Bug#727708: systemd and upstart, a view from a daemon upstream

[Russ Allbery <rra@debian.org>]

Ian Jackson <ijackson@chiark.greenend.org.uk> writes:

> So overall my conclusions at this level are:

>  * socket activation is an attractive implementation target for an
>    upstream daemon author.

>  * upstart's SIGSTOP protocol is an attractive implementation target
>    for an upstream daemon author.

>  * systemd's readiness protocol is an unattractive implementation
>    target for an upstream daemon author.  I think this is an important
>    weakness, if it remains unaddressed.

I've indicated my disagreement with the last point elsewhere, but I wanted
to note that I agree with both of the first points.  I have conceptual
problems with upstart's synchronization protocol, but it's certainly easy
to implement (in the form of a new flag) and test.

For comparison purposes, the *total* burden, from my upstream perspective,
of the two options was:

* systemd: 14 lines (8 lines of code, 6 lines of build system)
* upstart: 12 lines (6 lines of code, 6 lines of documentation)

Since upstart synchronization required adding a new command-line flag, it
needed documentation of the new flag; systemd's synchronization support
didn't strike me as something that required documentation beyond a note
that it was supported.

Both of these are effectively trivial, and my current intention is to add
support for both protocols to the daemons that I maintain.

-- 
Russ Allbery (rra@debian.org)               <http://www.eyrie.org/~eagle/>