Richard Lowe wrote: >> There's another issue though - there is no requirement that OSO and SWAN >> usernames match up, and there's nothing to prevent someone outside >> registering with a username that is the same as an existing SWAN >> one. > > I'm not sure why that matters.
Because people will tend to assume there is always a match, because in many cases there will be. >> This is doubtless going to cause confusion. Personally I'd prefer that >> we made it mandatory that OSO and SWAN userids match up, even if that >> means renaming some existing OSO ones. > > I think that's bad for many reasons, not least that those cursed with > the borg usernames are then stuck with them and that, in theory, your > IT policy is such that said usernames should never be visible outside. > As I understand it (second hand) your sentence above mandates every > employee violate that, though perhaps I'm wrong. Borg usernames can (and should) be changed in any case. The problem we have is that we have no reliable way of identifying Sun employees in the current OSO database. There is a employee_id column, but there's nothing to check it is filled in correctly, or even at all. Email addresses are no use either. And if we don't mandate that SWAN and OSO usernames are the same, the problem gets even harder still. And before anyone tells me the distinction between Sun employees and non-employees doesn't matter, I'm afraid it does. For example there are bound to be legal reasons. There's also the question of SCAs - Sun employees will be covered by the Sun one, but all other committers will have to have filed one before being allowed to commit. The fact that we don't check this at present is a flaw, not something we want to continue doing. -- Alan Burlison --
