[Qgis-developer] Sponsoring Development.

Ramon Andinach custard at westnet.com.au
Sat Jun 18 08:53:31 EDT 2011


Hello all (developers and non-developers both),
Consider this a reaction to "have you considered sponsoring development". From a non-developer. Who is still trying to understand how the sponsoring development idea works.

It is likely to be a bit long. If you have time, I'd appreciate it if you could read and comment on the long version. If not, this is a short version:
1. For me it would be easier to sponsor development, if I knew what it would cost.
2. Numbers of people are asking for the same features, and numbers of people are saying they're small and can't afford development.

I've had a bit of a conversation with Paolo, and now realise that 1. is a chicken and egg situation, and hence awkward, but 2. could be helped by making groups of people who want a feature to sponsor it.

So I'm wondering, from those who've had experience;
a.) how does one get around, the "I'd like a cost estimate but the cost estimate takes time and about half the work" problem?
b.) has anyone tried putting groups of sponsors together? If so, how did it go?




----
This is the long version. 
It's mostly what I'd been through with Paolo, but expanded a bit more so my thought process is (hopefully) a bit clearer.[*]

I'm attached to one of the many small companies that make use of QGIS. I had thought that it would be good to provide support to QGIS, and I had been thinking in the form of sponsoring development to go after either a few annoying bugs or one or two features that I'd like. While I was thinking about this a few thoughts occurred.
1. For me, and I suspect many others, it would be very useful to be able to go my boss and say, "I'd like us to support QGIS develop <a useful feature>, and it will cost this much." ie, have a good estimate.[!]
2. I'm also noticing;
  a. a number of features have been requested by a number of different people. A good example of this is multi-column legends in the composer.[@]
  b. a number of different people have (essentially) said, "we're really small and can't afford to sponsor development."

I remember reading on one of the QGIS lists that many feature developments were not as expensive as most people thought, I recall €2-300 being mentioned.[#]

Somehow I've put these together and bits together and realised that if there were 10 users who wanted <a useful feature> then suddenly they only need to provide €20-30, which is much more manageable.

I've asked Paolo for some thoughts and Paolo pointed out that:
3. There have previously been problems with payment after the development was done, and
4. To get a really good estimate, this often costs about half the total cost of the project so they may unwilling to work through to that point with out some certainty of payment.

----
So from my non-developer stand point this makes 1. and 4. a problem. But for the purposes of going forward, there are two ways of short-circuiting this that might work:
5. Developer takes initiative - to attract development, features wanting sponsorship get worked to a point where an estimate of total cost can be made. (Probably not viable - see [%])
6. Users take initiative - For the purposes of getting a group of sponsors running, start with an assumed cost, which can be refined later. 


So, from the User Initiative end (6.) a work-flow might work like this:
7. A feature is nominated for work. A bunch of people that want that feature get together and work out the following:
  a. A minimal set of things the feature must do.
  b. A nominated lead for the group.
  c. A contribution that each member of the group makes, based on an assumed cost. (Say €300). 
  d. Decide on what to do if the development cost more than expected.
  e. Decide on what to do if the development costs less than expected.

8. Find a willing developer.
  a. Sound developer out on scope, and adjust according to what is actually achievable.

9. Once, we have developer available and willing, then each member of the group contributes at least half of their total commitment to a nominated place. No development until money in bank.
  So, hypothetically, if there was a group with 10 members and a €300 project, then each contributes €15 now. This should be enough to make people feel committed to seeing something out of the development.

10. From there, the group would need to be able to see progress. In my industry it is normal to see weekly status reports from contractors doing projects over more than a couple of weeks, and I'd like to see that happen. Nothing fancy, just "this is what I did this week, this is what I'm going to do next week, this is what's left to do," style of thing.

11. Then I suppose the group would need something they could test.


----
If we can get to a point where there's a process that people thing might work, then yes, I'm prepared to help trial it.

Any thoughts?
Particularly I'm curious if there are thoughts on:
1. Getting around the "a cost estimate makes it easier for me to get funds" against "a good cost estimate is most of the work" problem.
2. Viability of this idea.
3. Thoughts on managing groups of sponsors.


-ramon.

______
[*] For the record, I was asking for advice, and we rapidly got to the point where it was getting complex (as you can see) and Paolo suggested that it be brought out on the the mailing list. Please don't think that this is a final-ish suggestion.
[!] In my industry a good estimate is within about 10%. So, I'm not expecting dollar perfect.
[@] I had heard that some of the Swiss users might be getting together on this one, so this is just an example. On that thought, big swiss users or little swiss users?
[#] but I can't find it, so I could be hallucinating. Either way, I'll run with these numbers.
[%] Yes I do know there are a limited number of developers, who are already over-loaded. If this way did happen, then the cost estimate would include work to the point of estimate, work to finish (and I think it would also be reasonable to include an "I'm taking the initiative" fee here too).



More information about the Qgis-developer mailing list