Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Saiko Kila]

What was point of material changes in .34.08? I'm rather disappointed with some of them. There are even some outright wrong changes, for example cedar is based on a tree which is not cedar at all, not even a conifer - and it still grows in coniferous forests in DF. It's like basing a wolf on Tasmanian wolf.

I suppose the only change which matters in game terms is solid density of materials. Unfortunately, DF simulation is rather weak in this area - real life Scots pines (Pinus sylvestris, one of the most important pines of the Old World) for instance, as timber, have density usually in range 400-550 kg/cubic metre, but can be less or more. Siberian larch (Larix sibirica) can weight 590 kg/m³, just as the larch in raws, sure, but it also can weight 450 kg/m³, and freshly after cutting it is 850 kg/m³. Back to cedars: Canadian Western Red Cedar weighs even 550 kg/m³ as fresh logs, but its timber is only 330-460 kg/m³. So some logs weight more like fresh ones, and some more like heavily dried or seasoned.

Similarly stones have different densities, but since they are more resistant to water penetrations the changes are proportionally smaller. Gabbro can weight 2700-3300 kg/m³ for example, the DF version weights 2920, close to average values.

Since there are such many natural variations, I don't see a point in changing values to some arbitrary ones, when DF cannot emulate such variations. I suppose it may be added in future, like there are variations in size of dwarves, but for now such changes are premature. At least when it comes to such variable materials as timber. I admit I don't like the changes for two reasons: one is hauling update, which makes it matter, but the other is... less than scientific reliability.

[PoodleIncident]

Toady knows this; whether he cares is another matter. The original values for stones were all the same, iirc. Surely this is better?

[Noodz]

afaik, uristocrat has been collecting this information, so applying the changes to the raws was straightforward for toady.

I agree that natural materials, especially timber, would be better modeled not as having a a probabilistic density based on natural values' mean and standard deviation, each log/boulder instance would have its density decided upon creation and would carry that property. But i get the impression this would place more burden on processing, as each object produced from these objects would have to carry additionalal properties.

All things considered, material properties updates have little cost and add some depth to the game. so why not?

[Garath]

have you considered to stop complaining and search for a thread where contributions like this are welcomed. Seriously. He's a programmer/mathematician. If he got something else wrong, feel free to make a suggestions thread or find a thread that goes along the same lines, but don't come here to complain without contrubuting to a solution, that's low.

[slothen]

I'm not really clear on exactly what densities were changed.  The best I can figure is that before, differing materials had differing densities, and that now those densities have been tweaked based on player input from that thread.  If the players do research and say "hey Toady, these raw values should be updated to better reflect reality."  And then he changes them, then that's a good thing.  If you have issues with the accuracy of the values, then that should go in the thread that has that discussion.

What was point of material changes in .34.08? I'm rather disappointed with some of them. There are even some outright wrong changes, for example cedar is based on a tree which is not cedar at all, not even a conifer - and it still grows in coniferous forests in DF.

Then this has merit, but it still isn't clear to me how the .34.08 change is a step in the wrong direction.

So some logs weight more like fresh ones, and some more like heavily dried or seasoned.

Well, the trees that dwarves cut down are kind of strange and only behave like other materials only once they're turned to logs.  Given the variation of density depending on age of the log, it would seem reasonable to take all the wood material definitions and change them to be uniformly that of dried wood and not fresh-cut.

Similarly stones have different densities, but since they are more resistant to water penetrations the changes are proportionally smaller. Gabbro can weight 2700-3300 kg/m³ for example, the DF version weights 2920, close to average values.

Most would take this as an instance where it has been done right.

Since there are such many natural variations, I don't see a point in changing values to some arbitrary ones, when DF cannot emulate such variations. I suppose it may be added in future, like there are variations in size of dwarves, but for now such changes are premature. At least when it comes to such variable materials as timber.

If you're suggesting that two otherwise identical instances of wood items should have differing densities based on a sampling from a normal distribution or by tracking the time at which they were cut down, then you are insane stark raving mad.  Attempting to emulate these variations is a terrible idea, and not even one of those low-priority but cool ideas.

Defining gabbro's density as the centerpoint of an observed range of densities is the opposite of arbitrary.

[Kestrel]

have you considered to stop complaining and search for a thread where contributions like this are welcomed. Seriously. He's a programmer/mathematician. If he got something else wrong, feel free to make a suggestions thread or find a thread that goes along the same lines, but don't come here to complain without contrubuting to a solution, that's low.

Sorry Garath, I don't mean to pick on you personally by pulling out your quote, but this just got to me.  I really loved how patient members of this forum were toward questions that might be slightly out of place or redundant, but somewhere in the past nine months people have been getting rather hostile.  I just don't think responses like yours improve the character of this forum.

Again, didn't mean to single you out!  It's much more of a generalized feeling, but your post kinda provided a good example for this.  I've been guilty of the same thing too.

[NW_Kohaku]

I'm not even sure what value would have made you happy here...

There isn't going to be a different density for different stages of wood.  Materials reference a single material definition for individual types of wood, and that referenced definition will be applied to all "cedar" wood types, regardless of whether they are trees, logs, or chairs made of "cedar". 

If you want to argue a value, do so here, in its thread.  It's certainly possible Uristocrat got some value wrong, but there's no point in arguing different densities at different stages of treatment of wood, because the game is not about to model that. 

