Tech Prototype Issues

[LDiCesare]

Starting the thread referred to in Demo6 Technology mode
The goal of this thread is to express:
1)What techs do we want/need in the demo for what reasons.
2)What prereq techs are needed for these.
3)What additional code, if any, needs be done in order to have the thing working.
4)What we need in which xml file.
5)What activities (as defined in the tech model) feed the research in these areas. In effect, what do we give research to in order to improve the tech.

The whole thing getting priority stamps.

I imagine model leads/coders will ask for point 1 and tell what they need in point 3/4, and Rodrigo will provide answers to point 2.

Here is my go for military model:
(High priority)
1- Military tactics tech (or application?).
2- Probably no requirement.
3- My problem is to be able to pump research points to Military tactics.
4- Get it into the technology.xml file
5- Related to 3: What activity is there/can I use?

1- Warhorses
2- ?
3- Code done.
4- Just need the tech in the technology.xml.
5- ?

(Lower priority)
1- Techs to increase particular units, like metallurgy for spears (phalanxes, legions) or ranged fire tactics for archers.
2- ?
3- No new code needed.
4- I just need to have these techs in the technology.xml file and know them so I can reference them in the military file. I could reference them right now, but it's no use cluttering xml files with unused stuff.
5- ?

(Edited to add in at the beginning of the Thread point 5)

[Lemmy]

1- Warhorses
2- ?
3- Code done.
4- Just need the tech in the technology.xml.

a prereq for war horses could be some sort of domestication.

Techs to increase particular units, like metallurgy for spears (phalanxes, legions) or ranged fire tactics for archers.

what do you mean with increasing particular units?

[LDiCesare]

By particular unit, I mean a unit type (actually element), like:
Warhorses => bonus for chargers and cavalry elements
Ranged fire tactics => Bonus for Archers, maybe war chariots.
Metallurgy => bonus for heavy spear elements and any element using armour.

[Gary Thomas]

And also:

5: What activities (as defined in the tech model) feed the research in these areas. In effect, what do we give research to in order to improve (or allow) warhorses?

As an aside I am not really sure that such a thing as a warhorse is properly part of the development of cavalry. Any horse strong enough to bear a rider could be used for cavalry. Previous to that development, smaller horses, or ponies, could pull a chariot. It would be rather nice to have this sort of detail realistic, since we have gone to so much trouble to make it possible.

As far as I see, the tech is "horses". The various types of riding, from a couple of ponies pulling a chariot to a destrier of the late middle ages, should be specified by the increment in the level parameter of the horses tech, with the use of saddle and stirrups coming into the same progression.

Cheers

[LDiCesare]

Gary, I edited the top post to include your 5th point.
As to horses, I think we can do a lot of bickering here. The problem comes from the fact that warhorse is a word in English, but it might be better to say war horse or horses for war or horse warfare. That includes breeding horses, using saddles, stirrups and so forth. So we can rename to horse warfare? Or simply horses, but then ploughing and warfare don't breed the same horses.

[Mark_Everson]

Hi Gents:

For the moment, my take is that a single tech of "horses" is the right way to go. We can always change it later if that doesn't work out in playtesting.

I have done some further steps toward actually using tech in the econ model. I immediately came across one request of type (3)

3) Tech turns are not currently run automatically, at least in the version of the code I have. I had to execute TechnologyFactory.update() by hand myself to get anything to happen. (High Priority)

I ran the same test as Gary did earlier, Put 100 RPs into the Food activity and see how Farming grows. I had it growing very quickly, but it turned out to be the same as Gary got in his tests, so that part looks like its working

Richard, My request are for you to put in another activity. BTW an up-to-date technology.xml should be in the "clash testbed" stuff I sent you a while ago.

5) Activity called Manufacturing, and it should give its points (for now) to Metallurgy -20% and Production - 80% Between the farming and production stuff I think we'll have a good basis for experimenting on tech using the econ model. (Medium Priority)

I think we also need to work out between us Some notion of what an RP is so we can figure RP inputs to activities with some rational basis. I'll put that in the other thread though.

