[Bug 71384] improve account storage GInterface

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=71384

--- Comment #4 from John Brooks <john.brooks at jollamobile.com> ---
(In reply to comment #3)
> >   account_name = mcp_account_manager_get_unique_name (self->priv->am,
> >       "ring", "tel", "account");
> 
> This means that if you have more than one modem, the accounts' names are
> assigned at random, making the account object path uncorrelatable with
> anything useful.
> 
> Does oFono let you attach arbitrary data to modems? If it does, you could
> tag the modems with their object path tails and display names.
> 
> I realize this is necessary if you want the first account to be
> ring/tel/account0 for compatibility or something.

Reasonable point. Don't worry about the compatibility case; nemo is the only
one that cares, and we can fix all of that software to handle it properly, or
patch the name back to what it was. Ofono at least provides object paths for
modems, which could probably map easily to account names.

> >   PARAM ("always_dispatch", "true");
> 
> John: do you need this, or is it cargo-cult? I never liked this special
> case, so if we can get rid of it, I'd be in favour; or if you need one of
> its (possibly several) effects, I'd like to pin down which one(s) you need,
> and document why.

Yes. I use the connectivity management in MC, and the ring account must always
be connected regardless of connectivity. From a quick glance, the name is bad,
but I don't see that it has any other effects?

> >   return TP_STORAGE_RESTRICTION_FLAG_CANNOT_SET_PARAMETERS |
> >       TP_STORAGE_RESTRICTION_FLAG_CANNOT_SET_ENABLED |
> >       TP_STORAGE_RESTRICTION_FLAG_CANNOT_SET_PRESENCE |
> >       TP_STORAGE_RESTRICTION_FLAG_CANNOT_SET_SERVICE;
> 
> I think this suggests how we can supersede McdAccount::always-on:
> 
> * if we CANNOT_SET_ENABLED, then Enabled is forcibly set to what the
>   storage backend says it is (instead of TRUE, as always-on does)
> 
> * if we CANNOT_SET_PRESENCE, then AutomaticPresence and ConnectAutomatically
>   are forcibly set to what the storage backend says they are,
>   and RequestedPresence is forcibly set to
>   (ConnectAutomatically ? AutomaticPresence : (OFFLINE, "offline", ""))

If the behavior change isn't a problem elsewhere, that would be a good
solution.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.