subdomains for testing things

[Daniel Pocock]

On 08/02/18 09:44, Max Mehl wrote:
> Hi Daniel,
>
> # Daniel Pocock [2018-02-08 10:32 +0100]:
>> Could we have a dedicated sub-domain for anything like this that is
>> running as a test?
>>
>> Using a domain like "community.fsfe.org" runs the risk that it is
>> perceived as or used as if it were a supported service.
>
> In theory yes, but in practice I'm afraid this would cause more work and
> barriers for our volunteers who want to establish a new service.
>

How does using a domain with the word "test" in it somewhere create more
work?  What about the possibility that people using the service by
mistake creates more work too?

Creating a barrier is a deliberate thing though: it mitigates the risk
that wider parts of the very large FSFE community start to rely on it
before the organization makes a deliberate decision to accept and
support the service.

Some effort also has to go into choosing domain names for services,
especially in a case like this where it is not clear that
"community.fsfe.org" is the only choice.  That is bureaucracy but it is
hard to avoid it.


> To avoid confusion, we explicitly decided to *not* advertise this or
> other beta services in our communication. When I shared the link on
> lists like this, I always noted that this is still not in production.
> I'm not sure that the technical layer you proposed will add any further
> clarity.
>
> Please also see our guidelines for "volunteer run services":
>
>  https://wiki.fsfe.org/TechDocs/VolunteerRunServices
>

There is a distinction between people volunteering to maintain a service
and the association choosing to rely on a service.

This is particularly important in cases where two services do something
similar (e.g. Discourse acts as an alternative to the existing Mailman
service).  If half the group uses one service and half the group uses
the other, you split the organization or you double the amount of effort
required to community.  Metcalfe's law[1] comes to mind.

As another example, the GA would have to make a decision if we were
going to stop using our GA mailing list and use Discourse instead, it
wouldn't just happen automatically.  So a decision by one group of
volunteers (the people who want to run Discourse) impacts other groups
and therefore it is a decision for the wider organization to consider. 
Does that create barriers and bureaucracy?  Yes, but isn't that better
than the organization splitting over different communication tools?

Regards,

Daniel


1. https://en.wikipedia.org/wiki/Metcalfe%27s_law