Coal burners mess up the economy

[Teppo Mäenpää]

Background:

Coal burners make coal out of trunk, with a low efficiency. Coal can also be mined; therefore coal made by burners is a small enhancement in most maps, while trunk is essential in the economy in many ways. Coal burners easily grasp all the available wood and therefore harm the economy.

Actual description:
I think that coal burners should only work when there is no shortage of trunk.

Additional information:
If I was the dictator of this project, I would apply this right away. However, we have a feature freeze and it would be surprising if everybody shared this view..

[Teppo Mäenpää (kxq)]

This is a small subset of what was discussed in Bug #1236538. While that bug is a mess, this is a clear case, in my opinion.

[SirVer (sirver)]

what is wrong with reducing the target quantities for trunks and increasing for coal? I mean, when a player builds a coal burner, she is usually already starved for coal and wants probably that it eats all the trunk. With your change it seems there is no way to make the burner agressive again, right?

This is a feature change and should not go into b18, no matter what.

[Teppo Mäenpää (kxq)]

In short: Reducing the target quantities for trunks and increasing those for coal does not help.

Longer answer:

Having some trunk is important. After this change, burners shut off if there is less than this much trunk.

Without this change, the target for trunk does not really matter. If there is demand of coal, the burners easily hog all spare trunks. Of course, if the player has more than enough coal, then the burners stop.

In my experience, it is rather common scenario that a player uses most if not all mined resources to imrove soldiers, and still would like to build. In such case, the burners harm the economy.

The downside of this change is, of course, that burners do not work at all if the number of available trunks is below the target value of the economy. I do not see this a problem, since even in maps like Fellowships one can reduce the target number of trunks when the need for coal exceeds the need to have trunks at hand.

At least some people do build burners also if they have access to some coal.

Yes, this is a feature request and not really a bug. I am not at all surprised by your verdict on b18.

[_aD (ad-simplypeachy)]

I believe the coal burner works as expected and needs no change. It consumes input wares until the economy target is met.

Set the coal burner trunk priority to low and trunks will go to other sites more often than the coal burner. You ought to check to see how much difference running a coal burner really makes - just because you have zero trunks in stock does not mean that your wood economy is broken. If you have plenty of trunks being generated then you may have zero spare, but still enjoy a healthy economy.

[_aD (ad-simplypeachy)]

I wanted to quantify this discussion somewhat, so set up two crude experiments. Each was run at 1x speed and with six lumberjacks running at around 100%. In each test there were no other trunk- or wood-consuming sites.

Experiment one:
I created four Empire mill construction sites, not connected to the roads. I then paused the game, connected them and unpaused, timing how long it took for the construction of all four to complete.

Starting with 40 trunks in stock the experiment took 5 mins 41 secs to complete.
With no trunks in stock and four wood burners operating it took 4 mins 56 secs to complete.
(Yes, the results seem backwards!)

Experiment two:
I added two sawmills and two weapon smithies. The weapon smithies had all wares except wood set to zero. I set the wood lance target to 50. The weapon smithies were full of wood but 0 wood and 0 trunks were in stock.

Without any wood burners operating it took 22 mins 9 secs to bring the wood lance stock to 50.
With four wood burners operating it took 25 mins 15 secs to bring the wood lance stock to 50.

The first experiment showed that the supply of 12 trunks needed to build the four sites was not adversely affected. The second experiment, which obviously requires many more trunks, shows a detriment to the wood supply. In my opinion, this detriment is proportional and doesn't seem to suggest that the coal burners are "greedy".

[Launchpad Janitor (janitor)]

[Expired for widelands because there has been no activity for 60 days.]

[SirVer (sirver)]

I think aD_ experiment is support for the opinion that the coal burners are not broken. I agree though that they are hard to use correctly in some sense and are special in the way that they consume more wares faster than most buildings. So I can also understand Teppo's opinion. For now I think it is best to not change them and instead have open ears if we might to revisit this topic in the future.

[Teppo Mäenpää (kxq)]

I made an experiment too. Unfortunately with barbarians, since I had forgotten what _aD did.

Construction time of two babarian fortresses increased from 100 to 139 seconds in my setting. The setup mimicked what I remembered from _aD's experiment. I repeated the play few times, and the results hold within few seconds. I would be curious to see the experiment 1 savegame.

I did my timings using gametime.

[SirVer (sirver)]

39

[SirVer (sirver)]

geez, sorry for that last comment. somehow came on the post button :/.

39% time increase when you add a big consumer of wood doesn't sound too bad to me - or rather expected. Could you detail your experiment a little more, I do not see how this number alone supports your cause.

[Teppo Mäenpää (kxq)]

I have tried few games.

Basically, the percentage goes up with the following:

- Make a setup which consumes some trunks and some planks/hardwood

- Empty the warehouse somehow

- Either connect or do not connect the burners

- wait a predetermined amount of time after touching the burners.

- Then do the actual timing.

In my opinion, those are reasonable steps. However, this is a road to cherrypicking.

>what is wrong with reducing the target quantities for trunks and increasing for coal?

That does not help with the problem described in the bug report.

> I mean, when a player builds a coal burner, she is usually already starved for coal and wants probably that it eats all the trunk.

Currently, it would be stupid to build burners otherwise. However, if the change was applied, and a player would like to convert all his trunk to coal, then reducing the target trunk count would do the trick.

>With your change it seems there is no way to make the burner agressive again, right?

With target trunk count of one, I would say that they are aggresive enough: One trunk goes to warehouse, all burners start working, and then submit a request of 6 trunks each.

Of course, it would be possible to add a "burn nicely / burn aggressively" button, but that would be an overkill.

I still would like to do this change. Why a "don't fix" status? I do not undestand that.

[_aD (ad-simplypeachy)]

I believe my experiment's results on comment 5 were enough to justify this behaviour as "working as intended". Trunks are only kept in stock so that they can be used. If a coal burner needs to meet the coal target it, burns them.

[Teppo Mäenpää (kxq)]

_aD: Your scenario is a road block test. Making the roads more sensible cuts some 40 seconds of the construction time (test 1), compared to the few seconds difference you saw.

Your tests do not tell anything of the actual question, sorry.

[Teppo Mäenpää (kxq)]

The right way to test this is, in my opinion, the following:

- Play a game against human players, on a map without any coal present. Pick a wood-consuming tribe* (Atlanteans, not Empire). Human opponent, since current AI handles this badly. Try to make some hero-grade soldiers. Observe that everything stagnates while coal burners are doing their best.

- Apply the patch, and repeat.

In my opinion, there is a great difference. Also, I cannot see any drawbacks**. This is true also for maps where coal is present.

The suggested modification, attached to this bug report, is rather uninvasive. What I would like to do is something like this:

https://bugs.launchpad.net/widelands/+bug/1236538/comments/19

That would already be a significantly larger change.

_aD: I do not really understand, why are you opposing this. Do you oppose all changes where you do not see clear benefits to those games you typically play? If no, then I would really like to understand you opposition.

*) Empire needs some wood, I agree. Please try something else anyway.

**) almost nothing.. longer story. Try it.

[Launchpad Janitor (janitor)]

[Expired for widelands because there has been no activity for 60 days.]

[Teppo Mäenpää (kxq)]

I still think that this is a problem, and the current behavior is not well thought-of:

- Coal-less mines have been uncommon

- Comment https://bugs.launchpad.net/widelands/+bug/1344179/comments/3 hints that way.

[Teppo Mäenpää (kxq)]

typo in the above comment. Intended to say, coal-less maps.