(In fact, for a thread titled "what was the point", it begs the question of what the point would be in having wood suddenly get more dense just after chopping to represent a drying process that doesn't even take place in-game...)

Uristocrat posted his source for cedar in this post, and sources his information to the USDA Forestry Service.

And for the record, someone already brought this up in that thread:

Depending on definition, the specific gravity of cedar should either be:

470 - Redcedar (dry vs 440 green)
560 - Lebanon Cedar

Interesting.  I personally prefer the Lebanon Cedar, personally, given that it's the most famous.  But that makes the numbers merely a change from 570 to 560, which is well within the normal variation for materials like this.  Still, that's some fascinating tree information!  Can I ask where you found those numbers?  The only source I found pegged Lebanon cedars at 580.

[Girlinhat]

This is a case where simulated simplicity takes importance over accuracy.  It would be possible to have things be different semi-random densities, for instance gabbro could range from 2700 to 3300 density.  But that would add enormous strain on everything.  Right now, all items merely reference to a parent item template.  If you opened up the memory and looked at your stone stockpile, it would NOT say "This is gabbro is is so-and-so dense and so-and-so density and so-and-so value."  It only says "This is gabbro.  This is gabbro.  This is gabbro."  The program them looks to the template for gabbro and grabs data.  If you added random variance, it would need to say "This is gabbro density 3000.  This is gabbro density 2700.  This is gabbro 2900."  You can imagine that for all the stone generated, this would easily quadruple the size of memory consumed and contribute much more directly to FPS decay.

Adding age and drying process would be much much worse, as you then include specific density AND age, but ALSO a per-tick or per-season check on the decay/drying rate of items.  Then you have a memory hurdle AND processing hurdle.

If you're arguing that the declared values are wrong, then I fail to see your point.  Designating things in the median range is what works for simulations, since handling all the specifics would be too difficult.

[Excedion]

To be honest I don't think people really care to much about having such overly realistic values. The minute change in density of wood between fresh cut and dried would probably be unnoticeable never mind the fact it would require the game not only to keep checking and referencing different values for different stages of the drying process. I mean as if the game doesn't run slow enough already and you're complaining toady didn't sacrifice even more FPS for such a minute thing? Arguing for declared values is one thing, arguing for a density overhaul and material based states is quite another.

[Loud Whispers]

To be honest I don't think people really care to much about having such overly realistic values.

They are not needed, but damn are they lovely.

[King Mir]

I think natural variability would at best be a detriment to game play. It wouldn't effect player behavior most of the time. But there may be times when a player wants a precise weight for a particular purpose. In the real world, people would weigh some rocks, and pinch a bit off or add a bit as needed but in DF that's not possible. Having materials with a fixed mass is useful in those cases.

For example, someone may want to skip a minecart full of wood across a channel and catch it on the landing in a single tile. If the weight of the contents are too variable, then you cannot make the channel and landing the perfect width and distance. That's one less way to have fun in DF.

I'm in general in favor of more accurate simulation, but not in this.

[Ubiq]

What was point of material changes in .34.08? I'm rather disappointed with some of them. There are even some outright wrong changes, for example cedar is based on a tree which is not cedar at all, not even a conifer - and it still grows in coniferous forests in DF. It's like basing a wolf on Tasmanian wolf.

...

Back to cedars: Canadian Western Red Cedar weighs even 550 kg/m³ as fresh logs, but its timber is only 330-460 kg/m³. So some logs weight more like fresh ones, and some more like heavily dried or seasoned.

So, you're unhappy that the specific gravity of cedar in the game is based on a tree that isn't actually a cedar and then point out the difference in green versus dry weights for cedar by using a localized member of a juniper species?

[Loud Whispers]

So, you're unhappy that the specific gravity of cedar in the game is based on a tree that isn't actually a cedar and then point out the difference in green versus dry weights for cedar by using a localized member of a juniper species?

No he's unhappy that the specific mass of cedar in the game is based on a tree that is not a cedar.

Not the gravity.

The gravity is constant.

/Physics/

[Garath]

So, you're unhappy that the specific gravity of cedar in the game is based on a tree that isn't actually a cedar and then point out the difference in green versus dry weights for cedar by using a localized member of a juniper species?

No he's unhappy that the specific mass of cedar in the game is based on a tree that is not a cedar.

Not the gravity.

The gravity is constant.

/Physics/

wasn't it that the specs of cedar in colder biomes were based on a cedar that didn't grow in colder biomes? since there are different varieties of cedar that grow in various biomes. It was based on cedar just fine, but saying all cedar has density x was a bit broadly taken, right?

-snip-
Again, didn't mean to single you out!  It's much more of a generalized feeling, but your post kinda provided a good example for this.  I've been guilty of the same thing too.

no offence taken, since you're right, I just felt a bit testy as some untrue myths have been spreading and constantly come up in various questions/suggestions and the questions on the df gameplay questions thread are getting a bit repetetive (no flux, animal training how, why not making steel). I think I need to take a break from answering questions.

[Ubiq]

Not the gravity.

The gravity is constant.

/Physics/

Wow, don't care.

The point remains that the wood referenced still isn't correct as it isn't even a member of the cedar family either and, further, is specifically a temperate species while the cedar of the game is a temperate/tropical species. White-cedar actually has a better claim to it because of its regional distribution even if it isn't the tree that we think of as cedar.