Crazy Example 1:
- Go to Wincor-Nixdorf, Diebold and NCR. Show them that Windows on ATMs had bring them more problems than before.
Since they already have an investment on OS/2 apps for their ATMs, give them a business case to support an Open Source OS/2 project, so they can have their own OS for ATMs.....
:?
Heck, even showing them that funding continued development of a closed source OS/2 would be beneficial.
We (the OS/2 and eCS community) are and have been the recipients of various fixes and updates to OS/2 largely because of such things. An IBM client (who refuses to or has not yet migrated off OS/2), through their support contract or on a pay as needed basis, engages IBM to fix or update various components in OS/2 or write needed drivers.
Thus, I think such a method could still be beneficial to the eCS community with the right parameters in mind by such a company - such as an updated (64 bit anyone?) kernel and continued driver support - which we (via Serenity) would then benefit from.
Note: I am not saying it would be the best solution (I'd rather the right people had access to the source to update it on (our) own). But it is a feasible solution in ensuring that OS/2 runs, takes advantage of, and will continue to run on the newest hardware. It is a start...
The problem still arises, in how to convince such companies that such a strategy is beneficial. I am sure convincing them that they would have avoided such headaches by sticking with OS/2 is the easy part.
I am not so sure that convincing them to stick with (or switch back to) OS/2, when they'd need a company willing to support such a large infrastructure, would be nearly as easy.
And though I would love to volunteer for such, mine and my bro's company is a mere 2 people (he and I) and some outside contractors on an as-needed basis. And sadly, we are talking about needing a company that can do such support, installations, setups and maintenance on a nationwide (pick your country) - or even global scale. Including writing
new third-party software that will run on the ATMs with all the neat "bells and whistles" - which while possible, will require OS/2 programmers on a level that we seem to be lacking.