Also, if there's a genuine need to keep IE around for longer, then you have some options, e.g. deploy Win 10 LTSC for specific legacy use-cases, or publish IE11 via Citrix.
Why?
Surely it's not a greenfield project.
For that type of project, it might be worth it to write a detailed spec for how the back-end talks to the front-end so in the future it's possible to write a more modern client.
All of that on top of a system which has likely changed in the intermediate years. Basically, you might as well milk the existing system for as long as possible and then replace it with a completely new system that probably runs on newer hardware (1990s instead of 1980s), is much more maintainable and does things in a more modern way.
Not sure when was that app written, but the control unit was assembled just months ago.
'In development'. Does this mean v1 hasn't even shipped yet?
yes
Edit: I've seen a later comment mentioning it was started a while ago. But still... this is a dead platform. To me this is on the level of launching a new app targeting Windows XP. It's just a disaster
None of it could be upgraded.
Which would have been fine except that it had dependencies, and those dependencies had dependencies, and... you get the idea. I got the distinct impression that were it to fail or be unplugged every system in the business might slowly and inexorably fail over a period of days or weeks, eventually grinding all trading activity to a halt.
They were trying to figure out what to do about it and, fortunately, the guy in charge of replacing it really seemed to enjoy that kind of problem but, for all I know, they're still running it.
People should know that hardware will become obsolete or fail, that software will become unsupported, the supplier might go bust etc.
You need to be able to run generations of IT in parallel so that you always have newer hardware/software getting used/proved alongside the old nonsense and if you will need to replace an aging system, you better have seen it coming enough years earlier to give you time to find a replacement.
Also, large corporates have a habit of over-customising what they want so a new system has to be completely bespoke = slow and expensive to replace.
Depends on the constraints. Qt was around and mature for many years in 2006. Ruby on Rails had it's first release mid 2004. Django mid-2005.
Again it depends on the constrains, what you can "the best choice in 2006" is limping now. Meanwhile an app built in 2006 on Qt/RoR/Django would be on a platform that is currently still being considered for green field devt.
Without knowing all constraints I'd be willing to say that IE+applets was not the best choice in 2006, neither was: GWT, WebSphere or any of the MS GUI toolkits around in that year.
I'm still on the fence wrt Vaadin. It was not around in 2006, but soon after. Never used it, but it seems to still be actively maintained to this day.
But the lack of obviousness does not make IE+applets the best choice in 2006.
The timescale for a supplier to be in-business in order to prove their safety-critical credentials, the time it takes experienced developers to learn a language or framework and prove that it works, the time it takes to find, specify, build and deliver a project in these circles means you would be hard pushed to find anything that is less than 10 years old being used outside of startups.
now it's time to sell it hard to management
I see no devastation here...