A general multiplayer design discussion

[halbert5150]

So I dont want to be critical of anyone who has taken the time to create any of the current mutiplayer routes / sessions out there. These things take alot of time to create and Im always thankful to those who put forth the effort. I do have a couple of criteques that I wondered if anyone would care to comment on.

First of all I don't see the point of creating a multiplayer route that is huge. Ala UMR or MRL. Once again I am not critical of those who took the time to create these mutipleyer sessions but I thought the point was to interact with other drivers. Most of the popular multiplayer routes are so huge you will never pass by another player or have to coordinate a meet or anything like that. You might as well be playing by yourself and just chatting on teamspeak with each other. Only if the session fills up with the maximum amount of players will this ever happen, and lets face it that just doesn't seem to take place.

Secondly, I would like to see a tighter focus with the theme of the layouts and rolling stock. If the route has modern locomotives then the rolling stock needs to be modern as well; not woodside boxcars with roofwalks. I suspect that rolling stock choices are made by whats already on the download station and not by what the creators would prefer generally. If you create 50 new rolling stock assets then thats another 50 items that will have to be downloaded.

Lastly, and this is something that auran would have to address, but I wish there was a way for me to be able to drive my payware locomotives in a multiplayer session. I don't know exactly how this could be accomplished, obviously it would take a good deal of reworking of the code for multiplayer, but I think it would be cool. I dont think the 3rd party payware creators would mind either, its free advertising.

So what do you think? Please don't make this a grip session about existing multiplayer routes. Thats not my goal, but rather maybe we can come up with some ideas to make the next few a little better.

 

[jfriedrich]

I whole heartily agree about being able to use our payware purchases in multiplayer sessions.

 

[nicky9499]

There are so many banes and restrictions in multiplayer that I'm surprised anyone even plays it at all.

1. Only same versions can play with each other.
2. Only same builds can play with each other. (Read: Pre-SP1 and SP1 divide among users)
3. Only DLS content allowed in route.
4. Route must also be on DLS.
5. Only DLS rolling stock and locomotives allowed.

Can whoever thought up of these rules please enlighten me as to the reasons behind them please? There are so many way to do multiplayer and this is one of the worst I've seen (nothing surprising though, this being Trainz).

Take flight simulation for example. Doesn't matter whether or not you have FSX + Acceleration. Doesn't even matter if you're FS9 or P3D. All of them connect using a common method and basic functionality is maintained across all players. Pilot A has a payware aircraft that Pilot B doesn't? No problem! The MP client matches it to a stock or AI model with a similar livery. Both of them are flying to KLAX but only Pilot B has payware scenery? No problem! Pilot A's sim loads a built-in scenery.

Can you imagine if flight sim was done N3V-style? Nobody would be able to fly their $90 PMDG 777s online - only crap reskins from freeware file libraries are allowed. FS9, FSX, P3D pilots cannot interact with each other. Heck, those running different builds of P3D also cannot fly together.

---

Back to Trainz. Again, I'm no programmer, I'm just looking at this whole thing from a KISS perspective.

I'm not saying TRS06 users should be able to play with TS12HF4 peeps or whatever. TS10 (afaik) has some MP functionality. TS12 had it right from that start -> this means all versions of TS12 should be able to connect to each other and achieve basic functionality. So what if player A is driving a JR BNSF Dash9 and player B doesn't have it? The program should look for similar content locally (eg. Chris' reworked Sporbust Dash9s), failing which a generic model (Auran SD40) is loaded in its place. And what if player B does indeed have a JR BNSF Dash9 as well? The game loads exactly that. Why do we need the "DLS-only" rule?

Routes. You can keep your current stinking system, but please allow a second "open" category of MP routes where the route builder takes responsibility for dependencies. For example, I go to great lengths to make sure that my route is released with all dependencies accounted for, either on or off DLS. Every end-user will end up with the same content set required to run the route. Why then are they not allowed to connect to each other?

Even if a user is lacking some scenery items like a shed or some grass texture, like above the program would either look for a similar replacement or ignore it completely. Does a missing shed impact operational gameplay in any meaningful way? The only thing an MP session should enforce is the track network. If a player is missing an asset that changes the track network such as an interactive industry, trackside object (triggers/markers) or signals, he or she is not allowed to connect. The onus is on the route builder to ensure that his users get all dependencies required by his route and for the user to follow all instructions/click all required download links. So why do we need the "DLS-only" rule in this case too?

