User bus conclusion

[Lennart Poettering]

On Tue, 09.11.10 21:08, Havoc Pennington (hp at pobox.com) wrote:

> 
> Hi,
> 
> On Tue, Nov 9, 2010 at 6:40 PM, Lennart Poettering <mzqohf at 0pointer.de> wrote:
> >> Isn't the new user bus simply saying that the session bus has a fixed
> >> address?
> >
> > This is definitely the main point. But there's more. For example the
> > user bus is strictly defined to be non-networked. For the session bus
> > this was still kinda left open.
> 
> I agree with Thiago; I think you're setting limits/rules around
> sessions. (Sessions will not have apps from other machines as a
> session member; users may only have one session per machine; some
> details of how you want to start a session and discover it.)
> 
> Renaming to user bus obfuscates things. (The subject line of the email
> for me didn't match what you guys propose in the body of the email.)

Well, initially I actually wanted to simply redefine a little bit what
"session" in the context of dbus means, but Ryan convinced me that due
to the changed semantics we should introduce a seperate name for this
kind of bus.

> User bus is misleading. If you have network-wide user accounts, then
> those users can have N buses, 1 per machine, not a single bus. The
> buses typically would all share the same home directory.

Well, this is certainly is true. But I am not sure that when we pick a
name we really need to ensure that the full semantics are can be read
from it. But then again, it's also fine if we did.

> If you insist on a new name for the bus in the dbus APIs, please use
> USER_ON_HOST or something, not just USER, for the enum value. But I
> think just recycling SESSION is OK here. What you're doing is changing
> how the OS will define a session. The bus still corresponds to a
> session.

if we want to include a hint that the bus is machine local in the bus
name, then I'd vote for USER_LOCAL or LOCAL_USER or so, rather than
USER_ON_HOST.

Lennart

-- 
Lennart Poettering - Red Hat, Inc.