Print

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

[Footkerchief]

Hey! well it's my first time to post here on forums, but I've been playing DF for a long time now.

This project sounds amazing, I wonder if I could make sprites for the creatures? \o/
Maybe it's useless for so many critters, but I'm up for it. I've done some art stuff for other games like for CIV3 warhammer mod

I think we're open to this, although I'm not sure how sprites work in this respect.  Does anyone know if we can have a custom sprite for each and every critter when the new release comes out?  That would be amazing.

Vermin are still stuck with ASCII tiles, but that should be remedied in the nearish future when we get full graphics support (#4 on Eternal Suggestion Voting), so time spent making vermin sprites won't be wasted.  The one graphics upgrade we're getting in the upcoming version is caste support -- you can now have separate sprites for male and female creatures. 

Anyway, yeah, it would be really awesome if this project had creature graphics.  Feel free to tackle it however interests you, but my advice would be to start by making generic graphics -- a generic squid, a generic antelope, a generic crab, a generic lizard, and so on.

Firstly, we should start coming up with prefstrings for the entries, since we have info on each creature and that will not be a copy and paste affair. 

Are we going to want to put prefstrings in the creature templates as well?  That is, would we want prefstrings that apply to whole groups of creatures as well as specific creatures?  I guess it'd be sort of weird to like a specific creature for a generic reason, which highlights the shortcomings of the prefstring placeholder -- dwarves only like specific creatures, not groups of related creatures, so you inevitably have weird situations where a dwarf loves green bottle flies for their iridescence and hates blue bottle flies for their nasty habits.

Secondly, we should probably start doing some research to figure out how to balance the frequencies for animals.  Maybe start by defining, for each biome, the percentage of creatures that should fall into each class, and then determine the percent distribution of each order in each class, etc.  Once these ratios are established, it would be fairly simple to come up with appropriate frequencies for each creature in a spreadsheet.  Maybe for the sake of realism, size should be accounted for so the ratios would be biomass, rather than number of creatures.

Honestly I don't know if it's worth tackling the frequency issue yet.  The FREQUENCY tag is an extremely blunt tool -- it doesn't let you do anything biome-specific, for instance -- and without knowing what tools it'll be replaced by (and when), we run a very large risk of wasting our time.  Frequency is just one of many, many issues that simply won't work well for a while.  Symbiosis and aquatic creatures are two other big ones.

-Leptodirus hochenwartii (Amber Cave Beetle): http://en.wikipedia.org/wiki/File:Leptodirus_hochenwartii.jpg
--Including being one of the few true cave dwelling creatures. It also looks like it is made of Honey/Amber. Looks like an Ant though.

I am entirely for a better name then Amber Cave Beetle... though the only one I thought of is "Amber Statuette"

"Amber cave beetle" sounds great and I can't see any reason it shouldn't be included.  Throw it in!  We already have a generic representative of its family, the "round fungus beetle," which you can keep or remove at your discretion.

[Pizdzius]

Vermin are still stuck with ASCII tiles, but that should be remedied in the nearish future when we get full graphics support (#4 on Eternal Suggestion Voting), so time spent making vermin sprites won't be wasted.  The one graphics upgrade we're getting in the upcoming version is caste support -- you can now have separate sprites for male and female creatures. 

Anyway, yeah, it would be really awesome if this project had creature graphics.  Feel free to tackle it however interests you, but my advice would be to start by making generic graphics -- a generic squid, a generic antelope, a generic crab, a generic lizard, and so on.

nice, I didn't know about vermins, it'd be great. So, getting back to generic creatures, where should I aim?
Size, I use the good ol' 16x16, but I know people prefer the bigger tilesets. Give me a hint on that, should I make two options or just resize and don't waste time?
Also, I don't know if the creature string will work in the same way. Default, child, zombie, skeleton etc.

[Footkerchief]

As I have thought more about this, I'd like to offer my services for spreadsheet/database management for this project, if it is deemed necessary.  We already have a lot of creatures listed with links to descriptions, but we could also start adding info on biomes, average weight, lifespan, prefstrings, flavor text, etc to a spreadsheet or database which could be on google docs or something similar.  I'll start putting something together today if people sound interested.

I've been mulling over the idea of a creature database or whatever too.  The main thing I want to avoid in any such approach is having our own data storage format that requires manual translation into raw entries.  The approach I keep coming back to is this: one table for creature templates, one table for creatures with foreign keys (or a join table, possibly) to the template table representing template invocations at the beginning of the creature entry.  The tables would each have a column for the actual text of their raw definitions, plus additional columns for project metadata.  (There would probably be another table for folder-like nested taxa to which creatures could belong.)  The reason I'm normalizing templates and not other stuff like biomes is that a) a fully normalized database would be a horrific mess, and b) I'm expecting templates to be by far the most critical tool at our disposal.