In it's current state I am not going anywhere near MP because of how poorly it's been implemented. I'm sure I'm not the only one.

 

[jhuravitch]

Lastly, and this is something that auran would have to address, but I wish there was a way for me to be able to drive my payware locomotives in a multiplayer session. I don't know exactly how this could be accomplished, obviously it would take a good deal of reworking of the code for multiplayer, but I think it would be cool. I dont think the 3rd party payware creators would mind either, its free advertising.

It's too bad that Trainz doesn't utilize a similar system as Run8, having high quality payware available but also a free version for everyone. Those who buy the payware see them in game, while those who don't still see a train in game, only the free version of the rolling stock appears in place of it.

 

[JIb228]

I agree with every thing.

This is largely why I have made the switch to Run8.

Secondly, I would like to see a tighter focus with the theme of the layouts and rolling stock. If the route has modern locomotives then the rolling stock needs to be modern as well; not woodside boxcars with roofwalks. I suspect that rolling stock choices are made by whats already on the download station and not by what the creators would prefer generally. If you create 50 new rolling stock assets then thats another 50 items that will have to be downloaded.

Take a look at the Fox Mountain & Little River Railroad's route and session.

Here's the topic on that: http://forums.auran.com/trainz/show...-Mountain-and-Little-River-Railroad-s-(MPish)

It had a common theme, and was even a smaller route with only about 20 miles split between the 2 railroads. One of the big problems though was the ghost trains. Another thing was, and this was reported to me from crews off of UMR, that if you got more the about 10 people in the session it would become extremely laggy. FMR&LRR was built really for only about 6 people in mind.

I have also learned that even when MP is working, both the servers and the routes/sessions, that only a very small hand full of people even join. We're talking I would announce weeks in advance that I was going to host a MRL session all weekend long. And only 3 people would show up over 3 days. Even UMR only ever had the same 5 people it seamed to me.

When it comes to having fun, building what I want to build and running what I want. I play Trainz. At this point, until MP is improved for Trainz, I'd say bite the bullet and buy Run8. It is worth it. I know a lot of people say "Oh its just MSTS." Its not. The locomotives, and rolling stock, are all most on the same level as JR content. The MP system is rock solid, and there are no issues with content or 3rd party servers. (Im looking at you N3V MP severs and you habit of shutting down on the weekends.) Yes its lacking some of the tools and features that Trainz has. Yes its only, at the time of posting this, 2 routes. Which can be merged into a much larger route. Which doing this you would think would make it even harder to run into people in a session, but its not. On average larger sessions can have up words to 40 people in one session, on both routes.


Ive ranted before in why Train MP is flawed, and doomed if not improved. I have some hope for T:ANE.

One of my older rants on the issues with MP: http://forums.auran.com/trainz/showthread.php?107261-Multiplayer-Consensus

 

[chris2001trainz2010]

I agree with ALL of this thread 1000000%

 

[halbert5150]

So have there been smaller multiplayer routes designed for 3 - 4 people created and do they work better?

Auran was always very reluctant to add multiplayer in the first place. Back in the days when I was part of TPR thats the first thing we requested whenever we had anyone's ear. I don't think they ever thought it was important and there hearts were never invested in the idea. I was part of the beta testing for multiplayer and the bug report page had everything listed that is currently still giving us problems, but they released it as it was anyway. I was shocked. The trainz new area also has nothing on it about a major focus to update or a fix multiplayer. So I think we are stuck with what we have. Thats why im trying to find out the best way to make it runnable.

I own every rail simulator with the exception of run8 and its not eye candy graphics that keep me coming back. Its surveyor and the ease at which I can create my own content and put it into the game. I think im not alone in that, but in addition all I have ever wished for since the first day I ran UTC was the ability to run trains on the same layout with my friends.


Hal

 

[JIb228]

So have there been smaller multiplayer routes designed for 3 - 4 people created and do they work better?

Yes and no.

Yes in that you get to interact with people more often, and you see them. FMR&LRR was a lot of few the times I ran it. You get one guy in Durand yard sorting cars. Some one working Hammansport. And a local heading back to Durand. It gets fun.

But the ghost trains, invisible trains*, and other issues just killed it.

*Some times a car could be seen by players A and B, but player C would not see it. Player C loads up thier own session and can see the car. Players A and B join player C. Now Player B cant see the same car.