[Mark_Everson]

4) I think we need to set tech upkeep to zero for now. Some thinking about RPs and upkeep have lead me to the conclusion that figuring out what the upkeep value needs to be for any tech is a nontrivial task involving upkeep being a function of tech level and probably population.

Since it is not a critical part of the model I propose we set all upkeep to 0 for the moment, and consider bringing it back later when things are further along.

[Richard Bruns]

The existing Simple_Tech_Test.xml seems to come close to fulfilling these requests. There is a military tactics tech, and different levels of it are requirement for various individual unit applications. There is also a "warhorses" tech that I will soon rename to simply "horses." Different unit techs require certain levels of horses or metallurgy, and higer values in these techs should increase their effectiveness.

The current "production" activity feeds into the production tech only. Do you want me to add an additional Manufacturing activity or simply change it so it also feeds the Metallurgy tech as well?

I believe that I have said that the tech model is a tool for other model designers to use as they see fit. That's why I built the online tech xml file builder. It allows any team member to view, modify, and add techs to the tech tree. The only restriction to user functionality is that I haven't figured out how to let the user upload existing xml files, I have to do that with an ftp program. This website has been up since last December, and I am pretty sure that I announced its existence. But I haven't gotten any requests to change it, or seen any indication that it has been used at all. Please feel free to load up the tech file and concoct any changes you desire. The program automatically changes the filename of all modified files by adding "New_", just change that to your name and I can see all the changes that anyone made. I made that web-based GUI for others to use, so you wouldn't have to depend on me to make changes you want.

http://paws.wcu.edu/RB22393/xml/Editor_Start.asp

[Richard Bruns]

I changed global upkeep to zero and added the metallurgy tech as a recipient to the production activity. It took me about two minutes, and I did it using only the online tech editor. As far as I can tell, the "tech" text file that came with the testbed stuff is the same as the default that loads into the online editor.

[Mark_Everson]

Hey Richard:

Originally posted by Richard Bruns
I changed global upkeep to zero and added the metallurgy tech as a recipient to the production activity. It took me about two minutes, and I did it using only the online tech editor. As far as I can tell, the "tech" text file that came with the testbed stuff is the same as the default that loads into the online editor.

Great, thanks, I've downloaded it, but am too fuzzy right now to check it out.

I believe that I have said that the tech model is a tool for other model designers to use as they see fit. That's why I built the online tech xml file builder. It allows any team member to view, modify, and add techs to the tech tree. (snip) I made that web-based GUI for others to use, so you wouldn't have to depend on me to make changes you want.

I just checked out the Tech_Draft.xml file, pretty cool! The reason I haven't used the xml builder to this point is that the tech model is really complicated and has many interdependencies between the numbers that need to be observed, or things can run away... If I absolutely needed to, I'm sure I could figure those things out, but it would take some serious time. And you're already the resident expert on what you can and can't do with the model! If we do the requests to you, you can balance things as new items are added (FE if one tech becomes a helper for many things it may need to be made more expensive). So I at least am likely to leave it to you, for fear that I would screw things up

Cya,

Mark

[Mark_Everson]

Hey Richard, something I missed in the last post...

Originally posted by Richard Bruns
The current "production" activity feeds into the production tech only. Do you want me to add an additional Manufacturing activity or simply change it so it also feeds the Metallurgy tech as well?

I'd prefer that the current production activity be renamed to Manufacturing (making it easier to tell activities from techs) and that the newly-named Manufacturing put its points into production and metallurgy as outlined above.

Thanks,

Mark

BTW my initial tests are showing that the Farming tech improves very rapidly (when it has enough RPs to improve at all). This is partly because for now Food RPs are generated proportional to population, and as Farming tech grows, so does the population, who become mostly farmers, giving a feedback cycle that explodes.

Part of it is that Food feeds not only Farming, but also Farming's only helper, Biology. If Biology weren't improved along with farming I'd think diminshing returns might set in. It seems that one thing that would help stability is that helpers be not fed by the Activity that a tech passes RPs to. Of course that is somewhat contrary to what would be desired in a number of cases. We'll try to get you a version with the tech stuff hooking up to some military and econ soon. Then you'll be able to play with the parameters yourself, and see what happens, and what can improve undesired behavior.

