Re: Next action, survey?

[Katherine Cox-Buday]

On 9/6/23 3:34 AM, Simon Tournier wrote:


Well, from my point of view, the next steps are:

  + Propose a survey (questions), open a new thread (or a bug report) for
iterating.


I'll open a bug report and start logging notes/suggestions against it. 
In it, I'll include your points:



  + Review how to distribute the survey >   + Prepare the skeleton for 
generating a report about the stats from the
survey.

Maybe the person or group of people in charge of this survey could be in
touch with Emacs folks.  They did a survey on 2022 [1].

1: https://lists.gnu.org/archive/html/emacs-devel/2022-10/msg01824.html


I'll start reaching out to people, specifically people who have done 
surveys for projects similar to Guix.


I'm not an expert in how to design surveys, and I assume there are some 
out there who could help us design one to do things like avoid skewed 
data (e.g. your point about how the survey is distributed), and not 
frame questions in a way that elicits a certain response.



A survey and the associated report will act similarly: list what we more
or less already know, and then make easier the path for improving.


Agreed! I think the trend-analysis of survey responses over time will be 
most interesting.

Next action, survey?

[Simon Tournier]

Hi Katherine,

On Tue, 05 Sep 2023 at 20:58, Katherine Cox-Buday  
wrote:

> OK, great! What are next steps, and how can I help?

Well, from my point of view, the next steps are:

 + Propose a survey (questions), open a new thread (or a bug report) for
   iterating.
 + Review how to distribute the survey.
 + Prepare the skeleton for generating a report about the stats from the
   survey.

Maybe the person or group of people in charge of this survey could be in
touch with Emacs folks.  They did a survey on 2022 [1].

1: https://lists.gnu.org/archive/html/emacs-devel/2022-10/msg01824.html


> Before I opened this can of worms, I was planning on trying to work on 
> the Go branch/packages 
> (https://lists.gnu.org/archive/html/guix-devel/2023-08/msg00058.html). 
> But since I'm the cause of all this, and it will indirectly help me with 
> Go things (which is why I brought it up), how can I help?

Having a survey will not be wasted, IMHO.  As you said, although there
are bias – the first one is the way to distribute it ;-) – it will help
to inform about the status.  Having some stats about Guix users and
contributors will explicitly spot what we probably already know but
having that will help, for sure.

Well, something unrelated.  For instance, I spent some time to run an
experiment for checking a real-world worst-case scenario when
fallbacking to Software Heritage [2].  This lists all the frictions.
All these frictions were more or less already known but having
explicitly collected them makes easier the path for improving.  That’s
somehow the measure you were talking about.

A survey and the associated report will act similarly: list what we more
or less already know, and then make easier the path for improving.

2: https://simon.tournier.info/posts/2023-06-23-hackathon-repro.html


Cheers,
simon