Print

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

[PTTG??]

The really funny part of all this is how people have been gushing for days about how cool it would be to use the music descriptions to auto-generate DF music with software or by hand... Then Toady gives us descriptions with exactly the amount and type of information that would needed to actually start to be able to do it - and people complain about "too much detail".

Point of order, they aren't the same people.

[Miuramir]

Thoughts on the music... overall, this is awesome and incredible, but very much a work in progress  

* This is probably in progress, but a quick reminder note of what general type of instrument each is would be good.  Even the very basic info ("reed aerophone") would give a starting idea, and a bit more detail would be good (e.g. "conical bore double-reed aerophone in an upper register" would let us go "well, it's sort of vaguely oboe-ish".) 

* I am concerned that the rhythms described appear to be disjunct from the meter of the poetry (where present).  I would think that if you're dealing with writing music to a poem, there would be some sort of routine that would give it decent odds of having the rhythm(s) based on variants of multiples or divisors of the stress of the poetic meter.  (Or, conversely, in cases where the music comes first, create a poetic form that lines up well with it.) 

* Toady mentioned something about improvements to vocals, but one point is that there doesn't seem to be a way to readily describe certain sorts of actual choral music in the current structure. 

** I think there needs to be a "chorus" block that represents one-or-more singers doing a task.  E.g. you might try and describe something typically Western as "A singer in a high register sometimes does the main melody and sometimes is silent; a chorus in a middle register sometimes does the main melody and sometimes provides harmony; a chorus in a lower register provides harmony and rhythm, and should accent the beat."

** As alluded above, "silent" needs to be an option, for pieces that have a soloist for some parts but not a lead vocalist (or instrumentalist) for the whole thing. 

* There doesn't seem to be a way to represent non-vocal, non-instrumental human noises, which are a big part of various musical styles (particularly early ones).  Hand clapping, finger snapping, foot stomping, and so on should be in as rhythm choices at least. 

* The rhythms tend to be extremely complex, and involve long, eccentric patterns.  I'm not saying these are beyond examples of human music, because there's some pretty weird stuff out there... but statistically speaking, I think the generator needs to be tuned to prefer forms that are comprehensible and performable by ordinary folks a bit more.  In particular for styles and uses that in-game are supposed to be used by ordinary dwarves. 

* Similarly, "common folk" songs should probably dial back on the ornaments a bit.  We're getting a lot of trills, grace notes, arpeggios, etc. for someone who's not a professional opera singer equivalent. 

* Historical music that changes *scale systems* (not just key) in mid-song is pretty rare as I understand it.  Among other things, before valves and modern keyboard instruments it is difficult to produce instruments that even do multiple keys, and multiple scale systems in the same song would be demanding and rare.  A society that relies primarily on vocals and/or true chromatic instruments might be an exception.  In our world, IIRC that would primarily be the early organ variants, such as the ancient Greek hydraulis. 

[Footkerchief]

And perhaps one day Dwarf Fortress itself will be able to generate .midis from these details. Gotta plan ahead for these things.

Given the precision of the chords and rhythms, it looks pretty damn close.  Arrange the chords into some kind of progression, arpeggiate them as per the beat, and run them through placeholder instruments... I think that would do it?

It would be a shame to have all those words and no sound.

[Robsoie]

Complexity is only good if it is meaningful. If it is only complexity for complexity's sake, and not for gameplay's/story's sake, it becomes a chore.

Agreed 100% , when it serves the storytelling, it's great, when it's not, it's just overkill for complexity sake.

By example, there are descriptions in current DF that are lacking and do not live up to the potential of their purpose in term of story telling, so much they would be more worth than music when it comes to story telling .

In one of my most long 40.x adventure, i managed to get into a necromancer tower, and found 2 books that sounded like they had some actual storytelling worth :



But that was all for the description, only "an appetizer" instead of actively telling me a summary of what this was really about.
So i had to cheat and check the legends by fakely-retiring my guy, so i can access the legend mode and finally cross the various information of it to actually figure out what was going on with these intriguing book subjects.

