bridgewater-brunel.me.uk - Simutrans-Experimental - Pak128.Britain-Ex 0.9.0

[jamespetts]

Are things more stable for people now?

[ӔO]

only slightly more so.
Now I can stay connected for about 1 to 10 mins before desynching.

[Sarlock]

Was able to play for about an hour earlier today and stayed connected the whole time.  Will try again later.

[ӔO]

maybe the server game needs to be restarted after a certain amount of connections or hours?

I notice the game continually using up more memory until it crashes (at around 1.5GB) due to the game running out of memory.

restarting the game frees up quite a lot of memory. For me, the server game starts at around 1GB, then eventually creeps up to 1.5GB, even though the game does attempt to free up memory every now and then.

[jamespetts]

Hmm - the server is currently using about 1.3Gb of RAM, the server having a total of 3Gb available. I cannot immediately see how restarting or even reducing memory consumption would help. However, I am restarting now to test whether this makes any difference - I should be very interested in whether this assists.

[ӔO]

the restarted game was consuming under 1GB, so there is definitely not enough memory freeing up going on.

It didn't really change connectivity, however.

[jamespetts]

Are people still unable to connect for more than a very short time?

[asaphxiix]

it is quite more stable now.

[jamespetts]

Interesting - thank you for the feedback.

[jamespetts]

Updated to the latest version of Simutrans-Experimental (11.11).

[Sarlock]

Any plans to start a new game?  I've tried to start something up but it's too late in the game for any possibility of profit with small starting capital.  Patiently awaiting a restart

[prissi]

It may be wise to start a new thread each time you start a new servergame. Not least for clarity.

[Sarlock]

I think this has been the same game running for many months.  Year is 2026 now.

[jamespetts]

Sarlock - I was wondering about starting a new one. I was hoping to be able to do so with a new pakset release or with the new passenger generation code, however, as each server game takes a very long time to get to the end of time, as it were. I should be interested in others' views on the point.

Prissi - Sarlock is correct: this whole thread is about the one game.

[Sarlock]

Could you run a second server game?  Are you running this on a local machine?

Having a long running game certainly has the advantage of taxing the overall system's capabilities (as we discovered with the 50000 airplanes zooming around the map  ).  It is also a great way to test the organic city growth in a complex game.

With a lot of new features coming soon, I certainly can see your desire to wait.  Perhaps a shorter month-long game in the interim?

[jamespetts]

I run the game on a rented virtual server. I only rent a single core, so adding another game would mean both games running on the same core, which would impair performance. It may be necessary for the next game to be a shorter one if we start in 1750 and find that the game does not work very well in that era and some pakset adjustment is necessary, for instance.

I should be interested in a variety of people's views on the subject, however.

[ӔO]

I wouldn't mind having a new game, however it would be nice if the pakset bugs discovered so far are fixed with the new game.

I think the main ones were:
- lack of truck power
- cities consume too much power
- airport depot

[jamespetts]

Is the cities consuming power thing pakset specific...?

Edit: See here for some more detailed discussion on the topic.

[wlindley]

Despite what sounds like major upcoming game-engine changes, I support restarting at 1750 sometime soon. It would be splendid to see how the progress over the past almost year of Real Time will work a new world.  After a few months away, seeing the current world after two hundred fifty years is quite a lesson in real-world development -- we are all still "stuck" dealing with some of the decisions and compromises made centuries ago!

[jamespetts]

Thank you for your input.

What do people think about the (likely) eventuality that we will get through a good few decades of a new game, then the new major release will be upon us before the game gets any further? Would people want to go on then with a game calibrated with the 11.x series, or start afresh with 12.x even though the then current game will have many more years left to run?

[Sarlock]

I wouldn't have a problem with a reset after next major release... I see it as a test bed for new features.

Either turn down the bits_per_month to make a faster game or start in a later year (1920+/-) with more starting money so that we have more transportation options available.

[Junna]

I wouldn't have a problem with a reset after next major release... I see it as a test bed for new features.

Either turn down the bits_per_month to make a faster game or start in a later year (1920+/-) with more starting money so that we have more transportation options available.

Yes, more starting money for everyone except AEO! He doesn't need it!

[jamespetts]

Yes, more starting money for everyone except AEO! He doesn't need it!

AEO should start bankrupt and see if he can build an empire from that

[Sarlock]

This does bring up the small flaw with the interest model: a large company reaches a point where interest becomes so significant an income source that it allows the player to completely disregard any need for profitable operations.

[jamespetts]

That is why we need to simulate inflation...

[ӔO]

Yes, more starting money for everyone except AEO! He doesn't need it!

hey, I tried to dump some 90% of my bank account, but my attempts crashed the game

[jamespetts]

I wonder whether, if we make this realistic enough and then set AEO on it, he'll work out how to solve lots of apparently intractable real life transport problems...

[asaphxiix]

I do think the old game has reached a point worth concluding at, not least because the station coverage is no longer quite relevant.

As for the next game and the major release after that, I would gladly volunteer to set up another vserver when that time comes (or even before that), perhaps even one that can host more than one game for a variety of tastes etc., so the next game won't have to be stopped.

[asaphxiix]

I wonder whether, if we make this realistic enough and then set AEO on it, he'll work out how to solve lots of apparently intractable real life transport problems...

And I was wondering whether AEO's profession is a transportation engineer, or an engineer of another sort.

[jamespetts]

Asaph - thank you for your input. Any more Simutrans-Experimental server hosting would indeed be welcomed. I should rather like to get the next pakset release out before restarting the server: I am currently working on finishing off the project that has been on the go for an awfully long time: finishing off the Midland Railway carriages. I then need to integrate Junna's recent creations before releasing the next version.

[ӔO]

lol, no I'm just a gamer.

[Sarlock]

That is why we need to simulate inflation...

Inflation and interest rates are related to one another.  Having an interest rate only seems to serve the most profitable players that have accumulated enough cash reserves.  Inflation is already a factor in the game with vehicle prices increasing over time - though there is a corresponding potential for higher cash flow/profits as well.  If anything, an inflation rate should be in place instead of an interest rate, serving to erode the purchasing power of accumulated excess cash reserves.  The player should always have incentive to create and maintain a profitable transportation network.

[jamespetts]

I have now upgraded the server to the latest version of Simutrans-Experimental, 11.12. Users will need this latest version to join the game.

[ӔO]

seems like there is a problem with trains not using choose signals anymore.

[jamespetts]

Is this specific to the server, or is this a general issue with 11.12?