News:

Use the "Forum Search"
It may help you to find anything in the forum ;).

Reduced forum activity in recent years

Started by jamespetts, October 17, 2015, 11:01:08 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

jamespetts

I have only just noticed some of the recent discussion here. I note with interest Prissi's work on making things work better with 4k monitors, which is a very worthwhile thing, as Simutrans (especially Experimental, with its higher computational load) works well in the desktop arena, and desktops are very much heading towards 4k (and possibly even 8k) as being quite common.

The graphics are indeed old-fashioned, but the relatively low-fi graphics allows for much greater quantity of output (and therefore simulation depth) than beautiful, detailed 3d graphics; Cities: Skylines looks beautiful, but there is only one type of 'bus.  Train Fever is more interesting: it is totally modern in graphics, but much more limited, from what I understand (I do not have a copy) than Simutrans in terms of scale, I think in part because of its modernity. There is a great advantage in working with older code in Simutrans in that it was optimised for much slower computers: we can therefore have games on a huge scale with lots more simulation depth than is possible with more modern code and graphics. (The UI could possibly be updated somewhat, but this is likely to be a major job and one that it would be tricky to port to Experimental, although, as noted, the improvements to allow Simutrans to work better in 4k would be important to port).

On gameplay: I agree with Isidoro that the gameplay in Simutrans is deeper than Cities:Skylines (although I have a copy of the latter and enjoy playing it). Skylines does not have any sense of progressing through time, for example, and simulates a wider range of things in a smaller geographical area in a more limited way. One thing that Skylines does particularly well, however (at least by comparison to other games), is road traffic simulation, which is one thing that I find particularly enjoyable about the game. Rail transport simulation is very greatly lacking, and air and water transport simulation are extremely rudimentary (airports and docks being treated as utilities rather than transport nodes).

As to Experimental features, it is a conscious decision not to have released a stable version for a while because I wish to get the core balance features implemented first. Releases take a huge amount of time and effort, which time and effort would better be spent on improving the balance critical features first. Given that, without the pakset being fully cost balanced, Simutrans-Experimental is really only half a game, given that balancing is a gargantuan task, given that balancing before implementing major balance-critical features is futile and given that my time is extremely limited by comparison to the amount of time needed to achieve what needs to be achieved, this course of action rather recommends itself quite strongly. What is needed more than anything else to get Experimental into a state where it is reasonably stable and playable is people to work on the core balance features in the code. With just me working on it, I am afraid that it is likely to be years. (The fact that I spent most of last year not working on Simutrans owing to moving house has not helped much either).

I at least have the great advantage now that there are several people who are compiling from source and playing the development version, reporting bugs as they go, which makes tracking down problems much easier and more efficient than waiting until after a release.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Sarlock

The massive map sizes permitted in Simutrans is its greatest appeal to me.  The recent SimCity incarnations have been very visually appealing and have many nice attributes, but lack the large map size that appeals to many simulation/world recreation enthusiasts.  Even SimCity 4 had significant map limitations that impeded with large world simulations.

I feel that Simutrans is perfectly situated to be the platform of choice for many of these enthusiasts.  It has a wonderful, robust and well tested transport simulation system, can support very large maps and is very customizable.

Its real limitation graphically is the difficulty with which a new "modder" can create new graphics to add to the game.  In SimCity 4, with the BAT, this is a fairly simple process and a new user can learn the basics fairly quickly.  In order to create high quality 3D type graphics for Simutrans, however, one has to learn a 3D application like Blender, which takes years to master, then undertake a fairly lengthy conversion process from Blender to PNG files, create the DAT file and port it to Simutrans.

If there is an area to focus, I would focus my energy here.  We would attract hordes of modders wanting to create their simutations/world recreations in Simutrans and this would bring us many skilled and capable artists and programmers.

Adding alpha channel support would be a wonderful feature in the future, but as I have demonstrated with my work, even without an alpha channel the graphics can look pretty amazing using a detailed 3D model in an editor like Blender.

I will continue my work and see what else can be accomplished, but one man can only do so much in his spare time :)
Current projects: Pak128 Trees, blender graphics

jamespetts

A few people working together (Zeno, I believe, and The Hood, with a small amount of work from me to refine it slightly) have produced (some time ago now) automatic rendering scripts for Blender that significantly reduce the amount of time that it takes to get graphics into the game, especially when combined with ImageMasker (for 1x1 tile graphics, including all vehicles and ways) and Tile Cutter (for larger graphics, such as larger buildings).

