isn't dbus harmful in a similar way to systemd?
It was pushed into linux by Red Hat, and now tons of programs depend on it unnecessarily.
it's also believed that dbus was created as a way for Red Hat to get around GPL restrictions.
>Strategically, it allows RedHat to sell its customers an opt-out from the GPL, while still effectively linking to GPL codebases. That hasn't been tested in Court of course, and I for one wouldn't believe any lawyer nor snake-oil salesman that it would stand up to inspection.
>If anything, I'd fire the lawyer who dreamt it up or okayed it, as he's a liability.
>Localising RPC doesn't stop it being a function-call; that's the whole point of RPC. All they're doing is exactly that, and labelling it IPC.
>Indeed when I ran the above analysis of the motivation, past an embedded developer, his immediate response was "Yeah, that's exactly why we [his company] love dbus: because we can ignore the GPL."
>systemd: removing dbus breaks dependency 'dbus'
really makes you think...
It is.
Pretty sure kbus is our only hope.
>>61748598
It's actually a very closely linked threat. systemd does most of its harm by reimplementing core OS functionality as DBus interfaces. There's been countless security bus due to AccountsService being complete shit, but the more core problem is that it's making different programs dependent on DBus services that only systemd provides.
I'm not convinced that Red Hat is trying to get around the GPL through DBus, there are much simpler ways than that.
at this point, I struggle to think of anything positive red hat has contributed to linux in the past several years.