Running On A Tiny Tablet!

[mikeebb]

This is an experiment, but so far it works! Unbelievable! ORTS installed and runs in a $100 pink Win10 (32-bit, not 64) 2-in-1 (has detachable keyboard) tablet with a 10" display, Atom CPU (quad-core, 1.33 GHz but speed varies and can briefly reach 1.8), basic Intel HD graphics, 2GB RAM, and the install on a 32GB Micro-SD card. In the EB 1 activity of Bob Wirth's Chilicothe Div. (daytime, clear weather, default MSTS trainset), the train starts (with much slip), moves, speedometer very jerky, in debug F5 the frame time is maxed but the other performance measures are roughly consistent with each other. Stated frame rate is 8, but not really very jerky overall. I'd actually rate it as almost playable!

This is in stable 1.1.x OR using the non-LAA binary (it's a stock 32-bit system after all); haven't tried an update yet (downloading now). Given that (per Task Manager) it's redlining one processor core and using large percentages of the other 3, I probably can't run it for long without hitting thermal throttling, but still, it's fun. And the main menu does have an issue: it's a tad too big to fit the height of the 1280x800 screen. Works, but the bottom row of buttons is inaccessible. Again - this is a smoke test; picky details like reaching all of the buttons aren't really relevant.

This test demonstrates that you don't absolutely need a mongo gaming rig to run ORTS, at least at present; it *runs* though maybe not with fabulous numbers - I suspect I could optimize a few settings and get it up to 10 fps - in a computer that barely competes with a phone these days. The old ASUS i5 laptop does double the frame rate, with Win10 64-bit Home in 4GB RAM and older Intel graphics, and the mongo rig (Win10 64-bit Pro, Core2 Extreme, 8G RAM, nVidia graphics) will exceed the monitor refresh rate in nearly all cases. So the mongo rig does run a LOT better. But in this barely-a-computer rig ORTS actually runs!

BTW, 32-bit has an advantage if you're into ancient software. It has the NTDVM which allows running DOS and Windows 3.1 (16-bit) software. Train Dispatcher 2 works like a charm in the tablet. Can't run Railroad Tycoon, though; the keyboard has a non-standard layout outside of the basic typewriter keys, and no numeric pad (which is required to build anything).

Cheers!

[lateagain]

You need to get out more Mike!

Interesting test though in terms of spec needed etc. However apart from lack of Numeric keyboard the lack of any separate keyboard on most tablets would seem to be more of an issue? BTW it always baffles me when folk have breakdowns because there rig won't reach 60 f.p.s. Feature films used to be shot at 24 fps (TV in the UK films at 25fps).

BTW did the touchscreen allow use of the controls in the cab?

[EricF]

I tend to agree on framerate overkill, although with a couple of caveats. Frame rate is somewhat in indicator of control response -- that's because control/position refresh may be linked to the absolute screen bitmap, and faster refresh allows for finer control response. But even so, a frame rate of 30fps is usually fine for most computer games. Microsoft Flight Simulator X is best at 20fps and up (which is pretty forgiving, really) but if frame rates drop much below 20fps, control inputs can get laggy. First-person shooter games and 3rd-person view tactical games often need faster framerates to keep the player control/movement responsive and in-step with the display.

Computer monitors also have some characteristics that differ from film in how the image gets perceived by our eye. Basically, it's the difference between a scanned top-to-bottom image of a computer vs the flash-the-whole-image-at-once method of film projection. Faster frame rates hide the top-to-bottom scan from our eyes better. 60fps is good for that. 30fps can show visual artifacts to some eyes if you move the viewpoint rapidly, due to the scanning effect.

But for a train sim -- which doesn't need fast-twitch, ran-and-gun perspective -- high frame rates are nice for subjective smoothness, but generally anything above 18fps looks and works pretty good. That's why over in another conversation over at ET I had the opinion that as long as Open rails runs adequately on hardware from 2010 or so -- around a seven-year typical computer lifespan -- there's no need to worry about moving toward higher-end hardware specs. There are a lot of older, basic computers with more than adequate performance out there, particularly in the i-series Intel and corresponding AMD families that can run it and will continue to do so even if OR's ability to take advantage of fancy hardware advances. They won't be left behind very quickly. Advanced hardware capabilities are nifty if they're available, but dialing the settings up or down seems to allow OR to run reasonably well on a wide range of hardware, especially if you don't obsess over the highest possible framerates.

