Open Rails in Linux

[NW 2156]

Evening, With both the latest release of ORTS and Wine 2.0/2.1 it would now seem open rails can be made to run in Linux. https://appdb.winehq.org/objectManag...sion&iId=34781

The payed for version of Wine from Code Weavers also reports being able to run ORTS. https://www.codeweavers.com/compatib...ver/open-rails

I have not yet tried it for myself but I am hopeful it will work.

Robert

[cjakeman]

I have not yet tried it for myself but I am hopeful it will work.

Love to know how it works out. I see quite a few enquiries on the website hoping for Open Rails on Linux.

Best wishes,

[shadowmane]

I struggled with this for 2 days a few weeks ago. I couldn't get winetricks to install .net 3.5 sp1 no matter how hard I tried. I haven't tried since then, but I will take on the challenge again when I feel ready for the headache. It might not help that I'm going through the whole process on a Pentium D machine. I'm determined to make this work, though. The machines I have all have Windows XP on them, and I hate XP with a passion. It was the reason I ditched Microsoft for good in the first place.

Alternatively, I'm toying with the idea of running Open Rails on an Amazon Web Services instance and streaming it through Steam. Might work, might not. It will give me a powerful virtual machine to run it on, though.

[ftldave]

You can stop hating XP. It is so old that it's no longer sold or supported by Microsoft. No need to hold a grudge that long!

I don't know whether to pity or admire the efforts that Linux lovers go through just to try to get it to work, to do the things Windows and Mac OS X do out of the box.

I mean, it's kind of like wondering what to say when one of your friends proudly displays a little pile of homemade soap she spent 16 hours cooking up?

[shadowmane]

Your comments aren't helpful. I'll invite you to take them and go for a long walk off a short pier.

[lindsayts]

I struggled with this for 2 days a few weeks ago. I couldn't get winetricks to install .net 3.5 sp1 no matter how hard I tried. I haven't tried since then, but I will take on the challenge again when I feel ready for the headache. It might not help that I'm going through the whole process on a Pentium D machine. I'm determined to make this work, though. The machines I have all have Windows XP on them, and I hate XP with a passion. It was the reason I ditched Microsoft for good in the first place.

Alternatively, I'm toying with the idea of running Open Rails on an Amazon Web Services instance and streaming it through Steam. Might work, might not. It will give me a powerful virtual machine to run it on, though.

I had a go at trying to get Openrails running on Linux some time back (12 months or more ago) and I also could not get winetricks to install .net 3.5 sp1, install simply hanging. This is somewhat sad as the new route editor works perfectly under wine, so I have not yet even tried to compile a Linux native version of it.

The WineHQ seems quite specific and its also current so I may give it another go.

Lindsay

[ftldave]

Your comments aren't helpful. I'll invite you to take them and go for a long walk off a short pier.

In a forum you're going to see many points of view. Here's one that sums up what happened since 2008, "the year" for you know what that didn't happen.

Desktop Linux

[lateagain]

Your comments aren't helpful. I'll invite you to take them and go for a long walk off a short pier.

As the original poster states that reports are that this now CAN BE DONE..... perhaps you can explain how your comments are helpful? Have you followed these latest reports? Have you considered that most people will already have an OS and how few might have sufficient problems with that to even consider the complications of creating a dual boot Linux machine? I take it that would be needed? If you want to discuss why you're interested you may attract interest. Otherwise the XP hating comments seem pretty valid Slagging people off for daring to question your post is not what a forum is about.

[mikeebb]

This thread is good news (at least the early and late parts of it). Not sure it will work in Linux in a VM (which is where Mint lives at the moment) but will try it.

As for Windows (not just XP) hate, it's not worth the blood pressure. Most people using a PC (a dying breed, according to market surveys, though still a very large one) use Windows. That's a fact. However, if you use old hardware (or the newest, if you're wedded to a non-10 version of Windows - see recent reports about Win7/8 on Kaby Lake CPUs), or just don't want to deal with all the maintenance and configuration hassles of 10 (assuming you want to keep SOME information private), Linux is a valid alternative.

As with anything, there's a cost, and with Linux in my experience you pay with time and the need to develop some moderately serious geekiness - it's not free even if you don't pay for it in cash. But then, all Windows versions have required a bit of tweaking and management to make them run well (as opposed to just running, slowly). 10 requires a slightly different approach to the tweaks, that's all. A general purpose computer is not an appliance. So set up a VM in Virtualbox or whatever and try Linux out - some modern distributions (distros) aren't difficult to use at all. For average computer users, the software they like now will run (at least under Wine) or has a pretty good equivalent. Having OR shown to work in Linux is a good step forward in making Linux equivalent to Windows, for me. Now if only I could get ACD's activation process to work for Canvas, I'd have no excuse to stay with Windows...