One does not really need to master Blender for producing Simutrans graphics, as a basic knowledge will suffice to produce the low-fi graphics that will suffice for the 128x128px .png files used (perhaps slightly more proficiency might be needed for good graphics at a 192px tile size). Blender is not too difficult to get to know - certainly much easier than learning to code.

However, the most difficult bit is actually making all the various graphics consistent with one another (despite taking great pains to ensure graphical consistency in Pak128.Britain, it turns out that many vehicles have to be re-scaled because they were on many inconsistent scales, which process is taking a long time).
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Sarlock

It took me about two years to get fairly proficient at Blender, even though I am still a long ways off.  Then again, I am not young anymore and don't have much free time to learn this stuff, so the learning curve is much steeper and longer.  It doesn't take long to create low-fi graphics, but it takes much longer to create detailed models.

I've tried using the scripts before but didn't have much success.  I'll have to try them again and see if I can make them work for me.

Being able to push a button in Blender and have it use some preset parameters and directly put it in to a PAK file for Simutrans would be huge.  I've tried to learn some Python, but again, I am not young anymore and I don't have the free time  ;)
Current projects: Pak128 Trees, blender graphics

jamespetts

The workflow is not quite as fast as "make a .pak of this", not least because you also need a .dat file. The workflow (for 1x1 objects) is:

(1) make the model from a template with the lighting/rotation already set up;
(2) run the export script (a button in Blender if the addon is installed, along with some settings such as how many rotations, whether a vehicle or not and the name of the file);
(3) import into ImageMasker* to consolidate all into a single image;
(4) import into the GIMP to set the background colour to the correct transparent colour (use fill tool with tolerance of 0 on the black pixels) then save;
(5) create the .dat file from an existing template, modifying the filenames from one to another using a search/replace function; and
(6) recompile the pakset fully using an automated script for doing so with one command.

Because of the ability to do lots of steps (3) and (4) all in one go very quickly, it is easier to produce a decent number of graphics at once in steps (1) and (2), process a large number together in (3) and (4), complete (5) all at the same time, and then run (6) once.

* ImageMasker was actually intended for a sligthly differnet workflow involving using anti-aliasing with a background and special mask images, but it is not used this way in Pak128.Britain. Using the proper workflow for this adds steps, but can reduce the appearance of fading to black around the edges of objects.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

Quote from: dekema2 on November 23, 2015, 03:49:13 AM
I agree. There is nothing wrong with the basic mechanics of Simutrans to me, it's really the UI and the graphics that feel suck in 2000.

I think Simutrans just needs a polishing to bring it into 2016.
The graphics style of the simutrans world, i.e, the paksets are fully in trend at the moment. Look at all the indie game smash hits of late, which embrace the pixelated look. However, these also have very modern and very well thought out UIs -- no wonder being designed now and by full-time devs. These UIs might at times look old, but are modern in the sense of efficiency and reduction on the relevant, and most imporantly by not being hand-knit but using the full tool kit adoptable game engines offer.

The basic mechanics of Simutrans however, is not very much at home in these days. The trend went towards more limited interactions that still generate a rich complexity. Sandbox games are notoriously difficult to maintain. Games like FTL or Don't Starve have very skillfully implemented game mechanisms to reduce the complexity (in the sense of orders of freedom of a player) quietly. The latest simcity tried also to curb complexity (not very skillfully though). Cities Skylines or Europa Universalis are exceptions.

jamespetts

I like to think of Simutrans (perhaps especially as Experimental) as another one of those exceptions, for the perhaps niche group of people who like seriously realistic and economically accurate transport simulation games.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Ters

I find the simplicity of the economic model in Simutrans more bothersome than the graphics. But as I've written many times before, the Simutrans community is widely diverse. Some play it like a businessman, some play it as a model railroad. These two extremes pull Simutrans in opposite directions. Then there are those that want Simusociety, which is sort of orthogonal to this.

sdog

Quote from: Ters on November 25, 2015, 06:58:10 AM
I find the simplicity of the economic model in Simutrans more bothersome than the graphics. But as I've written many times before, the Simutrans community is widely diverse. Some play it like a businessman, some play it as a model railroad. These two extremes pull Simutrans in opposite directions. Then there are those that want Simusociety, which is sort of orthogonal to this.
Well, a lot of orthogonal vectors span a nice vector space.

