[benilov]

Grassroots sounds nice, but the hard work isn't writing the code - it's convincing those within government that user-centered services and open source are the way to go. I believe that this can only be achieved given a strong mandate, from within government itself.

[danmaz74]

Based on my experience in Italy - with my former employer we did some work with public administrations - my idea would be to start with cash-strapped local administrations and the most forward-looking of their suppliers.

These administrations basically have all the same needs, but many of them are served by small suppliers that always have to reinvent the wheel with some half-cooked proprietary solution.

If there was a well-designed, free open source GPL solution that already coverered 80% of the requirements, that could give a competitive advantage to those who based their offer on that solution. They would then compete only on service and customization.

As I see it, the problem would be mostly to kickstart the development of that GPL platform - no for-profit company could do that, it would be against their interest. But if a big and committed enough community started it, then it could create a positive feedback loop that could be very difficult to stop. IMHO this could be a field where the GPL could give its best (I'm not pasdaran of the GPL, and released my own little open source contributions as MIT).

[benilov]

Well, the GOV.UK code (https://www.github.com/alphagov) is your 80% (or more likely higher) solution. It's based around many micro-services that integrate over HTTP, so it's extremely extensible.

However, even with the software being perfect, you still have very high barriers of entry to overcome: 1) the local authority is likely to be locked into a 5- or 10-year contract with its current supplier 2) the choice of supplier is rarely (if ever) based on technical excellence/licensing strategy. Usually the people commissioning the IT don't tend to have a very deep understanding of technology, licensing, etc

[danmaz74]

If the gov.uk code covers a lot of the requirements for local administrations in continental EU, that's terrific. Honestly, for now I'm just thinking about this idea, but I don't have the time to study it.

As for the contracts, at least in Italy I've never seen one longer than 3 years, and practically by law the deciding factor is, most of the time, price (which is a big problem on its own).

[benilov]

Well ok, the GOV.UK is the 80+% solution for publishing; transactions is another matter, although hopefully when the Digital by Default Service Standard starts kicking in (specifically the open source provision: https://www.gov.uk/service-manual/digital-by-default#criteri...) then a lot of code around transactions should start appearing.

[danmaz74]

That's interesting, I'll look into it if I'll ever have the time ;)