[lateagain]

Interesting input Mike but I still don't see what Linux "brings to the party" for OR? Not being argumentative here. It just strikes me that I've no obvious (I may well be missing something!) issues running OR in Win7. Most machines ship with an M$ OS anyway so apart from the geeky pleasure in proving it will work I don't see why anyone would want to change?

BTW what's a Kaby Lake CPU? and if it's new? surely machines equipped with it won't ship with Win 7 anyway?

[NW 2156]

The potential could well be in drawing the attention of developers who could contribute to ORTS from the Linux world. This can be seen in following OpenBVE which runs in all three OSes quite happily.

Robert

[lateagain]

The potential could well be in drawing the attention of developers who could contribute to ORTS from the Linux world. This can be seen in following OpenBVE which runs in all three OSes quite happily.

Robert

....well from where I sit they'd be doing something more constructive working on OR? ....so good point!

[mikeebb]

Interesting input Mike but I still don't see what Linux "brings to the party" for OR? Not being argumentative here. It just strikes me that I've no obvious (I may well be missing something!) issues running OR in Win7. Most machines ship with an M$ OS anyway so apart from the geeky pleasure in proving it will work I don't see why anyone would want to change?

BTW what's a Kaby Lake CPU? and if it's new? surely machines equipped with it won't ship with Win 7 anyway?

Well, you have the key point right there up front: I'm moderately geeky and like to keep up on the competition. Yes, I know, Apple is the real competition, but you have to buy a whole new computer for that - you don't for Linux. Purpose is really to maintain an option should MS decide (as their Terms of Service allows them to) that my old hardware will no longer receive updates in Win10. I'm not going back to 7; it's already in extended support and hits EOL in only another couple of years.

Kaby Lake is the latest round of Intel hardware. The tech press (see Ars Technica, for instance) has been discussing the fact that MS now blocks support, if not installation, of Win7 and 8 on Kaby Lake chips, and Intel will not provide chipset drivers for Windows prior to 10. So if you buy a Really New Computer, you have to take Windows 10 with it. It's not a disaster, since despite threatening to do that for the later Skylake CPUs they apparently didn't do it and most Intel CPUs currently being sold are Skylake hardware. And if you have 7 or 8 working on older hardware now they continue to be supported until their stated EOL. It's just that if you buy a new computer with the latest CPU you have to take Win10 with it (and pay for it, either on your own if you build or from the OEM) - or install Linux. The geeks are freaking over this, but most people won't notice.

The story initially included the new AMD Ryzen CPU in the no-Win7-support category, but 7 & 8 have now been show to work with it including receiving updates. So far at least.

If you're running Win10 (or 8 or 7 with no plans for new hardware), and have it tweaked to run OK for you, then unless you're geeky Linux probably isn't worth worrying about. Keep using Windows; it's fine if properly configured. I just like to keep my options open.

[lateagain]

This takes us rather off topic ...EXCEPT that reality dictates that our hobby ONLY exists on machines built to support a much wider usage. The relationship between OS publishers (of any name) and hardware manufacturers seems to put the consumer right at the bottom of the priority list? Their R&D has to be paid for so they force new machinery and new software on the consumer by stopping support on their established product. I can see that Linux offers a "safe haven" from their greed ....BUT.....

THE big issue I have with new OS's is that they render software and utilities that worked superbly in previous OS's useless. Cut down "generic drivers" offer scant performance for expensive peripheries such as Scanners and Printers and usually render some excellent software that shipped with them unusable . I've had to bin two perfectly good photo quality scanners due to this and have found updates to OS's provide much poorer interfaces with printers. Most manufacturers are forced to develop new drivers for their kit but even Canon bailed out on an updated driver for the last scanner I was forced to bin.

As a technician I firmly believe in the creed of "If it ain't broke, Don't fix it", but sadly the so called progress of OS and hardware producers "breaks" more useful stuff than it fixes real problems? We got Win 8 & 10 to help M$ (fail to) sell their own cell phones! NOT to help the desktop PC user.

[EricF]

"Kaby Lake" is the current name for the latest generation of Intel CPUs. (They're based on the "Skylake" design from 2015.) Microsoft has stated that with the Kaby Lake series forward, they won't engineer support for Windows 8/8.1 or earlier on that platform. In other words, if you build a system around a Kaby Lake chip, it will need Windows 10 or whatever the current Windows OS is called. You may or may not be able to run Windows 7/8/8.1 on it, and Microsoft won't make any guarantees. The most controversial aspect is that Microsoft won't offer any operational (bugfix) or any security patches to the older operating systems under a Kaby Lake, AMD Ryzen, or newer CPU.