More seriously, without a nice model-railroad packaging, it would be too serious to draw many people. Likewise if it were a pure simulation, being gamy at times is not a bad thing for a game. A simulation of some aspects of society make compellingly complex problems, that relate to real life.

I don't tink the absence is related as much to the nature of simutrans or it being out of fashion, it is much more a typical demographic problem. As an immature project it was easier to draw very commited people. These stop playing or being interested after a random time. As more developed game it is easier to engage to players, however, it is also more difficult to get very dedicated players, as here the project has to compete with younger less developed projects.

sdog

This thread stopped at trying to analyse why the forum is less active. Do we have also ways to get it more active?

What comes to mind first, as a short-term and immediate measure is to increase activity on our social media platforms.

I'm doing things only from time to time on our plus site, which is mostly dormant because of that.

People like screenshots, and andz is keeping up the screenshot contests, which is very helpful. But that is not all, I see more interest in new developments. There are often new impressive buildings. Lately mostly from Sarlock and the pak comic team.

I'm afraid that I've lost overview of what is currently being worked on the development side of simutrans. Would someone who is following that write a few words on an interesting topic? Preferably someone who isn't a dev who would waste valuable development time, but one of the many devotees, commenters, or 'lurkers'?

jamespetts

Interestingly, the forum statistics show that this month is the second best this year, after January.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

Quote from: jamespetts on November 30, 2015, 07:42:22 PM
Interestingly, the forum statistics show that this month is the second best this year, after January.
Perhaps because the new version was released this month.

jamespetts

Perhaps indeed, although I have noticed an increase in discussions relating to Experimental, too, for which there has not been a new release, so perhaps not that alone.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

Quote from: jamespetts on November 30, 2015, 08:29:23 PM
Perhaps indeed, although I have noticed an increase in discussions relating to Experimental, too, for which there has not been a new release, so perhaps not that alone.
That correlates strongly with the increased presence of experimental's lead developer, who until recently occupied busy with creating a house-sized time-machine for his famous tea parties. :-)

AP

Quote from: sdog on November 30, 2015, 09:38:32 PM
That correlates strongly with the increased presence of experimental's lead developer, who until recently occupied busy with creating a house-sized time-machine for his famous tea parties. :-)

From a purely personal perspective - I play simutrans experimental & pak-gb - there are major known balancing issues with experimental, whilst these were being investigated / addressed there was a lot of playtesting going on, but things definitely went quiet (as might be expected) once that process slowed down. Once the balancing is done (and the new features incorporated), hopefully it should be viable for long-term play.

Once you've got used to the extra features / benefits of experimental, it's hard to go back to standard.

jamespetts

Quote from: AP on December 06, 2015, 12:40:24 PM
From a purely personal perspective - I play simutrans experimental & pak-gb - there are major known balancing issues with experimental, whilst these were being investigated / addressed there was a lot of playtesting going on, but things definitely went quiet (as might be expected) once that process slowed down. Once the balancing is done (and the new features incorporated), hopefully it should be viable for long-term play.

Once you've got used to the extra features / benefits of experimental, it's hard to go back to standard.

Out of interest, what do you see as the main outstanding balance issues in Experimental?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Quote from: jamespetts on December 06, 2015, 01:10:01 PM
Out of interest, what do you see as the main outstanding balance issues in Experimental?

The biggest issue I see is "Unpredictable Economic Returns".  In experimental, whilst there is a lot more sophistication/intelligence evident in the simulation, it is genuinely hard to make a profit. There are a lot of complex ways to spend money, many ways to spend it badly, and few ways to reliably get it: e.g.

       
  • Some industry chains inexplicably offer better returns than others.
  • Industries with issues of unproductive months due to amount of stock in transit.
  • Issues with distances between industries (why deliver milk to the dairy next door when you can sell the same milk for dramatically greater profit on the next continent)
  • Pre-rail, one must have shipyards everywhere because of the way restrictions (sea vs river0, yet they each cost a fair bit of maintenance. Ditto with stables if you don't start with free intercity roads on the map.
  • Cost per unit of storage is not equal across all units (notably staging inn).
For longer games other issues come into play