If you really want to run with friends, Run8. If you want the ease of which Trainz gives us to make our own content, stick with Trainz.

 

[Trackman53]

You can not compare FS multiplayer and Trainz multiplayer. FS has had MP built-in for several versions. Trainz for only one.

And there a big difference between a plane that can be any where within 50' of its true location and a train that must be placed within inches or less.

And I seem to remember pilots complaining about aircraft warping all over the sky.

Best thing N3V can do is just drop MP all together. Not that many played as stated above.

 

[halbert5150]

"Best thing N3V can do is just drop MP all together. Not that many played as stated above."

Trackman I couldn't disagree with you more. I think people don't play multiplayer because it doesn't work right; not the other way round.

Ill get run8 and ill spend time with it. Also I just discovered openrails this evening and openrails has a nice multiplayer in it as well. Just spent 40 minutes playing that and it works nice. So far openrails is only compatable with msts content but I guess Ill be learning how to create that type of content as well.

 

[ZecMurphy]

1. Only same versions can play with each other.
2. Only same builds can play with each other. (Read: Pre-SP1 and SP1 divide among users)

This isn't quite true. If you are running the same trainz-build number (as of TS2009, this is showing as the Content Manager 'Version' number ), then you should be able to play multiplayer across multiple releases of the same version of Trainz (TS12 Steam vs TS12 from other stores). TSMac 1 was unfortunately too far different to work cross platform, and the MP was included as an 'extra'. TS:Mac 2 is a newer trainz-build, and hence is currently on it's own.

3. Only DLS content allowed in route.
4. Route must also be on DLS.
5. Only DLS rolling stock and locomotives allowed.

This has been explained so many times it's not funny. First off, these three that you've outlined aren't quite correct. All content for the multiplayer session (including the session itself, the route, and all dependencies of all items used by both) must be built-in, included in a DLC pack (essentially seen as built-in by Trainz), or downloaded and remain unmodified from the DLS.

All of these rules apply for the same reason. Ensuring the content is entirely unmodified/exactly the same. How would you like to fire up a multiplayer game, only to find Player 1 has modified the map and removed a length of track in the middle of the mainline. All trains now derail over that length of track, irrespective of the track being visible for you.

Or how about we put a different enginespec on a locomotive? Suddenly the same train in player A's game accelerates and decelerates at a different rate to that train in player B's game. After a short time, the trains could be several miles apart in the two games...

Let us use Mojave as an example there. You've got two trains preparing to pass at one of the sidings/crossing-loops. Player A is sitting in the siding waiting for player B. On Player A's game, the train has just passed him, and he departs. But in Player B's game, he's still a mile away. Player A departs, and runs head on into... The train that passed him!

Keep in mind that comparing to a flight sim is not overly appropriate. There are some pretty big differences in what they can do compared to a trainsim. For example, physics are less of an issue for the 'AI' (other players) aircraft. Except at the airport itself, you're unlikely to have to worry about them being in your way, to the same degree that you need to worry about another player's train on a single track line

You also don't need to worry about things like vehicle lengths, acceleration and stopping distances, etc, because aircraft aren't exactly parts of consists

And before you mention systems that 'share' the content between machines. Imagine having to copy the entire session and route (including all dependencies) every time you fire up the map, so as to ensure that everything is going. Some sessions/routes, including all locos and rolling stock, could go well above 500MB... And that's not counting the issues we'd have with payware creators not being happy about other people suddenly receiving copies of their content

My point is, there is a BIG difference between MP for a rail sim and a flight sim. Simply to ensure it will always work for everyone (in terms of gameplay), we have to take these steps. The options are there for creators to allow their content to be used in MP sessions. They can release on the DLS, or sell their content via our DLC system (I note that a large portion of JR's payware is now available via the DLC system for anyone who wishes to use it in MP...).

Keep in mind that other rail sims with MP so far impose much the same rules. Either you can only use official DLC's/built-in content in MP (where they could then easily provide semi-accurate placeholders, particularly in terms of physics, each time a DLC is released...), or you have to follow essentially the same rules as Trainz...

I'm not saying TRS06 users should be able to play with TS12HF4 peeps or whatever. TS10 (afaik) has some MP functionality. TS12 had it right from that start -> this means all versions of TS12 should be able to connect to each other and achieve basic functionality.

