Re: Systemd

[Ralf Jung <post@ralfj.de>]

Hi,

> The bottom line is I think it's fair to say you *are* locked into
> systemd journald in a way you weren't locked into syslog.  Claims that
> you can just write another one exporting the same API don't ring true,
> because I suspect the API will be about as stable as an internal kernel
> API.  Tracking an internal kernel API from outside of the tree is
> really, really hard work.

Just one little remark here, it was my impression that "you can write
another one" referred to implementing the journald API. That API is
covered by the stability guarantee, and has enough users by now that
some form of backwards compatibility is necessary even for practical
reasons. Running logind without journald, I can imagine that this is
close to impossible (is there a compile-time switch?), but as mentioned
earlier in this thread, one can configure journald to do almost nothing.

In fact I am surprised that what we seem to end up with is an
alternative implementation of some internal systemd APIs (called
systemd-shim), which *will* break in backwards-incompatible ways -
instead of a reimplementation of the API that devs (of policykit or
various DEs) care about, namely logind. I still think that's the
long-term solution that will (should?) emerge - we'll see ;-)

Kind regards,
Ralf