suspect most tools do not support usernames of this format. I suspect quite a lot of things will go wrong if you try to implement this sort of thing on your system. I also suspect it will work reasonably well if they are pure virtual users that only exist to receive and send email, or as part of a kerberos domain or similar. I'm not sure any of the suspicions have any bearing on question #2, however: should adduser be allowed to add usernames of this format, when told to with suitable --force flags? Thanks, -- ----------------------------------------------------------------- | ,''`. Stephen Gran | | : :' : sgran@debian.org | | `. `' Debian user, admin, and developer | | `- http://www.debian.org | -----------------------------------------------------------------
Attachment:
signature.asc
Description: Digital signature