And see that the elf in question was an herbalist that wandered from elven places to other elven places and finally became the druid of his last destination, still wandering he defeated a goblin but on one of his travel was then killed by a night creature.

The corpse of that druid was found by the necromancer author of the book and was then raised to undeath, and checking at the dates, it really looks like that was the 1st named undead the necromancer created, making the creature of some importance to that necromancer, as i can imagine for story telling purpose this was his first success and display of power, a milestone in his life.

Basically lots of interesting story telling stuff ... the kind of description that should be in the description of the books themselves instead of forcing you to retire/save cheat and go check the legends, because they're meaningfull in understanding the characters stories, and who the necromancer is (well was due to the bugs) .

[Knight Otu]

* This is probably in progress, but a quick reminder note of what general type of instrument each is would be good.  Even the very basic info ("reed aerophone") would give a starting idea, and a bit more detail would be good (e.g. "conical bore double-reed aerophone in an upper register" would let us go "well, it's sort of vaguely oboe-ish".) 

Yeah, Toady specifically notes that the instruments will be described in the forms in the final version of the descriptions, and that more general instrument requirements (I guess in the form of "a metallic wind instrument") will become possible as well.

* Similarly, "common folk" songs should probably dial back on the ornaments a bit.  We're getting a lot of trills, grace notes, arpeggios, etc. for someone who's not a professional opera singer equivalent. 

I'm guessing that's where the note comes into play that Toady only took one musical form from each civilization, and that some forms of civilizations are closer to the Bald Vessel form in complexity.

[Appelgren]

I enjoyed figuring out the cross-rhythms/polyrhythms  (I'm a huge Captain Beefheart fan).
But I can't wrap my head around the unusual scale systems (like dividing the octave into 24 notes and then building random tetrachords).  It comes dangerously close to being"grey goop" for me. Or intriguingly alien but ultimately samey goop. Adding some familiar scales/modes to the mix would make the random scale systems more interesting to me.. Just like forgotten beasts seem more special in a world where theres also ordinary bears, cats and capybaras etc.

[Man In Zero G]

Point of order, they aren't the same people.

Didn't mean to imply that they were, it was meant as further illustration of my point: What's important to some players is not to others. The technical parts are important to the people who want to try to actually make this music, but not to the people who don't care about it.
The "funny" part of it is: these details, like practically all cosmetic details in DF, are something we're likely to have to go into a separate screen to look for. In other words, the presence or absence or these details will not negatively affect the people who don't want to see them beyond "I'm not going to press that key... but if I do, I won't scroll down". Whereas their absence would negatively impact the people who're jazzed about actually making this music, and their presence can only enhance these player's experience. If the removal of a detail will not affect your game experience, but will harm another player's, why ask for it? The complaints about the quantity or complexity of the detail seem kind of silly in that light.

All that said, sure, the wording and descriptions could be clearer. Absolutely. And that can be done without losing the level of detail.
It's also the first draft he's shown and it even says "subject to change!" right there at the top. This stuff is not exactly how we're gonna see it when it gets to us. It's still being refined.

[thvaz]

I thought the poem rules were unnecessary until people started doing awesome poems with them. Maybe something like that can happen with these song rules?

[Footkerchief]

I enjoyed figuring out the cross-rhythms/polyrhythms  (I'm a huge Captain Beefheart fan).
But I can't wrap my head around the unusual scale systems (like dividing the octave into 24 notes and then building random tetrachords).  It comes dangerously close to being"grey goop" for me. Or intriguingly alien but ultimately samey goop. Adding some familiar scales/modes to the mix would make the random scale systems more interesting to me.. Just like forgotten beasts seem more special in a world where theres also ordinary bears, cats and capybaras etc.

It seems to have solid precedent.

Also, this one looks fairly conventional:

The uto hexatonic scale is thought of as two disjoint chords spanning two perfect fourths. These chords are named ngub and ozse.