This would certainly be nice, but we can't ensure both players will be seeing the same thing. There are progressive improvements to physics in various updates to Trainz. We've also changed the way that some content is handled, particularly DLC's. Some of the built-in content also had changes made, which can change the way that Trainz works. All of these can build up to both players seeing entirely different things in the one game...

So what if player A is driving a JR BNSF Dash9 and player B doesn't have it? The program should look for similar content locally (eg. Chris' reworked Sporbust Dash9s), failing which a generic model (Auran SD40) is loaded in its place.

Again, see my example on Mojave...

And what if player B does indeed have a JR BNSF Dash9 as well? The game loads exactly that. Why do we need the "DLS-only" rule?

Can you categorically tell me that both players are using the EXACT same asset. No modifications. No changes at all. In particular, the mass tag and the enginespec. Is player 3 when he joins going to have that exact same unmodified content? How do we know those three are using unmodified content? Well, the only way is to have an example asset (DLS) to compare to, or locked content (built-in/DLC).

Routes. You can keep your current stinking system, but please allow a second "open" category of MP routes where the route builder takes responsibility for dependencies. For example, I go to great lengths to make sure that my route is released with all dependencies accounted for, either on or off DLS. Every end-user will end up with the same content set required to run the route. Why then are they not allowed to connect to each other?

Again, can you assure me all of that content is unmodified in ever player's copy of Trainz? Can you assure me all players are playing unmodified maps?

As I can tell you now. ANY issue with ANY content used in MP WILL be blamed on us. We're first stop when people want to lay blame. We've stated many times that implementing an 'advanced users' option is really no option for us. We simply can't support it, because it becomes to unwieldy too quickly. When you have those issues I've outlined above, and have to say it's because you've turned on XX, we get abused. Heck, this happened with compatibility mode! Turning it on degraded performance, so we copped flack and abuse because an option in our software caused a large performance hit.

Even if

a user is lacking some scenery items like a shed or some grass texture, like above the program would either look for a similar replacement or ignore it completely. Does a missing shed impact operational gameplay in any meaningful way? The only thing an MP session should enforce is the track network. If a player is missing an asset that changes the track network such as an interactive industry, trackside object (triggers/markers) or signals, he or she is not allowed to connect.

This is all well and good until you get some objects that do interact with the gameplay that can be missing. Examples such as scripted scenery items that are interacting with scripted objects on the track. Rather than trying to single out specific items, we make it a VERY simply rule. Everything must be DLS, DLC, or built-in. Easy to remember what content needs to follow it (everything)...

The onus is on the route builder to ensure that his users get all dependencies required by his route and for the user to follow all instructions/click all required download links.

To a degree, yes. But more so, the onus is on us as the software developer... Simply because we are the first stop for complaints about things not working. In part, this is due to people not wanting to offend the creator of the route... More so we're just a very obvious place to lay blame for issues caused by not ensuring things work properly, in this case for gameplay.

Regards


EDIT: Corrected information about 'cross version' multiplayer. My apologies for any confusion, I was thinking of TS12 'standard' edition vs TS12 'Steam' edition.

 

[nicky9499]

Thank you Zec for your in-depth reply and for shedding some light on this dark corner of Trainz.

This isn't quite true. If you are running the same trainz-build number (as of TS2009, this is showing as the Content Manager 'Version' number), then you should be able to play multiplayer across multiple releases of the same version of Trainz

Again, I was specifically referring to different trainz-build numbers (and versions, to a lesser extent), for which I observe there is quite a divide among users ever since SP1 came out.

How would you like to fire up a multiplayer game, only to find Player 1 has modified the map and removed a length of track in the middle of the mainline. All trains now derail over that length of track, irrespective of the track being visible for you.

See my point regarding track network integrity.

Or how about we put a different enginespec on a locomotive?

Let's step back, look at it from a KISS perspective and question the entire approach to online train movements. Why do we need the enginespec in the first place? If I'm in an online session, why does my computer need to worry about the physics of other players' trains? All that is needed for basic functionality is the roughly correct-looking train in a roughly correct location. The program needs to sync movements, not control commands and letting the asset's enginespec on each client decide how each train behaves.