That was a bit of a digression.  My direct thoughts on what you said are this:

• Having a wiki and a spreadsheet at the same time is bad.  Maintaining them against one another will be hard -- even if you try to automatically update the spreadsheet with a parser script, it'll get stymied by name changes etc.  It has to be a one-time switch.
• The spreadsheet needs hierarchical organization to arbitrary depth, like the wiki has.  That's the only thing keeping our current list readable.  I don't think I've ever seen a spreadsheet like that, but if it can be done, then I'm cool with switching.
• If such a spreadsheet isn't possible, then we can think about putting raws snippets in the wiki itself as an interim solution.
• Whether we're putting raws snippets in a spreadsheet or the wiki, they'll be problematic until we get our basic templates ironed out, because people will include code that ends up being redundant with the templates and so on.  So, I keep coming back to the need to get templates ironed out ASAP.  But we still haven't seen how a creature variation works.

I guess I need to bug Toady about that more directly.  Or maybe you could, Rainseeker?  You seem to get good results.

[Rainseeker]

What specifically did you want to ask him?  I could ask him to just look at this thread if you want.  Maybe highlight in limeGreen the stuff you would like Toady to help with?

[Footkerchief]

What specifically did you want to ask him?

We could really use a preview of the creature variations.  I'm PMing him about it now.

[Zantan]

• Having a wiki and a spreadsheet at the same time is bad.  Maintaining them against one another will be hard -- even if you try to automatically update the spreadsheet with a parser script, it'll get stymied by name changes etc.  It has to be a one-time switch.

Yes, that would be way too much work.  At the moment, it would be best to create some kind of template and decide what additional info to add to creatures, and make a one time conversion once we have a sufficient template.  However, in the mean time couldn't we replace all the lists in the wiki with tables, and slowly add a little information to them?  For instance, turn the list of herbivorous marsupials into a table, and do the same with the bovids, cats, etc.?

• Whether we're putting raws snippets in a spreadsheet or the wiki, they'll be problematic until we get our basic templates ironed out, because people will include code that ends up being redundant with the templates and so on.  So, I keep coming back to the need to get templates ironed out ASAP.  But we still haven't seen how a creature variation works.

It's not clear how much redundant tags would matter...that is a good question for Toady.

[Footkerchief]

However, in the mean time couldn't we replace all the lists in the wiki with tables, and slowly add a little information to them?  For instance, turn the list of herbivorous marsupials into a table, and do the same with the bovids, cats, etc.?

Yeah, that might be a sensible compromise for people who are anxious to start adding prefstrings etc.

It's not clear how much redundant tags would matter...that is a good question for Toady.

Well, it shouldn't matter to the game (which always handles redundant tags gracefully, as far as I know).  It's just a problem for us because it would violate Don't Repeat Yourself and make the creatures hard to maintain.

[Zantan]

Well, it shouldn't matter to the game (which always handles redundant tags gracefully, as far as I know).  It's just a problem for us because it would violate Don't Repeat Yourself and make the creatures hard to maintain.

If this is a serious concern, we could assume a certain amount of redundancy, and code for that.  This should be relatively simple, since Toady said that templates are able to both add and remove tags.

However, in the mean time couldn't we replace all the lists in the wiki with tables, and slowly add a little information to them?  For instance, turn the list of herbivorous marsupials into a table, and do the same with the bovids, cats, etc.?

Yeah, that might be a sensible compromise for people who are anxious to start adding prefstrings etc.

I could set up an account on the wiki and start doing this, but I'd want Rainseeker's approval as well, since he seems to have some degree of ownership of this project.

[Rainseeker]

I only claim birth rights to this project, and spiritual leadership... vision, in other words.  As far as the technical nitty-gritty is concerned, I want Footkerchief to take the reigns for sure and make those decisions.  He knows what he's talking about, I'll trust his judgment to work out any technical issues with the volunteer force as a whole.

[Zantan]

Okay, so what are good things to be filling in now?  Biomes, prefstring, flavor text, weight (does anyone know the relationship between weight and size in the next version?), maybe littersize and maturation time...anything else?

[Neonivek]

What specifically did you want to ask him?  I could ask him to just look at this thread if you want.  Maybe highlight in limeGreen the stuff you would like Toady to help with?

