> Alternatives to solve this problem is to make some product based on debian
> and sell it, and some parts of that work distribute back to debian itself.
> (Whole work cannot be distributed, since company will loose it's competitivness
> and loose market to another company).
> Not necessarily. Competitiveness is not only achieved by proprietary
> control of software. And the network effects of wide adoption may be
> much more effective in terms of getting developers paid (not just you,
> but many developers).
Wide adoption can be achieved only via popularity of given asset.
Popularity can be achieved only via good developers / contributors.
Hmm... Anyway - it can be also measured from quite many angles
(like marker share, etc...) - that may I will not continue this subject.
If I'll create some side site - metrics needs to be somehow in-place.
> or do you receive some money for improving debian and placing it into open
> source community ?
> Yes, currently for Linaro on behalf of ARM, but previously for other
> companies (Toby Churchill, Aleph One)
Ok. There are quite many companies dealing with different permutations of linux
- what Linaro does ?
> I had also some ideas about using intermixture approach - like
> I'm writing closed source code and set it's price. (How much I think it costs
> and how much I've put hours into that work).
> Anyone can make same source code as I did by themselves, or buy my own source
> code "license" from me. In case if they buy source code from me - it reduces
> my investment into source code, and eventually makes my source code open again
> for public eyes.
> "Almost free" idea refers of starting with closed/propriatory code till it
> becomes open due to people are investing money to make it open source code.
> Now I'm also wondering whether I have reinvented some wheel which was invented
> I'm not aware of anyone using exactly this model, probably largely
> because it's almost equivalent to selling the
service of writing
> software, so usually people just do that.
The main problem is that they use binary distribution after that, and
source code remains in premises of company owning source code.
No improvement or derivate works will be born out of that closed source code.
> Your way allows you to keep
> control of the licencing, so that you can open it up when you 'break
> even', but it also gives you all the risk of not succeeding. Most
> software authors find it easier to let someone else worry about the
> risks and licencing. Some reasonable (and I suspect increasing)
> fraction of them get to work on free code too.
Yes... Licensing and managing licenses might get difficult / expensive.
Also trying to convince someone that this is right way, not that one, also might have
it's own problematics.
That's why this needs to be half-automatic web site with automatic money and source code
flow in all directions.
"Developer for hire" -kind of web site.
You come to that web site. If you want to hire a developer - you describe the problem,
constraints, requirements etc tec - and make a post to that web site. Developers accept
the offer and work on given item. And then can return some demo / brief summary back to
owner of item. After problem is solved, developer specifies how much time he has spent
and how much the solution will costs. Each source code license decreases solution costs,
and after reaching a zero - source code is exposed to free. Free source code can be
commented, bug fixed, and derivate works based on it might be born.
Roughly like this... May be all I need now is some invester who would like to hire me for making
this kind of web site. (Applying the same description as above to source codes of that web site)
But really - I could use some free time on this one... Do you have a means to set up
custom web site / forum on some debian.org ? (E.g. debian.org/developer4hire/index.php ?)