Please, the big problem of governments and administrations is not using Linux or Windows, the operating system is irrelevant... The problem was probably reliance on crappy software written by shitty companies that only try to milk them of more and more money. The fact that it took them a decade to make the transition was most likely because of the shitty non-standard software they were using and the time it takes to retrain people used to working with backwards and user hostile software. If you are a gov accountant or clerk and you accounting software had such a bad ui that it took you 3+ years to properly learn to use it, of course you will resist any kind of change because you just can't imagine things can be any better.
And this is because governments and administration never choose any kind of product because it is "easy to use", "simple to maintain" or "easy to extend".
Even the switch to Linux was made for the wrong reasons, that is for trying to find something cheaper and better ("better" seems to mean "more secure" in this case). NEWSFLASH: As a government/administration, you should never look for cheaper and/or better things, you should look for something that is "barely good enough" (even when it's a bit worse than what you already have), but easiest to use without training, easiest to teach people how to use it, easiest to replace / throw away and cheapest to extend, even if it happens to be both worse an more expensive.
The fact that they happened to choose Linux, which is probably a pretty good fit for them, was an accident, as they did the same old stupid thing of going to look for something cheaper and better.
>And this is because governments and administration never choose any kind of product because it is "easy to use", "simple to maintain" or "easy to extend".
The same is true for enterprise software. It's not optimized for the users. It's optimized to appeal to ones paying.
Of course the lowest bidder is supposed to win, as all bids are supposed to fulfil the requirements laid out in the contest and these requirements are assumed to be fully sufficient for the task at hand. Theoretically speaking, of course.
Hm. I’m obviously not a lawyer and also not a software engineer in an enterprisey setting. But some ideas might be:
- all internal data storage formats have to be completely specified
- an open data format (RTF, PDF, txt?) has to be supported for import/export
- the source code has to be licensed suitably, e.g. GPL
- open/platform-independent toolkits (say Qt, GTK) have to be used
As I said, I’m not a lawyer and not a software engineer, so these are likely entirely unsuitable. And while they not ensure maintainable software, I think they certainly would not be detrimental.
I'd make damned sure that a format in the preferred form of the work for making modifications to it be specified.
PDFs can be modified, but you really don't want to do that. Far better to get source in LaTeX, DocBook, RTF, Markdown, or some similar format.
Better: insist that the tool not only be capable of outputting the chosen format, but that it do so by default, as a matter of course, and without any error, warning, or other dissuasive dialog (e.g., "some formatting changes may be lost") in the process.
While it's not the only company to do so, Microsoft's gaming of file formats for lock-in advantage is legendary (and yet another reason I avoid the company's products where at all possible).
Monetary consequences for failing to meet Service Level Agreements; contract-mandated maximum time periods for patches; support for the both oldest and newest OS-vendor supported versions within the window of the contract.
And this is because governments and administration never choose any kind of product because it is "easy to use", "simple to maintain" or "easy to extend".
Even the switch to Linux was made for the wrong reasons, that is for trying to find something cheaper and better ("better" seems to mean "more secure" in this case). NEWSFLASH: As a government/administration, you should never look for cheaper and/or better things, you should look for something that is "barely good enough" (even when it's a bit worse than what you already have), but easiest to use without training, easiest to teach people how to use it, easiest to replace / throw away and cheapest to extend, even if it happens to be both worse an more expensive.
The fact that they happened to choose Linux, which is probably a pretty good fit for them, was an accident, as they did the same old stupid thing of going to look for something cheaper and better.