Re: Interactive guide for new users

[Eduardo Mercovich]

Hi Eli.

[...]

> > >> 1. distributions first [...]

> > > I don't see any need to do this, certainly not as the first step. If
> > > someone wants one of those versions, they don't need us.

> > New users don't know they exist... we are only opening their eyes.

> I think they already do know.  Moreover, I think they will decide to
> try those before they get to see what you are planning to make happen.

Those things I prefer to measure them in an experiment. :)
Anyway, I totally agree that it's not our priority now.

[...]

> > Perfect. Is it there any agreed upon list of things on this front?

> Not only don't we have an agreed-upon list, we don't even have a
> proposal for an exhaustive list. [...] If we can come up with such a
> list, discuss it, agree on it, and have a reasonable grouping (if
> needed), that would be a huge progress towards making everything
> discussed here close to happening. 

Then, since such a list seems valuable (before discussing the value of a
common vision), I propose a process to generate that list based on some
simple tested and working principles (*1): 

1- Let's collect a list of all the proposed preferences, by any member of
this list (to start). Everyone is invited to propose as many as it
wants. 

2- Then we prioritize with the same group (this list) by giving each
person a certain amount of points depending on the total amount of
preferences. It could range from 40% if they are few, to ~50% if they
are many, say, more than 20. Just as an example, if the total collected
options are 25 (since many people will agreed on some basic ones and
have a particular one to add, it could be bigger, of course) we may
give ~14 points to each person. 

3- Each person can cast it's points as it likes: all of them to 1
preference, 1 to each preference while it's points lasts, or any
combination in between. In the previous example, someone can put their
14 points to show-line-numbers, other to 14 different options (1 point
per option), and a third 5 to show-line-numbers, 6 to CUA-or-Emacs
cut-copy-paste, and 3 to show-menu-bar. 

4- We sum the total points per preference. Order from more to less
points. If they are few (say, less than a dozen), they can go all. If
they are too much, we can cut using any agreed upon criteria like a kind
of Pareto (66%~80% of the total points awarded) or other (the 1st N
items). 

If this seems ok to all of you, with this process we will have gone from
many persons with different opinions to a mostly agreed of the most
accepted (or desired) preferences to be set for the general editor
preferences. And it shouldn't take more than a few days at most (we have
to see what tool to use, since I assume that proprietary tools are not
acceptable; Limesurvey allows this, even if it's a bit oversized..). 

This process can be repeated for each "section" or group that we may
like (like org-mode or whatever you may find valuable after this). 

Now, after the list-making, we can make the screens (they don't have to
be functional) and test them with users as I commented on a previous mail. 

What do you think? 


(*1) From Sociocracy, maximize (democratize) input and optimize decision
making, plus the acceptable options range; from agile, let's try to have
something good enough working to validate our assumptions and learn from
reality before investing too much or taking too much time. 
--
eduardo mercovich

Donde se cruzan tus talentos 
con las necesidades del mundo, 
ahí está tu vocación. 
(Anónimo)