For example, physics are less of an issue for the 'AI' (other players) aircraft. Except at the airport itself, you're unlikely to have to worry about them being in your way, to the same degree that you need to worry about another player's train on a single track line

acceleration and stopping distances, etc,

Again, I would reiterate my point above. Why do we both about the physics of other players' trains in the first place? In fact, when you compare railroads vs aircraft, the latter has much more data to sync because trains only have ONE degree of freedom and that is along the z axis whereas aircraft have SIX.

Imagine having to copy the entire session and route (including all dependencies)...issues we'd have with payware

That was never part of my suggestion. An asset being (non DRMed) payware doesn't make a difference if both parties have it. Each client uses whatever it can find in the local installation and failing which, looks for a similar substitute.

They can release on the DLS, or sell their content via our DLC system

Be that as it may, but right now I think many will agree that a lot of the best content out there are not found on the DLS. Creators choose to not distribute their creations via N3V's depositories either due to existing conditions or policies. This in itself is a completely different discussion for another time but my point is there's a lot of good stuff out there and route builders have to chose between quality or MP-compatibility.

Two examples off the top of my head; Felix's recent Emily Bay Model RR and Andy's Checkrail routes. They represent what is arguably the pinnacle of routes in their respective Trainz generations with long lists of 3rd party downloads to match.

This would certainly be nice, but we can't ensure both players will be seeing the same thing. There are progressive improvements to physics in various updates to Trainz.

In particular, the mass tag and the enginespec.

They will see the same thing if you sync the same things. If train A is at this particular spot you sync its length and position and that will be it. No need for "physics" and "enginespecs" and "weight" or "stopping distance" worries.

Is player 3 when he joins going to have that exact same unmodified content? How do we know those three are using unmodified content? Well, the only way is to have an example asset (DLS) to compare to, or locked content (built-in/DLC).

This is a classic case of thinking too much. Let's say player 1 is driving said JR BNSF Dash9. Player 3 also has this train installed but he's swapped out the horns and reskinned it to ATSF. Does that matter to player 1? No! Player 1's program will still display the JR BNSF Dash9. If player 3 blasts the horn, player 1 will hear the horn installed on his Dash9, not whatever horn player 3 swapped in.

Again, can you assure me all of that content is unmodified in ever player's copy of Trainz? Can you assure me all players are playing unmodified maps?

This is all well and good until you get some objects that do interact with the gameplay that can be missing. Examples such as scripted scenery items that are interacting with scripted objects on the track.

Do you mean stuff like Boat's ATLS system?

The only aspect of MP that has to be strictly locked down is the route's track network and associated assets. This probably includes said scripted assets. Anyone whose copy of the route may compromise the session is not allowed to connect until he does a "Revert" or reinstalls the route.

implementing an 'advanced users' option is really no option for us. We simply can't support it, because it becomes to unwieldy too quickly.

we are the first stop for complaints about things not working.

Then don't. I understand your point of view as support staff who gets the brunt of user feedback simply because you're on the frontline. Simply then, don't provide support for "advanced" multiplayer - and make the disclaimer easily visible. Any route builder that releases an Advanced MP route and players who participate in it know what they're getting themselves into. Put a "terms and conditions-ish" box before any Advanced MP session if that's what's required. SCS Software used to have big text on the main menu of their games if they detected a mod that read "No support is available for modded games." (If you posted to their forums with an issue they'd still gladly help though but that's besides the point.)

This of course doesn't include obvious backend issues like "it's the weekend and the MP server is down again".

---


I have zero intent of getting into an argument or raise any grievances, I am simply suggesting how I see multiplayer can be better implemented so that more people will play it - all of which is for the better. Right now we have this half-cooked feature that has great promise but ultimately fails to deliver and in turn makes players avoid it altogether, me included.

I'm not saying Trainz must outright copy exactly how Flightsim does MP, only that the concept behind FS MP is much more ideal and flexible and could work in a trainsim environment as well.

Cheerio,
Nicholas

 

[JamesMoody]

Again, I was specifically referring to different trainz-build numbers (and versions, to a lesser extent), for which I observe there is quite a divide among users ever since SP1 came out.

Compatibility between different versions breaks because the data formats have changed somewhere.

We build in backwards compatibility for reading, but we can't build in forwards compatibility for writing as that would require perfect precognition - and if we had that, I'm sure we could put it to a lot better use than making a railway simulator .