       
  • Excessive urban sprawl (server game became near 100% urban)
  • Inefficiency at removing obsolete infrastructure (can take as long to unbuild as to build -totally soul destroying!)
  • Compounding Success - Compound interest on capital - server games (players who do stumble upon a successful industrial model reap huge cumulative rewards - rather than losing some to dividend payments they get compound interest, which lets them adopt increasingly inefficient practices which exclude newer players, rather than becoming burdened by their inefficiency)

jamespetts

Thank you - that is useful. To give some idea of the progress on these elements:

(1) revenues per type of goods per unit of distance should be more or less balanced; can you give examples of where, if anywhere, this does not appear to be the case?;
(2) I think that there were bugs in 11.x that are now fixed that were the cause of this;
(3) I have done some work to address this in the development build by limiting the distance between industries (this can be set per industry in the pakset);
(4) if I remember correctly, I greatly reduced the maintenance for shipyards (for the unreleased development version) a while back after your earlier comments on this (or possibly someone else's - I cannot remember whose now);
(5) this will need to be looked at when the pakset is fully balanced, although I may have made some adjustments (I cannot immediately recall);
(6) I plan to rework the entire town growth model fairly soon, but, in the meantime, the town growth rates can be adjusted in simuconf.tab;
(7) can you explain this in a little more detail?;
(8) I am just removing the facility to pay credit interest now.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

The Hood

An interesting discussion, especially as I've not been around on here myself much of late. Increased responsibility at work means less time for simutrans amongst other things. But another factor is there do seem to be fewer people contributing. With the pak128.Britain project I find it harder to motivate myself to contribute when I'm the only person drawing new stuff. The aspirational graphics list is very long and it feels it will take forever to do. Downloads on sourceforge are relatively constant though, especially considering no updates this year. I guess people are playing but just not wanting to get involved.

Regarding balancing, that was next on my hit list for the set (rather than new graphics) but no promises on a timescale!

jamespetts

I have been focussing on coding lately, although I have been producing new and updated graphics: see the new signal graphics that I have produced (compatible with Standard) and the gradual re-scaling of early railway vehicles. I am also planning on adding some L&TSR locomotives and carriages at some point, and looking at adding some additional 'buses over Christmas, although consideration will first have to be given to re-scaling all road vehicles to match the trams, trains and boats, and "all road vehicles" unfortunately will include all buildings with parked cars, so careful consideration as to how best to do this will have to be given before proceeding (preferably with some co-ordination with the Standard pakset developers).
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

I've posted the last screenshot contest (SMBC) images on the g+ site. Overall this increased the page views by more than 200% to about 3.5k. There is still interest in Simutrans on social media, when we make ourselves heard.

We are missing videos however. Perhaps we could host a video contest in the pattern of the SMBC? Or perhaps animated screenshots, vine like animated gifs?

(I am mostly ignorant of videos. I find it bothersome to click them, when i can gain the same information in a fraction of the time. I fear however, that I am not really representative for social media users. Thus I'm not really the right person to start these.)

colonyan

To be honest I have been almost never neither playing nor visiting forum for 2 whole years. But since few days ago, Simutrans is thing again. Mainly thanks to pak.nippon being constantly worked on. It kept me checking back once in a few month.

Cities:skyline was very fine game but it felt little too artificial on how city develop. I never was able to get into train fever due to UI, monotone map.

I have vague new pak project idea which I will start really humble and hopefully get it rolling as a hobby project.
I'd like to exploit many dat file and config files to make something new for standard version using 64 size. I had some project but it never lasted from different reasons like vision for goal.

For instance new idea have things to do with...
- Maximum interaction between passenger and industry chains. Full use of factory boost.
- Minimum type of freight cargo/vehicle type and maximum item type.
- Minimum vehicle type number.
- Industry chain as the dominant growth. Focus on Industry and Passenger.
- Minimalistic freight counting. 1 wagon carry's 1 unit load. 1 passenger group per 1 car.  Abstraction to core.
- No mail and No power
- No Com and Ind city building. Only residential city building.
- Very wide but shallow industry chain.
- Wide variation of final product consumer building and mix of freights intake. Generalization as a commercial district.
- No age progress. Start from modern-present.
- All new graphics. Streamlining production. Efficiency!

Ok. Having said that it makes me more constrained to show progress in near future. Let's get things rolling...

prissi

YOu can make a city with com only, but only res or only ind will not work. Com counts as both, so only com will work.

jamespetts

Since the Steam launch, I note that May 2016 had the highest number of new members in any one month since March 2013.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

DrSuperGood

Quote
I note that May 2016 had the highest number of new members in any one month since March 2013.
Of which A lot of them were asking about something related to steam? Still I guess more people is better in some way but I just wish development was still as active as it was 1-2 years ago for the standard branch.