I was personally going to wait until after the next release before making the creatures... because there is just so much I need to know and be able to compare.

For example in order to do sizes (unless they match real life sizes) I need to see the differences in creatures. In order to do poisons I need to know the difference between deadly and annoying.

Lets just sum it up as "I need to have my hands on the raws"

The only REAL question I have for Toady right now isn't needed to complete the project.

[Footkerchief]

Toady PM'd me back about creature variations.  Basically, they're awesome and we can start making templates right now if we want.  This is important reading for anyone who wants to understand how we'll be doing the creature raws:

Spoiler: Toady's reply (click to show/hide)

1) What do they look like?  Can we see a preview?

Here's the current animal person one.  Removed tags need to be added afterward if you want the information:

Code: [Select]

[CREATURE_VARIATION:ANIMAL_PERSON]
[CV_REMOVE_TAG:NAME]
[CV_REMOVE_TAG:GENERAL_CHILD_NAME]
[CV_REMOVE_TAG:GENERAL_BABY_NAME]
[CV_REMOVE_TAG:CASTE_NAME]
[CV_REMOVE_TAG:CHILDNAME]
[CV_REMOVE_TAG:BABYNAME]
[CV_REMOVE_TAG:SMALL_REMAINS]
[CV_REMOVE_TAG:DESCRIPTION]
[CV_REMOVE_TAG:PETVALUE]
[CV_REMOVE_TAG:PENETRATEPOWER]
[CV_REMOVE_TAG:VERMIN_EATER]
[CV_REMOVE_TAG:VERMIN_HATEABLE]
[CV_REMOVE_TAG:VERMIN_GROUNDER]
[CV_REMOVE_TAG:VERMIN_FISH]
[CV_REMOVE_TAG:VERMIN_SOIL]
[CV_REMOVE_TAG:VERMIN_SOIL_COLONY]
[CV_REMOVE_TAG:VERMIN_ROTTER]
[CV_REMOVE_TAG:TRIGGERABLE_GROUP]
[CV_REMOVE_TAG:PET_EXOTIC]
[CV_REMOVE_TAG:NOT_BUTCHERABLE]
[CV_REMOVE_TAG:SPEED]
[CV_REMOVE_TAG:SWIM_SPEED]
[CV_REMOVE_TAG:MUNDANE]
[CV_CONVERT_TAG]
[CVCT_MASTER:BODY]
[CVCT_TARGET:QUADRUPED]
[CVCT_REPLACEMENT:HUMANOID]
[CV_NEW_TAG:LARGE_ROAMING]
[CV_NEW_TAG:SAVAGE]
[CV_NEW_TAG:SELECT_CASTE:ALL]
[CV_NEW_TAG:GRAVITATE_BODY_SIZE:70000]
[CV_NEW_TAG:CAN_LEARN]
[CV_NEW_TAG:CAN_SPEAK]

And here it is being used to make a cave fish man -- the body stuff doesn't work there because the original fish is not a quadruped, so it uses a custom variation on the spot, which is processed and executed right in there:

Code: [Select]

[CREATURE:CAVE_FISH_MAN]
[COPY_TAGS_FROM:FISH_CAVE]
[APPLY_CREATURE_VARIATION:ANIMAL_PERSON]
[CV_REMOVE_TAG:AQUATIC]
[CV_ADD_TAG:AMPHIBIOUS]
[APPLY_CURRENT_CREATURE_VARIATION]
[GO_TO_TAG:BODY]
[CV_REMOVE_TAG:BODY]
[APPLY_CURRENT_CREATURE_VARIATION]
[BODY:HUMANOID:SIDE_FINS:DORSAL_FIN:TAIL:HEART:GUTS:ORGANS:NECK:SPINE:BRAIN:SKULL:MOUTH:RIBCAGE]
[GO_TO_END]
[SELECT_CASTE:MALE]
[CASTE_NAME:cave fish man:cave fish men:cave fish man]
[SELECT_CASTE:FEMALE]
[CASTE_NAME:cave fish woman:cave fish women:cave fish woman]
[GO_TO_START]
[NAME:cave fish man:cave fish men:cave fish man]
[DESCRIPTION:A humanoid creature with the head of a blind cave fish.  They live in colonies in watery regions far underground.]
[POPULATION_NUMBER:30:50]
[CLUSTER_NUMBER:5:10]

The [APPLY_CURRENT_CREATURE_VARIATION] runs all of the custom commands you've saved up so far (variations from files are run immediately).  There's also the strange go to stuff in there -- some tags need to be before other tags, so this was a quick fix.  The proper way might be to have it understand the requirements a little better and collect the strings into ordered groups that are invisible to the user or something, but there will be one annoyance or other either way.