There's a business reason why they're doing this -- it costs money and development time to adapt the OS kernel and provide specific fixes and security updates across multiple CPU platforms. Microsoft has decided not to spend any of that on end-of-life operating systems for CPU architectures that are new and not specifically intended to run the older OSes anyway.

By way of more detail, all modern OS kernels -- Windows, MacOS(BSD-based), BSD UNIX and Linux are all compiled to support generational families of CPUs. It's because of the specific memory and processor bus architectures, as well as power management functions. Even in 2001, no one expected to make Windows XP run on a 486DX. But it would run on a Pentium, Celeron, or comparable AMD K5; those were all within the hardware generations that XP was designed for.

As all OS kernels have developed, they've laid down cutoffs where they won't run on older processors. Apple runs an approximately seven-year "window" of compatibility (although it's gone as long as ten years under OSX originally). Linux has been a bit more complicated, but there are still ongoing generational boundaries. Microsoft learned the hard way that continually revising and adding-on support in Windows XP because (primarily business) users didn't want to switch got expensive and consumed resources that should have been spent on modern development for modern processors. That then fostered more consumers to stick with XP because hey, it's well-developed and stable at work so it ought to be perfect for everybody. And so XP stayed, and stayed, and stayed beyond it's intended design life.

Microsoft doesn't want to repeat that. On the other hand, they've learned from the Linux and UNIX world the advantages of some of the finer points of kernel designs that are more modular and adaptable to updating to match CPU engineering advancements. So Windows 10 is, from an engineering standpoint, sort of a "line in the sand" where monolithic design stops and a more modular design of the kernel begins. Windows 10 (or whatever they decide to call it in another decade...) will always be recognizable to an OS developer as the same framework, but with different specific bits to a related family of CPU designs, but not necessarily all of them since its inception. The advantage is that as old CPU designs become obsolete and new ones become available, the OS can be upgraded to match with less engineering effort and cost since (like UNIX/Linux) the functionality doesn't change much, only the CPU-specific code has to. That's quite different from the Win9x vs NT families' differences, and even the evolution of Windows NT through 2000/XP/Vista/7/8/8.1. There have been some major monolithic architecture changes along the way for the NT family, all the while maintaining Win32 API support all the way in spite of it. It's been fantastic for end-users, but expensive for Microsoft.

So Microsoft's decision now is to break with the past in favor of moving ahead more economically while still not breaking Win32 compatibility. So you can keep using the same software as much as possible while continuing to upgrade the computer hardware it runs on. The hardware upgrades eventually demand OS upgrades; it's just makes good financial sense to focus on a reasonable spectrum of older hardware up to the newest as a development target.

Some people are complaining that they won't be able to keep installing Win7 on their Kaby Lake CPU-based computers because of the cutoff, as if they're being singled out. That's not the case. According to the new rules, my still-nearly-new Skylake CPU-based custom PC will eventually be unsupported by Windows 10. I'll have to upgrade the CPU and motherboard, at least, to keep going with updates. But the same has been true with my Macintoshes, and while I have a couple of very useful old laptops running Linux, they can't run the very latest distributions anymore since their CPUs are from 2004 or so. It's a development life-cycle process.

The nice thing about the old Linux machines is that, while they can't run the very latest major distributions with the newest kernel builds, they can still run some of the current-version software packages, or alternate versions compiled for older kernels but still with current features. Linux has a long history of allowing for near-infinite customization of the OS and the software it runs, which helps keep older hardware productive -- although at some point, security patching becomes a concern and old hardware still should be retired. Open-source development helps quite a bit for this in the Linux/UNIX world. That may be part of the reason why Microsoft has decided to take .Net and Visual Studio development into open source. This will provide more Linux/UNIX-like development flexibility for software while Windows moves to what's really an established OS development life-cycle that's been common outside the Windows world.

So, Microsoft has anticipated that people will want to buy a brand-new PC with a Kaby Lake CPU and have Windows 7 on it. And they're saying "No, don't do that -- don't put an OS from 2009 that's end-of-life in three years -- on a CPU from 2017. It wasn't designed for that and we won't support it. And by the way, don't expect our operating system to forever support aging CPUs that are functionally obsolete going forward either, because that's just silly and expensive." That's actually pretty reasonable. The problem is that they aren't saying it plainly that way. Actually, they're not saying much of anything plainly, and that's because the people in Marketing insist on editing everything the technical folks say, and they turn it into useless, incomprehensible marketing double-speak which gets all turned-around upon hearing it.

Short version: Microsoft is capable of making good decisions with reasonable justifications. They just need to banish their marketing people. And probably some of their lawyers. Or at least keep them muzzled and on leashes in public.