Re: Survey answers part 3: systemd is not portable and what this means for our ports

[Ondřej Surý <ondrej@sury.org>]

On Fri, Jul 19, 2013 at 11:14 PM, Russ Allbery <rra@debian.org> wrote:

I would *hope* a lot of Debian developers would do things like that, for

any of the options!  There's no substitute for actually trying the
software and seeing how easy it is to use, how well it works, and how
difficult it is to support.  There are a bunch of good reasons to install
packages, even if one isn't going to use them regularly.  Among other
things, it's often the easiest way to read the documentation so that one
knows what people are even talking about!

JFTR I did that (I already know upstart quite well from Ubuntu, so I just did install systemd on my work machine) and now I am replacing sysvinit with systemd on every machine I maintain. I have missed a supervised services from init(1) for too long (I can only +1 to Russ's mail about experiences with administration) and I have already used that to supervise crashing gitlab-sidekiq service.

I also did add support for systemd and upstart to php5-fpm (and kept sysvinit script) and I am quite confident I can support all three of them without any trouble. It has a learning curve, but I am willing to learn that (sysvinit scripts also have a learning curve - the idea that those are just simple shell scripts are just wrong.)

And if I could use upstart job to extort Ubuntu PHP maintainers to contribute back more, even better! :)

Anyway - this is the question for all proponents of systemd and upstart. It's quite obvious that we cannot reach full consensus (and we never will be able to reach one), but we have a body to make technical decisions - tech-ctte. Why we just don't shove it to tech-ctte and let the independent body decide? (Or GR, but I think that GR is an overkill.) This could stop this endless debate and force the participants to write the technical summary.

O.

--
Ondřej Surý <ondrej@sury.org>