Due to the scale of the maemo.org community and the less-than-optimal sprint system for the activities in the community I have lately pondered if there was a way to try and handle the many activities this community could be doing.

Initially it started out as an idea for Mer, but I think it might scale to the requirements of the community.

The principle is that in a community we have a lot of interwoven mini projects that often depend on the completion of others. Along with that, we have the problem that there's no 'staff' as such (excepting maemo.org paid staff which is a scare resource) to allocate onto projects. The other problem is that newcomers into the community rarely know where they could contribute or fit in. We also have problems quantifying the amount of work people put into the community into karma.

So how can we deal with this? We are many talents in this community and we could easily do very complex and big activities -but- we aren't that well organised and we might need tools to help us become so. This post is about a hypothetical web system that might do the trick.

Introducing maemo.org miniprojects

First off, I'd like to describe a miniproject formally. A miniproject can be created by anyone. A miniproject may be given priority by either votes and/or council decision - this means it will be marked as a priority miniproject in listings in order to give the community work direction.

A MP (Miniproject) has two descriptions, one short, one long. It has a list of other miniprojects that it depends on to be completed before the project can be started.

Along with those, a MP lists a series of roles along the goal of this role in the miniproject. A community member can choose to be one of these roles and will work towards the goal of this role.

Finally, a MP has a status. Initially the MP is in either WAITING or BOARDING status.

WAITING status

The project is currently waiting for the MP's it is depending on to complete. When they are complete, the project changes to BOARDING status.

BOARDING status

The project is ready to accept participants and will accept anyone joining into the project as filling a role. When the team is full, the project moves to PLANNING status.

PLANNING status

The participants create tasks (this may be other MP's) amongst themselves, decide dependancies of eachother's tasks and decide how to do the project. If another role is needed, one is added and the project reverts to BOARDING status. When planning is completed, the participants move the project to IMPLEMENTATION status.

IMPLEMENTATION status

The participants run through their tasks and mark them as completed when they are. In this process they try to give an estimate on the time they spent on the task in order to help quantifying work put into it. They may add additional tasks in this phase.

When all the tasks are completed, the MP moves to COMPLETED stage.

COMPLETED

The project is archived and MPs that depended on it's completion can now begin.

A practical example

Mini project: N900 X-terminal cheatsheet

Depends: None

Description, short: Provide a wiki page with N900 X-terminal cheatsheet
Description, long: Provide a wiki page with some tips and tricks or the N900 X-terminal for end-users

Roles:
* Wiki editor: Write cheatsheet page
* Wiki integrator: Provide links to cheatsheet from existing wiki pages
* Artist: Provide screen shots of relevant places and icons.
* End-user tester 1: Acceptance of end-user usability of result
* End-user tester 2: Acceptance of end-user usability of result

On the MP web site, a community member would be seeing different roles that needs filling. He clicks the wiki editor role for the cheatsheet project and signs up for the miniproject. When the team is filled, he discusses with the others how to go about the project. They work at it and the end-users test the results and confirms them by completing their tasks.

The idea is to provide a place that people can see what they can participate in, participate and help move the community forward.

The website should be simple:

- anyone should be able to understand the concept/workflow)
- a list of roles to fill in, a project should state what completing this project would mean to the community (what projects can -then- start
- lists of active projects in progress, etc.

What we really need is a collaborative area to do great things and lift heavier things together than it would be possible to do on our own.

For many things I think it would even be possible to template projects (pre-existing pipelines of getting things done, etc) - such as artist (make icons) -> integrator (upload to repos) -> tester.

I'd like to hear your comments on this idea or even ideas on how we could use this or implement it in practice. Flames are happily accepted if they're constructive.

Interesting. Does it require investment in tools; from the sounds of it it's more process and methodology-led? Lower tool investment means it can be used sooner.

This could fit well with some things I've been lobbying for recently; and hopefully we'll see traction on before the end of Q1.

Who are the stakeholders if questions come up during development? Presumably the stakeholder has some kind of "sign-off" on what the testers test.

Stakeholders are always a big issue, and you need to have them. You need to have someone who cares about the project as if it were their own, to see it to completion.

Regarding stakeholders I thought about a TESTING phase where in they approve/disapprove. But for simplicity's sake, it would be possible to have stakeholder as a role as well and have them participate in planning phase to guide the work.

As Jaffa notes, this could help Brainstorm, too:

http://talk.maemo.org/showthread.php...961#post480961

Agreed, and I would call that role Champion.

We have increasing needs for such people in various areas.

We have t.m.o, #maemo IRC and email.

We all have our go-to people for help and shared interests.

Can you point to any maemo community project that has shown discoordination? Even any project with at least 3 active developers?

Unnecessary structure is a time waste.

liqbase (due to lcuk's incredible development speed), MWKN (very hard to discuss things with Jaffa, because of timezone difference, and sparcity of internet availability), and those are the ones I've been remotely part of. I'm sure there are *tons* of projects that do not properly leverage the tools available (and garage is mess, the UI is crap to use)

Stskeeps, I think it is quite a good idea, however, I would suggest to try to simplify from miniproject to a simple action handling system.

Standard action definition:
- title
- description
- owner
- status
- deadline, priority

We could add contributors, comments, logged history, dependency on other actions and probably remove deadline and priority.

The standard action statuses: draft, open, work in progress, closed, cancelled. That's pretty much what you were writing:
- draft = waiting
- open = boarding
- work in progress = implementation
- closed = completed
- cancelled is for "false alarms", when during implementation it turns out that this action is not needed/not possible.

While it is only a small modification to your proposal, this would give the benefit to be able to use it across bugtracker, brainstorm, council sprints, community championed tasks (or even for self-imposed tasks, like a learning path).

Either way, I think it would be important to implement a solution to the problem you have raised.

( these notes were taken from my maemo.org prezi, which is currently in proposals-devel, usual warning applies: http://prezi.com/qefcq69cpjkm )

Working a project with a team is quite hard to handle especially if you are far from each other. Good thing that we do have online project management tools that could assist project managers on supervising their workers.


online project management