Cool to see how far you can get with very basic hardware!

[geepster775]

Well, if you are Dustin Hoffman in Rainman, and are in the midst of a cross country road trip, and you absolutely have to get your daily dose of Judge Wapner at 5 o'clock, then its great to know Open Rails, at present, can accommodate your needs from the passenger seat of the old Buick.

Just keep in the back of your mind that Open Rails is just now beginning its divorce from 'MSTS compatible'. And like all divorces, they are often traumatic and usually leave some of the lesser coping dependents crying.

Honestly, the last thing I care to see from Open Rails is slower development timelines to accommodate some lowest hardware denominator, or to be known as the last resort collector of individuals who were pushed out of other sims by various hardware bar raisings, and who are hobbled going forward in their ORTS experience by their collection of 'misfit toys'.

Like it or not, mongo is the future for many advancing train sim genres just like it is for many non-sim games. And the sooner we all plan properly, save appropriate funds, and get on it, the less painful it will be for all of us. Right now, there is enough MSTS compatibility in ORTS for those who cannot move forward, hardware-wise, due to finances to stay put where ORTS is now and at least experience a better simming environment than MSTS. But make no mistake, ORTS has totally fulfilled its obligations in the area of MSTS compatibility, and does not need to advance around that principle anymore.

[conductorchris]

I have a touchscreen laptop. Most cab controls work, although a lot don't.
It's a pretty cool interface when it works.
Christopher

[mikeebb]

I tend to agree on framerate overkill ... Cool to see how far you can get with very basic hardware!

It was raining, and I was vegging. I'm using the little attached USB keyboard (it's a 2-in-1 not just a plain tablet) not on-screen.

Then, one other thing happened. OR wanted to update, of course. I said go ahead. The update failed. And it wiped out all of the files (but not the folders) in the OR install folder. Including the selected MSTS content I had loaded for testing. Grumbles... Oh well, I hadn't actually expected to play it on the tablet anyway (have the laptop and desktop for that).

Then WINDOWS insisted on updating (Patch Tuesday, even though it's Windows 10), which I left running overnight. And for all those people complaining about Win10 restarting with no warning ... I have the opposite problem: even in non-active hours, it sits and waits after downloading for me to say 'Restart Now', so I had to do that when I got up.

Now let's see if I can get something useful done today...

[geepster775]

Touchscreen controls would be nifty if you were in the cab of a Gevo or ACe or a locomotive that actually had computer-screen equipped cabs, where you could touch-screen the in-cab computer console to get your F5 HUD info, or to work with DPUs or EOTs or cab audio. They are out of place in an SD40-2 or GP9 control stand, that's where a next-gen Raildriver would be more appropriate.

[lateagain]

It was raining, and I was vegging.
Now let's see if I can get something useful done today...

LOL! well trial and error all adds to the human bank of knowledge Mike. Bizarrely warm and sunny day here today for December so the dog got a longer walk.

I quite agree with other posters that most PC's have little more than a 7 year lifespan and certainly no development is going to "push the envelope" if you try and make it work on outdated kit. Simulators always run better on better kit but they haven't needed cutting edge specs like 1st person shooters etc.

As for MSTS divorce? Not how I see it at all. In the real world good sound designs in aircraft AND in locomotives have been "re-engineered". Better control systems, better more fuel efficient engines etc. If you have a solid design develop and tweak it because the ones that re-invent the wheel often bring a whole load of new glitches to be worked out. A Reality check means that whilst we have pretty decent "current state of the art" rolling stock no one is going to rebuild all that lot to higher specs anytime soon. Just as with MSTS those interested will develop new techniques, improve upon skin detail and weathering, sort out physics so that the stuff uses OR's coding etc.etc. What's needed is evolution not revolution IMO. BTW that's before we even look at routes!

[geepster775]

Why yes, Goeff, mommy and daddy are going to stay together 4ever! Keep telling yourself that from the safety of your bedroom.