The ngub trichord is the 1st, the 5th and the 6th degrees of the semitone octave scale.

The ozse tetrachord is the 8th, the 10th, the 11th and the 13th (completing the octave) degrees of the semitone octave scale.

[WordsandChaos]

These song descriptions are sodding great. Don't add anything to the gameplay?

The Elder Scrolls series has a hundred odd books that didn't add anything to the game. Still loved the fact they were there. Diablo II and III had a bunch of superfluous dialogue options/random bits of lore that gave backstory and info into the world and didn't add anything to the game. Still loved the fact they were there. The Lord of the Rings, Song of Ice and Fire, Warhammer, Snow Crash, Moby Dick, and thousands of other bits of media include other bits of 'fluff' that don't directly impact the main focus of the game/book/experience, but would make each thing more shallow if those bits of lore and fluff, etc, were missing.

Here's the fantastic thing about all of it: it's optional. It's not shoved into your face. You have to actively make an effort to go and look at it, and you can ignore it if you're not interested. And if you aren't that's fine. Not everyone read the pages and pages concerning who was wearing what suit made by which company in American Psycho. Not everyone read any of the books in the Elder Scrolls. Not everyone gave a damn about the pages upon pages of derail on whatever topic in Snow crash. The point is that some people were interested and did. This is why the Silmarillion, the World of Ice and Fire book, or the Black Library exists:

Because some like to delve deeper.

[Toady One]

Thanks to Putnam, Knight Otu, Dirst, Cruxador, Footkerchief, Man in Zero G, Vattic and anybody I missed for helping to answer questions!  Thanks also for the feedback on the poetry and music generators I've been working on.  Hopefully it'll be vaguely satisfying for the most part by the time we get a version up.

Any plans to extend the 256 ASCII-Tiles in the future? Probably to 512 or even unlimited tiles for optical differences. As there are so many tiles having multi-uses, it is quite confusing sometimes. I am not referring to adding graphics, just implementing other tiles not included in the ASCII e.g. horizontal brackets/letters or something like that. Same regarding colours. Any plans to increase them from 16 to e.g. 256 colours or even 24-bit? Some parts of the code like the color-definiton tokens seem to support them already. Or will that cause major recoding?
...
I am not referring to graphics, which for sure will cause major and long time dedicated work. Just extending it from 256 tiles to some more. Additional tiles could be derived from e.g. UNICODE. I am no specialist regarding copyright. But as most of the UNICODE-tiles are from various recent/historic alphabets, I can't believe they are more restricted than ASCII.

Besides, moving to Unicode would require rewriting a large portion of the display code that relies on 8-bit display tiles.

Even a basic addition of another character page is not a rewrite I feel comfortable doing.  The DF part is fine -- it'd take a while, but I could do the DF side of it once I knew what data to push where.  However, I have no idea how to set up additional texture atlases while still keeping the speed up in OpenGL, assuming it wouldn't require a completely different approach, so I can't really start.  It's way on the back burner for this reason.  The game could probably use another thousand characters at this point to remove fundamental overlaps.  That's potentially independent of the item graphics problem, though they both run into the texture atlas issue (if you want more than a few item pictures).  Item pictures are also complicated by material considerations -- those aren't prohibitive, but take us into graphics territory that I don't want to get mired in.

If temples are affected by material wealth within its zones and number of worshippers will there be a chance for positive blessings such as attribute/skill boost? Or is this related to artifacts? Also will dieties bless the fort with materilas/objects such as the procedurally genereated ones, will certain deities be more likely to gift certain materials eg. god of gems gives some rare gems?

I don't have a specific positive effects plan in place right now.  It's kind of lumped in with magic.

You mentioned that we might get the occasional Human elf or goblin on our doorsteps, will your fort be a possible destination for refugees?