We only ever change the interfaces if we are forced to (to fix a bug, or to add a new feature that then requires a new bit of data) - that's why you get a whole string of similar builds (hotfixes, localisations, and builds for different platforms) which DO work together, before a breaking change.

For example, one reason that TS12 SP1 and TS:Mac2 can't multiplayer together is because the location for "bell state" changed to be at the vehicle level (not at the train level). This changes the serialisation for both the train and the vehicle. TS:Mac2 can understand what TS12SP1 is on about, and will convert. But TS12SP1 cannot cope with the data from TS:Mac2.

For this kind of change (which requires a change in interface), our only alternative would be to decide to fix the interface in stone, never to be changed - which means issues like the bell would never be fixed. I'm sure that would be an unpopular decision.

The program needs to sync movements, not control commands and letting the asset's enginespec on each client decide how each train behaves.

Trainz does sync the current position.

But if this was all it did, you wouldn't be able to cope with hosting a worldwide game. In fact, it would still look jerky and juddery with everyone on a single LAN segment.

However, we also sync the current speed, the current acceleration, and the control inputs to predict how the acceleration will change over time. With all this data we can then do a full physics simulation to predict train movement - which as it's the same code calculating it at both ends, produces a very accurate result. That accuracy enables us to sync relatively infrequently so we don't need large amounts of bandwidth, and because the predictions are good, cope with high latency - which enables everyone to play together in a worldwide internet game.

Cut just one piece of data out, and we can't use the existing physics engine to predict movement. If we can't do that, we have to write a new one. That will not only cost us development time which could be better used making improvements elsewhere, but we'll need to parallel maintain it as well. And it will never be as good, so it will always result in more jerky movement.

Everything is a tradeoff.

Again, I would reiterate my point above. Why do we both about the physics of other players' trains in the first place? In fact, when you compare railroads vs aircraft, the latter has much more data to sync because trains only have ONE degree of freedom and that is along the z axis whereas aircraft have SIX.

Flightsim was played "multiplayer" for years with a bunch of disconnected PCs and CB Radio handsets. Actually seeing other planes in engine in Flightsim is of so marginal a benefit that it can be handwaved entirely, and the computers don't even need to be connected.

Given that kind of background, if Flightsim gets the position of an aircraft wrong by even several hundred metres during flight, it's probably not going to be noticed. You can also completely ignore planes you can't see - you don't have to know anything about them at all.

That doesn't work in a train-sim environment. That difference is the difference between a well planned meet at a loop and a massive derailment. So we get the position right, and do it without introducing geographic limitations - which means relying on more data. And we need it for every train in the world too - otherwise the signals would be all wrong.

An asset being (non DRMed) payware doesn't make a difference if both parties have it. Each client uses whatever it can find in the local installation and failing which, looks for a similar substitute.

How do we determine if both parties have the same asset? How do we determine if it's been changed?
How do we determine what is a "suitable fallback"?

If it's modified (and if it didn't come through our services, then we have to assume it is), then we have absolutely no way of telling locally if our copy of this asset is the same as someone elses. Some games take the viewpoint that every single modification should be copied from the server every time you join a game. Good luck getting in game this side of Tuesday with the amount of content our creators put into their maps and sessions. Also, that's not particularly nice for payware creators, because it will then share that payware to everyone who connects.

I think (...) a lot of the best content out there are not found on the DLS.

And that's a problem in itself. But there are so many other benefits from being on the DLS that we should attempt to fix it directly, rather than work around it just for multiplayer.

The only aspect of MP that has to be strictly locked down is the route's track network and associated assets.

That also means the elevation of the ground (because that can affect the track), and anything that can create a hole in the baseboard (because that can affect baseboard elevations). And any scenery object with attached track, including industries, level crossings, fixed track objects, bridges, tunnels, and so on. But not all such objects actually contain track. You've also sucked in things like the YARN road junctions, most catenary, and so on. And any dependency of any of those asset types too.

Then as I've pointed out above, we need train performance data to get smooth operation in internet wide games. That means every enginespec unmodified, every vehicle using it's original enginespec, and every vehicle and product weight unmodified. Also the product compatibility as the session creator intended, too.

It also needs to be everything with a script.

At this point, the list of "what's important to have unmodified" comes out very small. We're firmly in "... some of the sheep" territory here. Is it really worth computing the list of what's really necessary just to allow someone to change the billboards on a random building?

Anyone whose copy of the route may compromise the session is not allowed to connect until he does a "Revert" or reinstalls the route.