What good is refining aircraft and locomotive designs when the Z-train is going to be made obsolete by home package delivery drones? That's the level these other sims are reaching for. And the dysfunctional 'dragging' marriage of MSTS and ORTS is still worried about refining trusted locomotive designs. The others have 'gone drone', so ORTS had better cut that boat anchor from its leg chain pretty soon. lol

[mikeebb]

The future of OR doesn't have much to do with the original subject of the thread, except that what I discovered is a point-in-time and shouldn't be relied on for future reference. It was done for fun, not profit! Absolutely, hardware requirements will change (and increase) as time goes by.

IMO Open Rails seems to be following the Windows (and IBM mainframe!) path: improve but maintain backward compatibility. Yes, at some point, the improvements will become essentially a new system (arguably, is at that point in some ways), and the Good Stuff will be built for that. But a defining feature of OR, compared to most other train sims, is that the large library of stuff from MSTS still works more or less correctly. That makes a lot of models that people are no longer interested in working on still available and usable.

The Other Train Simulators, for the most part, follow the Apple model: regular (and frequent) updates that break what you already have, so you have to buy a whole new "set" every update or few. As commercial ventures, that's expected, and it's great for the purchasing "ecosystem" and results (if done right, respecting 3rd-party developer needs) in fairly rapid improvements, but it tends to limit asset development to a small group of people who do it as a business. And it's not so great for somebody who doesn't have the budget for constantly re-purchasing things.

On the whole, I prefer OR's current approach. Though I could see a split at some point into, effectively, a "legacy" thread that gets mostly just maintenance, and a "current" thread that receives all the latest updates. Essentially, a DOSBox for running MSTS stuff, and the mainline OR going forward. And btw I don't agree that OR is fully MSTS compatible yet - there are still some things to figure out - but it's more than good enough.

Oh yes, and it would be very nice if it could somehow be made to run on Linux (I've tried it in Wine and so far no joy, even with the MS dependencies covered). OR and ACD Canvas (illustration and GIS software) are the only things I have right now that just don't work in Linux.

[geepster775]

I do believe this "last version of Windows" is where Microsoft also jumps off the backwards compatibility bandwagon, too. Apple has such a head start with the new, restricted business model, who knows if Microsoft can even make up the ground, financially, given its lag.

OR will never be *fully* compatible with MSTS. Nor does it really really really need to be. It is as close as its going to get, right now. Consider it advanced, but 95% compatible. It is done with meeting its compatibility obligations, and need not worry about that anymore.

Multiple development threads is too much distraction for an all-volunteer effort. More reasonable to expect MSTS compatibles to simply stay below a certain version of OR, whatever that version may end up being, with the rest who are capable of leaving the .S file behind moving on to the newer OR versions.

And while I agree that the freeware library has made MSTS what it was, I have said before its simply not going to play as integral of a part in OR's future. I truly expect OR to morph into more of a smaller system primarily relying upon a handful of third party developers, too. Sure there will be more freeware-supplied options than the other sims, but nothing will ever touch the popularity of freeware libraries under MSTS again. The "space" is just too crowded and fragmented with options now, many of them better, albeit closed, options if you have the hardware. Nels is still waiting on 'enough' OR-specific library content to warrant the creation of OR library categories, and that 'enough' is never coming.

10-4 on the Linux comments.

[lateagain]

Why yes, Goeff, mommy and daddy are going to stay together 4ever! Keep telling yourself that from the safety of your bedroom.

Well actually my "mommy and daddy" did stay together until they died over 40 years ago now. I'm sorry that you SEEM to have some issues with divorce?

What good is refining aircraft and locomotive designs when the Z-train is going to be made obsolete by home package delivery drones?

If you'd read my post properly you'd have understood that my reference was to REAL WORLD practice with REAL aircraft and locomotives ....a matter of historical record.

That's the level these other sims are reaching for. And the dysfunctional 'dragging' marriage of MSTS and ORTS is still worried about refining trusted locomotive designs. The others have 'gone drone', so ORTS had better cut that boat anchor from its leg chain pretty soon. lol

An interesting analogy being as Drone Delivery is still a stupid publicity stunt by Amazon?!

Frankly I fail to see why ANYONE would see MSTS compatibility at some level or other is somehow an "Albatross around the neck of OR development"

IF OR developers deliver ALL that MSTS achieved, and BTW that other newer Sims are STILL to achieve, then at that time such compatibility will become less and less relevant.

