2 ms·
I guess we just need the DBUS developers to get their heads out of their asses, that's all: https://lkml.org/lkml/2015/6/23/657 https://lkml.org/lkml/2015/6/23/
by thyristan 2mo ago
I guess we just need the DBUS developers to get their heads out of their asses, that's all: https://lkml.org/lkml/2015/6/23/657 https://lkml.org/lkml/2015/6/23/657
- remix2000 2mo agoNah, DBus itself is plain bad interface design. Perhaps they could just piggyback off Binder if DBus itself weren't beyond saving. Also, despite a common sentiment online, it's not at all a complex process to implement a Binder userspace (given you have a sensible IDL in the first place, that is…) Wrt/ torvie's narcissistic snarks linked above, the most important task for a kernel is still to facilitate fast IPC. The only viable forms of IPC in Linux are UDS and Binder. The former is slow and clunky. The latter is not a valid choice for DBus. So DBus devs have to manage slow, clunky, ill-suited legacy IPC. When a kernel can't be bothered to accommodate one of the most important operating systems built upon it, that seems really bad.