This is probably for working on Siemens' Sibas 32 family of products. This ia a large product range for their rail portfolio. You're not going to change the break control driver system on a high speed train unless absolutely required. The oldest OS that qualifies for interacting with these systems seems to be Windows 3.11, so it seems knowledge of such a system would be part of the requirements of a field engineer, just like knowledge of first aid is a requirement even though in theory you are only going to touch a keyboard, but interfacing with high-power electrical systems can have serious physical consequences.
> The oldest OS that qualifies for interacting with these systems seems to be Windows 3.11, so it seems knowledge of such a system would be part of the requirements of a field engineer
Working with it, yes. Developing for it? Completely different game, as you'll have almost zero chance to use anything modern, and that includes the tooling.
People are building Amiga applications as well as games and scene demos for much older 8-bit home computers with modern development tools. Even if Windows 3.1 doesn't quite have the same enthusiast following I would expect that it's possible with a bit of reverse engineering and tinkering (especially since Win3.1 is still quite small and simple).
TL;DR: I wouldn't be surprised if it's harder to support Win8 today than Win3.1 ;)
> People are building Amiga applications as well as games and scene demos for much older 8-bit home computers with modern development tools.
Yeah but these things are built from scratch. Your average ages-old Borland or whatever suite with its build script? That's a lot of work to integrate with a modern IDE, and honestly I wouldn't risk it in that space either due to the potential for things to go horribly sideways. So you need at least the OS and build tooling combination that was used for the last certified build, and if certification requirements are really strict you can't do that in a VM but have to develop on age-appropriate hardware as well.
It's still state-run, and 90% of the reason why Deutsche Bahn AG is in such a sorry state is the German government. Want to see privatization done right? Look no further than Germany's southern neighbour.
In Czechia, the track (much like the highways) is managed by a state corporation, while the trains themselves are partly state-owned, partly private.
From my 25 years of commuting experience, the results are decidedly mixed:
* the track-managing state corporation, Správa železnic, is a shitshow that causes a lot of unnecessary delays by just ignoring needs of passengers and cargo companies alike; true bureaucrats,
* the highway-managing state corporation, Ředitelství silnic a dálnic, is very well managed by the latest director (Mátl) and really kicked off many projects in last years, which is notable in a very NIMBY-friendly country like Czechia,
* the state-owned railway company, České dráhy, used to be terrible, but competition forced it into providing quite decent services within last 10 years or so.
* the two main private competitors, Leo Express and Regiojet, are now visibly cash-strapped (esp. Regiojet), so the overall level of service has stagnated for several years. High costs of traction electricity don't help either. As of today, there isn't really a qualitative difference between boarding state-owned or privately-owned trains.
As you allude to, whether it's a public service or a publicly owned corporation is probably less important.
Switzerland just pays four times as much per capita for railway infrastructure.
And of course, it's small: a bit more than a tenth of the area, a tenth of the total operated railway length, a tenth of the population. There are economies of scale, but also system complexity that scales super linearly. Afaict, a train that takes four hours from start to finish is on the long side in Switzerland, not so much in Germany.
But I'm not saying that financing and management aren't the main problems. The current CEO of SBB has a background in computer science and railway transport. Many of the DB CEOs in the past 25 years had an MBA and a career in air travel.
It is a good recommendation to follow the "if it works don't touch it" motto, but it is also true that they have time and resources (you don't need that much!) to port this to a modern platform and perform all the QA activities that you need to be 99% sure everything works as expected.
Through one of my companies we were/are involved in many of these efforts (e.g. application virtualization) and the use of legacy software [1]. Win 3.11 is a relatively easy target. Much much easier than other projects we dealt with.
The problem with modern platforms is that they remove many constraints that kept complexity at bay. Combined with a large bureaucratic organization that is state owned and financed, Parkinson's law is going to hit very hard.
Hi, I was deeply involved in DBs Vendo project which completely replaced/renewed most of it's sales / customer-facing related backends and we were using Open Shift.
5/6 years ago... Everything could have changed again by now.
- https://news.ycombinator.com/item?id=39168469
- https://news.ycombinator.com/item?id=39160956