[OutOfHere]

The problem is that these companies aren't hooked on extensible open source software as they should be, on something that gets reused across towns. The problem space then ought to be limited to developing or configuring town-specific extensions to the open source software, plus the associated devops. As with anything else, the goal ought to be to morph the development problem into a configuration problem.

Switzerland recently made open source mandatory for the public sector. Whether it is mandatory or not in any geography, its use ought to be something to strive for, not only to maximize reuse, but also to achieve a higher quality result in the process.

[MadnessASAP]

The problem almost always comes down to how governments do procurement. In the name of fairness and preventing a variety of immoral acts it is extremely difficult for a government to point at an existing product and say "we'll take that". Instead they must create a list of requirements, debate said list of requirements, pass the list around various departments completely separated from the original procurement, and finally of you're lucky it'll be sent out for suppliers to bid on. Obviously any attempt, perceived or real, to write the requirements against an existing product will result in lawsuits.[1]

The result is that the product almost always ends up being some bespoke pile of scraps built to a years old list of requirements that only vaguely aligns to the original problem.

It's a market that requires extreme patience and a building full of lawyers, it's not something that many companies are interested in participating in directly and pretty much impossible for a open-source organization, not that they'd even want to.

[1] I've watched a company dream-up a non-existent solution just so they could sue the government for an "unfair" competition and be compensated. I almost wish the courts had forced the government go with the imaginary product just to watch the company collapse trying to deliver something they had no intention of and would not be able to actually provide.

[OutOfHere]

> Obviously any attempt, perceived or real, to write the requirements against an existing product will result in lawsuits.

The obvious solution here is for the town to give the open source software an initial try on their own, then include it in the requirements as a preexisting use. In any case, the bid and contract then is not for a new product but for integration instead.

[MadnessASAP]

A jurisdiction can't just do something, that's the whole problem. The moment a dollar needs to be spent, a resource committed, whatever. It becomes a procurement problem.

Further to that there's nothing stopping a supplier from either claiming integration with another product is unfair and prejudices the competition or simply importing the data into their proprietary system and leaving the FOSS solution as some vestigial object hanging on the side.

I'm not joking when I say the system is absolutely beyond fucked with no obvious solutions. Most of it is so mundane and boring you won't ever here about it but ask any public service employee about their experiences in trying to get absolutely anything, from pens to planes to platforms the stories are all the same and many have tragic comedies of when the whole thing went wrong.

[OutOfHere]

Jurisdictions do have their own software staff too whose hands are not tied. They're free to get something started with an open source project. I don't see anything being financially procured at this stage, so the procurement problem doesn't apply.

The contractor can claim what they want, but the legal case is strongly already in favor of the jurisdiction if something is already bootstrapped in-house using open source.

[concinds]

Open source also removes dependence on these software monopolies (with high barriers to entry for competition, due to the complexity and diversity of local regulations). Tyler Inc has no incentive to create non-garbage software.