2) Will sequentially invoked variations basically act like a nested template?  E.g., we'd want separate templates ARTHROPOD and CRUSTACEAN, where a creature invokes them in that order and CRUSTACEAN assumes that ARTHROPOD has already been invoked and modifies the ARTHROPOD stuff.

Yeah, it applies things as it goes to a list of strings floating around in memory, so you can nest them by applying them in sequence.  Following their evolutionary hierarchies may or may not be a good idea depending on how much the bodies morph.  Bodies are one of the more irritating things, since it's all one giant tag.  You can change and insert things within a tag, but you kind of have to keep track yourself what it looks like at any given stage.

3) Is there any way for a creature entry to deactivate a tag (like DIURNAL) applied by a template?  I guess you could create a special variation just for deactivating that one tag...

Yeah, you can see this in the fish man when it changes AQUATIC to AMPHIBIOUS.  For yours, you'd just do [CV_REMOVE_TAG:DIURNAL] followed by, say, [CV_ADD_TAG:ALL_ACTIVE] then [APPLY_CURRENT_CREATURE_VARIATION], which hands those two changes then clears the queue.  You don't need to add a replacement if it's not required.

4) Any other gotchas with the variations?

I'm sure there are...  the invisible cursor/go-to stuff is probably the most annoying.  CV_NEW_TAG and CV_ADD_TAG are the same and they both work.  I just got confused halfway through.

Okay, so what are good things to be filling in now?  Biomes, prefstring, flavor text, weight (does anyone know the relationship between weight and size in the next version?), maybe littersize and maturation time...anything else?

Those all sound good, plus coloration, diurnal/nocturnal/crepuscular/ALL_ACTIVE stuff, and maybe how long it takes to mature and die of old age.  Many of these won't be necessary depending on what ends up in the templates -- for example, if all squid have about the same maturation time (I have no idea if this is true or not), we can put the maturation time in the template and omit it from all the individual squid entries, which will massively cut down on clutter.  Likewise if the majority of flies are diurnal, and so on.

In terms of organization -- right now I'm leaning toward using the wiki for everything except actual generation of the final raws (which we can do by feeding the wiki's source text into a simple program).  We can attach templates to certain taxa just like we'll attach raws to creatures, and the creatures will be assumed to invoke all their parents' templates in order (creatures that don't want to use those templates can contain a special flag for the raws generator, like NO_TEMPLATES).  Still have to determine how we'll actually put those raws on the wiki pages. 

In other news, WP has a neat list of cave animals.

[Gorobay]

2) Will sequentially invoked variations basically act like a nested template?  E.g., we'd want separate templates ARTHROPOD and CRUSTACEAN, where a creature invokes them in that order and CRUSTACEAN assumes that ARTHROPOD has already been invoked and modifies the ARTHROPOD stuff.

For templates that are subtemplates of others, is it possible to do:

Code: [Select]

[CREATURE_VARIATION:ARTHROPOD]
(Do all the arthropod stuff here.)

[CREATURE_VARIATION:CRUSTACEAN]
[CV_NEW_TAG:APPLY_CREATURE_VARIATION:ARTHROPOD]
(Do all the crustacean stuff here.)

[CREATURE:CRAB_SPINY]
[APPLY_CREATURE_VARIATION:CRUSTACEAN]
(Do all the spiny crab stuff here.)
If that works, modifying the ARTHROPOD stuff in CRAB_SPINY would be guaranteed to work, instead of relying on the writer of the raws remembering to invoke ARTHROPOD. Basically it would remove a possible source of error.

Unrelatedly, is there any reason tardigrades are still on the list?

[Footkerchief]

^^^ My guess is no, because the dev log about creature variations said "it'll probably have to be augmented to make it more powerful down the line (nesting templates would be very useful)," implying we can't do that kind of nesting yet.  Fortunately sequential invocation is almost as good.

Do you have any thoughts about using the wiki as our raws repository like I described above?

[Zantan]

for example, if all squid have about the same maturation time (I have no idea if this is true or not), we can put the maturation time in the template and omit it from all the individual squid entries, which will massively cut down on clutter. 

Are we sure that aquatic life should be breeding right now?  Especially on lakes or seas, where it's relatively hard to hunt sea creatures, this could lead to an unwanted explosion of sealife over time.