The obstacle isn't the multi-species fort stuff so much as the number of people for that one, which is more of a hill-dwarf problem.  If I remember, the refugees can already choose your fort as a destination for camping next to, but they don't actually enter the fort.  It'll probably be best to put that one off until we have the broader infrastructure in place for hill-dwarves (which are just a group of numerous associated critters living near the fort).

They walk over the land on a path avoiding water

Does this mean that (invaders) avoid rivers? If so, will they simply give up if the fort is surrounded by rivers, or will they just go right over?

Ah, yeah, I should have been less vague -- Knight Otu clarified that I meant oceans and lakes.  The bridging/fording problem is being ignored until we dig in a bit.  Overall the regional pathfinding should be a lot easier than the fort pathing.  Even though we don't have the map at every resolution away from the play area, it doesn't change that often and the effect of bridges etc. can be tracked.  It already knows river width everywhere.

Are temple effects on immigration on the table for the tavern release? What about pilgrimages? By immigration effects, I mean things like getting more worshippers of a certain god if you have a big temple to it, or more devout worshipers if you have awesome temples in general. It seems like it would be more crucial for start scenarios where making a temple to a god is your primary goal, but on the other hand it could be a relatively low-hanging fruit. Pilgrimages on the other hand, fit quite perfectly with the "temporary visitor" idea being dealt with in the tavern release, but if they're to be relic-oriented like medieval Christian pilgrimages, it might make more sense to wait until some more artifact stuff is done.

It's possible -- as you say, it's a low-hanging fruit.  Those are numerous though, now, with all of the additions.  I'd put even odds on something basic to put some meat on the new temple locations, though they won't achieve parity with taverns yet.

Will there ever be magical songs to control your enemies and bolster your allies ala D&D Bards?

As with other "ever" questions, it's really hard to say, and magic is sort of an amorphous blob in the future, but we're one step closer to that now by actually having songs, anyway, he he he.  Magic will certainly profit by having all of these different threads to work with when we get there, and it'll be a fun challenge to generate intricate and interconnected systems that use the elements at our disposal.

Will military squad orders be updated along with the new justice system. For example, will the kill order be replaced with an attack order, where you can set the terms of engagement for your troops. The options would be something like, non-lethal, lethal, and no-quarter, so that you could target your own dwarfs without loyalty cascades and maybe attempt to capture invaders like goblins or kobolds. I was thinking being able to quell riots and beat dwarves would be fun along the lines of running a totalitarian fort.

Will there be prison zones, so you could have rooms designated for prisoners, like holding cells and cells. Would this be considered as part of the prison colony scenario? Will we be able to shackle, torture, and restrain prisoners based on civ ethics?

I don't anticipate doing anything with justice/military for this time, depending on how drunks work out, but when we get deeper into the law stuff we'll see some changes.  It's too far off to speculate exactly which ones.

With the look into the multi tile critters and the pathfinding changes that will require, were there any thoughts on tackling other parts of the pathfinding that is a bit lackluster? Like flier pathfinding, or things like one way or grasp_required kinds of pathfinding that came up back when ropes for climbing was mentioned?

And second just in-case that first one does get thrown out.  Will races without alcohol_dependent eventually take offence when visiting your fortress if there is no water whatsoever available, or will they perfectly happy to drink mug after mug like water like our dwarves do?

So far the general pathing experiments have all fizzled, and the only idea on the table for multi-tile pathers is them not being bright about long distances and using all available data for short distances -- like fliers.

I'm really not sure how the foray into drunkenness is going to go yet.  Some responsible parties may draw the line, but our critters have never been good at that.  I'll be happy enough to see the visitors served at all for now.

With the advent of Taverns, will some people who come in be lone/small groups of merchants? Obviously if they start coming in force as soon as you declare a tavern zone, it would mean a constant supply of things for rock goods; but if they only came with enough export profit, it could balance.

With music and such, will some people sell services for goods. Like paying for a performance(s) with trinkets, or for a traveling weaponsmith to work for roughly 10% of what he makes (or some other agreed upon condition)? On that note, if you have a renowned legendary flute-dwarf, can you sell his services for goods?

