> > > I don't think so - indeed policy says that packages shouldn't depend > > > on environment variables for their correct behaviour(3.9). I don't > > > disagree that http_proxy as an environment variable is common > > > practice, but I do disagree strongly that this should be made policy - > > > you'd have to do away with policy section 3.9 for starters... > > > > Uh? Environment variables should not be needed for proper operation, but > > you are taking that too far. In your world there wouldn't be a place for > > environment variables at all! > > In an environment in which the proxy is required then the environment > variables would need to be set for proper operation. It's a bit of a stretch > but not really an unreasonable interpretation. > > I would be against giving this system the "debian stamp of approval" by > putting it in policy. If we're going to make something policy lets get it > right. > > If we make the environment variables policy I say we make an /etc/proxies file > policy as well and require packages have a wrapper which checks > > command line flags (if applicable) > environment variables (required) > home directory dotfiles (if applicable) > /etc/proxies (required) > > Note that I'm not sure setting things in /etc/environment (or /etc/profile, > which is a worse idea) really works the way it should. does it override the > users' preferences saved in their dotfiles? That would be wrong.
Look. This is widely implemented... A few people like you are proposing things just to modify things, but without offering any improvement. You are trying to "dilute" the thing, to turn it in an endless discussion. The `http_proxy' thing is a fairly reasonable thing to do. Non-amibiguous, widely implemented, widely known. It's very similar to having a LANG variable, EDITOR, PAGER, TZ, and we already have that (don't forget that sensible-* is just a way of providing a way to get programs to comply with the policy without having to modify them). So.. what do we do? Do we approve this?