We have pretty much taken exactly that option. We can revert, or if you've only ever had a modified version, we can re-download.

I am simply suggesting how I see multiplayer can be better implemented so that more people will play it - all of which is for the better.

Good idea - but wrong direction. You'd make more headway by putting your efforts towards persuading the people who have made what you consider the best freeware to upload it to the DLS. Doesn't have to be DLS exclusive, just uploaded to make sure there's an authentic copy that Trainz knows about, and can trust. Then it's good to use in Multiplayer.

For local modifications, create a clone asset, rather than editing an original. Then you can enjoy the clone with your tweaked enginespec or whatever, and still run the original for multiplayer compatibility.

Or better yet, if you are a good enginespec tweaker, offer your services to loco creators - and help them improve their own enginespecs. Then they can be uploaded as official DLS updates to those locos, and everyone benefits from the improvements.

 

[halbert5150]

I didn't know I was bringing up such a sore subject. For me I just wanted to know with the system the way it is is there a better way of doing things. Thanks to your comments (especially from JIb228 thank you), I think its not going to be worth the time to even try. Man that suckssss....... big sigh...... go do something else.


I dont have run8 yet but ill get it and do some multiplayer. As I've stated that's something i'm real interested in. This does mean that Ill be spending less time with trainz and more with a different program. In the end that's for the folks at Auran to decide if that's what they want to happen or not. Alot of the stuff we have talked about in this thread they are the only ones that can do anything about it not us.

Thanks for your input,
Hal

 

[JIb228]

The two biggest issues to me that make MP a pain are the reliability of the servers, and ghost trains.

First off, you have to connect to the MP servers, then you can connect to the session. If the servers are down then no MP.

And of course the aforementioned ghost trains.

Both are issues you dont get in Run8. As I said, I keep build in Trainz, and keep up with my iPortaling. And maybe where these issues are fixed I'll go back to working more on MP routes/sessions. But for now if I want MP, I go with Run8. If I want freedom of content creation, Trainz.

 

[danny5]

Ghost Trains, Why do we have Ghost Trains?
We've had ghost trains when MP came out in TS10 and this has not been looked into for 4 years WHY.
I've tried to work out why it dose this.
I've looked at big routes,small routes,the same ghost trains.
with/without CMTM Sytem,ghost trains
lot & few rolling stock routes. ghost trains
with/without scenery ghost trains

 

[Trainlover16]

May I ask why or what may hold back trainz from allowing multiple users operate the same train. For example: an engineer operating a locomotive while a conductor can operate doors and other controls that a passenger car may have. This would allow for lots of interaction between players all in one train.

 

[JIb228]

May I ask why or what may hold back trainz from allowing multiple users operate the same train. For example: an engineer operating a locomotive while a conductor can operate doors and other controls that a passenger car may have. This would allow for lots of interaction between players all in one train.

Experience in Run8 says yes and no to this. For freight crews yes, it very nice and very fun. But passenger trains its very boring.

In Run8, and hopefully in T:ANE, freight crews can consist of a full 3 man crew if need be. One crew member sits in the cab, running the locomotive. Another crew member can act as the brakemen, cutting cars and throwing switches. And, if there is one, a third man can be the conductor, doing the "paper" work and radioing the dispatcher. This works VERY well in Run8, and can be a tone of fun.

Though for passenger trains that really only make station stops its kinda pointless. It can be done, and I have seen it done in Run8, but it looks boring to basically just sit there for a few hours until the train reaches its stop.

 

[Trainlover16]

I see what you mean Jib, but aside from stopping trains at specific spots on platforms, opening passenger doors and throwing manually operated switches can't the conductors for passenger trains do the "paper work", talk with dispatch, inform the engineer of speed restrictions/work zones, call signals, and operate cool controls in the coaches? This may be boring for most but for some this would be a dream come true. Passenger Conductors can also do reverse movements as well as yard work too. Unfortunately though none of the above can be accomplished with out being able to operate the controls of a train that is claimed by the engineer.

 

[JIb228]

All very true, and could be done/can be done in Run8.

As for Trainz, I dont see many passenger cars that have door controls. In in Trainz with out a dispatcher in the very anarchy type environment that MP sessions run I dont think this would work. Now if multi crew positions where available, and a dispatcher mode, then yes they would be very nice to have.