AP

Just re-reading this thread - realised I never answered James' question:
Quote from: jamespetts on December 06, 2015, 02:36:14 PM
Thank you - that is useful.
(7) can you explain this in a little more detail?;
It was regarding "Inefficiency at removing obsolete infrastructure ".

What I meant was, suppose you have engineered a canal route through challenging terrain, but now want to replace it with a railway, re-using the earthworks and routes through town centres etc. Getting all the boats "home" is time consuming. You can't over-write canal with railway, so there is a slow delete-and-replace game. In urban context the city may jump in with buildings or roads whilst you do this. On server games, other players might.

In another context, suppose your previously profitable empire is suddently losing money hand over fist. You can't fix it without vast amounts of game time which you probably can't deploy quickly enough (especially in a server game). You want to consolidate your position - ideally "sell/delete everything instantly" and rebuild a more profitable network. But the tools to do that don't exist. You either wait for it to sink or start a new company and lose all your hard earned cash - neither ideal.

jamespetts

Quote from: AP on June 19, 2016, 03:56:56 PM
Just re-reading this thread - realised I never answered James' question:It was regarding "Inefficiency at removing obsolete infrastructure ".

What I meant was, suppose you have engineered a canal route through challenging terrain, but now want to replace it with a railway, re-using the earthworks and routes through town centres etc. Getting all the boats "home" is time consuming. You can't over-write canal with railway, so there is a slow delete-and-replace game. In urban context the city may jump in with buildings or roads whilst you do this. On server games, other players might.

In another context, suppose your previously profitable empire is suddently losing money hand over fist. You can't fix it without vast amounts of game time which you probably can't deploy quickly enough (especially in a server game). You want to consolidate your position - ideally "sell/delete everything instantly" and rebuild a more profitable network. But the tools to do that don't exist. You either wait for it to sink or start a new company and lose all your hard earned cash - neither ideal.


Thank you for your feedback. None of these things are different from Standard in their implementation, so these difficulties will apply equally to Standard. I can see the merit, in theory, of a tool to replace one type of way with another (although that would take a lot of effort to implement to deal with the complex consequences of this in lots of small ways).

As to selling or deleting everything instantly, are you sure that this is a good idea as a feature? I can only imagine the consequences of accidental use of such a feature. It cannot be often that such a thing be merited, surely? Would it suffice to have an option manually to liquidate a company, rather than to have to wait for it to be liquidated automatically? This might be a workable idea and not too hard to implement.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

AP

Quote from: jamespetts on June 19, 2016, 04:21:15 PM
As to selling or deleting everything instantly, are you sure that this is a good idea as a feature? I can only imagine the consequences of accidental use of such a feature. It cannot be often that such a thing be merited, surely? Would it suffice to have an option manually to liquidate a company, rather than to have to wait for it to be liquidated automatically? This might be a workable idea and not too hard to implement.
On single player games it's fairly irrelevant, but on  several server games I've got to the point of "oh crap" when I've realised I've made an unholy mess of the thing. Sometimes we struggle on, sometimes it's just not worth it.  Manually liquidate would be fine, so long as the player keeps the liquidity i.e the cash they made, to plough into another enterprise. Otherwise players "walk away" which is not the desired outcome I'm sure.

jamespetts

Quote from: AP on June 19, 2016, 04:54:54 PM
On single player games it's fairly irrelevant, but on  several server games I've got to the point of "oh crap" when I've realised I've made an unholy mess of the thing. Sometimes we struggle on, sometimes it's just not worth it.  Manually liquidate would be fine, so long as the player keeps the liquidity i.e the cash they made, to plough into another enterprise. Otherwise players "walk away" which is not the desired outcome I'm sure.

Keeping the cash after a manual liquidation would require a fundamental change in the way that liquidation works. I do plan to change the company management features at some point to allow, e.g., mergers, but the idea of a solvent liquidation (which is what this would be simulating: similar to a members' voluntary winding-up in UK company law) is quite different from what we have at present.

Edit: I have added reference to solvent liquidation in the coding projects thread. This should probably only be possible where a player's net wealth is greater than zero.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.