Quite why this seems to make you SO bitter and twisted is frankly beyond me. These are HOBBY Sims? NOT life and death. I've a thick skin but you might want to reconsider the first part of your post as there may be youngsters on this forum that your comments might have upset.

This was about Tablets ....so I'm done on the topic.

[geepster775]

This was about Tablets

Yes, indeed, until a couple of you were found trying to build a bonfire to the merits of running simulators on crappy hardware. Thank heavens somebody had the common sense to come in and stomp that infant bonfire out.

Frankly I fail to see why ANYONE would see MSTS compatibility at some level or other is somehow an Albatross

Think of it more as an analysis of what is left to be gained from the MSTS-ORTS relationship. New recruits among the 5% who still run MSTS exclusively? Odds are low that they patronize the vendors or provide technical help to the community so they don't represent all that great of shot in the arm to the "open rails economy" anyway. We already have one vendor who will not provide club access unless you contribute adequately to the economy beforehand. The wave of the future?

[mikeebb]

Re. the tablet: having proven that it works (poorly) and having a failed OR update wipe out the installation, I didn't bother trying to reinstall. Took too much space for how (not) well it worked (moving from proof of concept to is it practical), and I have adequate alternatives in the laptop and desktop. So that's over with.

Re. MSTS compatibility and where this is going: OR is not a piece of commercial software, so there's little incentive to drive potential users away by messing with MSTS compatibility while moving the quality of the sim forward. Yes, as time goes on, some things will not be "what MSTS was" any more (timetable mode is already well beyond traditional MSTS activities, though it works on MSTS routes with MSTS trains). And a few trains are appearing that have so much detail that they won't work in MSTS any more. So far, not many, but some. That'll continue, and note that so far at least using a OR-only locomotive doesn't prevent you from hauling MSTS cars behind it, on a MSTS-based route; the mix-and-match is a good thing.

One thing I *have* seen is a fair number of obvious newbies (or trolls, in some cases?) in the fora who approach OR and MSTS as modern commercial programs where everything is done for you, then drive the old heads crazy with their questions which make it obvious that they don't want to fiddle with things to make them work (as most of us are used to). Many of them, IMO, would be better off with one of the commercial simulators that they can just run as a game, but they may not want to spend the money. My experience is that you pay for software one way or another: money, your own time, or sometimes both. Others might be convinced to figure out how things work, and in the process learn how to make some contribution to the hobby. ORTS' relationship to MSTS might be causing a problem for them now due to the need to fiddle with things, but that will improve and some utilities for managing and fixing the olde stuff might be updated to make it easier for those who need it to "just work." In all, though, I see no need (yet, at least) to draw a line in the sand and say "this far, for MSTS compatibility, but no farther."

[lateagain]

Mike I'm with you 100% on MSTS compatibility and I think you've hit the nail on the head with some of the newcomers (or should that be latecomers) to MSTS.

There's always been a section of society as a whole who are obsessed with "The New". Usually, but not exclusively, these are folk who are all too ready to ignore what they have around them because their self proclaimed expertise lies only in being the FIRST to loudly regurgitate all the hype around "The latest" which is of course only the hype of the marketing sector . As you get older you forget so much more silly hype and failed to deliver projects than the real benefits that actually come to fruition ....but those obsessed with THE NEW will never listen.

The commercial world in any sector, including games and simulators relies on these individuals to spread their marketing message. Look at ANY forum and the Fanbois post uncritical praise upon praise of the new and rabidly attack anyone who dares to post a considered or informed or (even worse) experienced view that doesn't support their sycophantic gushing's.

That, like it or not, will always be with us ....but what really saddens me is that their often spoon fed mentality means that they miss a whole area of Simulation that has made it such an enduring hobby. They miss the creativity and discussion about how to move things forward. Their blinkered attitude to critical discussion shows that they don't understand that nothing moves forward without critical analysis of where we're at and how we got there? Nothing, ever was improved by sycophantic praise . Stuff develops because folk say "Well that's OK as far as it goes ....but wouldn't it be better if it did X Y and Z too?

I studied design at College nearly 5 decades ago now. You learn to improve your work by constant review of it by your fellow students and tutors. Often harsh criticism, but often constructive criticism. However most projects involve starting with where we're at and trying to improve or move that forward. Few, if any, involve starting at square one.