Cya,

Mark

[Mark_Everson]

I've done some more tests with upkeep set to 0. The ones above were with the old upkeep values. (in my technology.xml file the old value for Farming upkeep was set to 1.0! which was likely the cause of needing large numbers of RPs for the tech to do anything at all...)

The immediate problem I see, and this is a gameplay problem, not necessarily a tech one per se, is the difference between a 1-square civ and a large one. Dawn has an initial population of 5k, and Rome in the other scenario has about 2M. For any number of RP generation that has Rome gaining Farming tech at a reasonable rate (maybe .1/turn) the Dawn scenario does absolutely nothing in tech. Not surprising given the math.

Of course one issue is that right now both scenarios start with the same tech levels. Fixing this may put things somewhat back on a more reasonable track. Improving stone age tech by 1 should be Much easier than improving that of classical antiquity. Another factor we could emphasize in dawn is that you really need to expand your population by a factor of 10 or so before expecting to see much of any improvement in tech.

Sorry for the stream-of-consciousness nature of these reports, but I'm trying to squeeze in what I can at work to push this along...

[Mark_Everson]

Anyone know... does the upkeep tag on an individual technology override the global upkeep value? If so we have a lot of techs with upkeep = 1.

Any problem with me setting all those to 0.0, or might I cause global destruction I have set them all to 0.0 for the time being.

Richard, there was a change that was required by the parser that I think Gary mentioned a while ago. You need to change your xml generator so that what used to be called activity "effect" tags are instead a "proportion" tag and the proportions have to be on a scale 0-1.0 so what used to be an effect 70 is now a proportion 0.7 .

I have copied over the Military Activity to the xml file with Laurent and my most recent mods. What does the global multiplier tag, and the same one on an activity do? I have commented them out for now, since it is not recognized by the parser. Many of the applications you had , Richard, are not in the current xml. I have not put them in since it might cause Laurent grief. I will email it out when I'm done with the mods, and have a chance to test it a bit.

The Simple_Tech_Test output file was time-consuming to work from since the xml codes are all run together. I remember that you didn't have a good fix for that, Richard, when it came up before. I guess maybe we'll just ask your opinions on values, and add things as we go along by hand. Is there any way to put characters into the file that we could then replace with new-line codes and spaces so that the output is easily formattable like the xml we're currently working from?

[LDiCesare]

Anyone know... does the upkeep tag on an individual technology override the global upkeep value? If so we have a lot of techs with upkeep = 1.

In the code, it does overwrite the default value. (I did that for military tactics and the effect was clear).

[Mark_Everson]

Originally posted by Mark_Everson
The immediate problem I see, and this is a gameplay problem, not necessarily a tech one per se, is the difference between a 1-square civ and a large one. Dawn has an initial population of 5k, and Rome in the other scenario has about 2M. For any number of RP generation that has Rome gaining Farming tech at a reasonable rate (maybe .1/turn) the Dawn scenario does absolutely nothing in tech. Not surprising given the math.

Of course one issue is that right now both scenarios start with the same tech levels. Fixing this may put things somewhat back on a more reasonable track. Improving stone age tech by 1 should be Much easier than improving that of classical antiquity. Another factor we could emphasize in dawn is that you really need to expand your population by a factor of 10 or so before expecting to see much of any improvement in tech.

I don't think I've ever used "respond with quote" on one of my own posts! But it was the easiest way...

I think the easiest way to deal with the 5k vs 5M population issue for now would be to add a number like the old civ "tech paradigm" Perhaps that is what the "multiplier" Richard had on activities was... A global multiplier (or looking forward to difficulty levels, even one done by civ) seems to me the easiest way to balance things so that people playing Dawn (5k) can see Some tech progress without the larger-pop scenario racing up in techs.

So long as the scenario can set the multiplier we can balance things out just fine. What do you guys think?