I read earlier that music can increase the bravery of dwarves, assuming this is true, Will we be able to assign military music dwarves? So we can march in to the beat of the drum British style, or a flute Spartan style?

We won't see little groups of traveling merchants for a while -- there's the hill dwarf side of that (as they visit you as the market center) and the "economy/caravan" side of that.  Neither one is going in for the tavern release.

We are hoping to have some sort of exchange for visitors staying on and doing something for you.  I'm not as sure about farming out your own people.

Military music is important, but sadly neglected so far.  I'm not sure what's going to happen there either.

Now that we will have temples, will we got a way to commision art works of a certain theme in demand? To be able to make
appropiate engravings and statues, without depending of the whims of the rng?

That's still a separate dev item, but it's up on the dev page.  The main minor obstacle is just setting up an interface for it.  Not sure when.

Will there be immigrants or dwarfs that are or choose to be priests, preaching to the people in your temple? How about barmaids/bartenders and wenches?

If we have wenches, can we also have... male wenches? Whatever the male equivalent of a wench is.

I don't know the extent to which we'll be messing with the religious stuff.  It won't be the focus until later.  We should see a number of things going on in the tavern, but they are interesting new jobs that don't really have analogs elsewhere in the fortress -- somewhere between the artisan/labor professions and entity position appointments.  I'm still mulling over exactly how they are going to function, be filled and be described.  Certain aspects of it might need to punted for the status/etc. framework, or they might be the vanguard for it.

Are there any plans to allow greater control over surface features in the raws - e.g. defining rock features, day/night cycles, light levels and other climatic conditions, and so forth? Roughly how far down the line might that be? For instance, will modders ever be able to make worlds in which it's always day or always night, there are no mountains, there's a ceiling instead of a sky, the world is divided into chambers, the sun is green, there are four moons, etc. etc.?

Once world generation has any variation in those things, it'll be possible.  I don't have a timeline for any of it.  The test myth generator contains certain implications along these lines, and we'll have to see how that interacts with world generation parameters and other raws when I glue it all together.  In any case, getting interesting fantasy-type world spanning features and other oddness is a goal.  It might be the push I need to get regular canyons in the game, he he he.

Will mugs get a non-trade use with the new taverns?

Will dwarves still drink from barrels?

We're hopeful for mugs and goblets and so on.  A dwarf without easy access to a cup will probably still guzzle like a bear.

relating to the talk of multi-tile creatures do you plan on changing how size influences combat such a giants attack hitting multiple dwarves and hitting their entire body or making it so a warhammer can't instantly cave in the entire skull of a massive creature.

That would be on the menu when we get to the overall change, yeah.

(regarding body size)
Why did you decide to make the raws use a different scale in the text files than in the game?

If I remember, there were 32 bit variables being blown out by multiplications for the largest creatures all over the place and I didn't want to track them all down.

(on reaction categories)
Is there finalized syntax for that yet, or will that have to wait?

It's fixed assuming it doesn't get any more interesting.  In any reaction's def, assign it to a category with [CATEGORY:<token>].

[CATEGORY:<token>]
 [CATEGORY_NAME:<text>]
 [CATEGORY_DESCRIPTION:<text>]
 [CATEGORY_PARENT:<token>]
 [CATEGORY_KEY:<key>]

The subtokens to define the category's details just need to appear below the CATEGORY somewhere once.  The nesting can go as deeply as needed.  It only shows reaction categories in the relevant workshops if your fort's entity has a permitted reaction within the category nesting, which will hopefully avoid some clutter in complicated mods.  We'll have to see what else is necessary as we go.  Things like material assignments linked to categories (to handle stuff like glass coming out into the raws) are elevated to "low-hanging fruit" status now, but aren't in the game.

How do procedurally-generated buildable instruments work? Is there a new [buildable] "token" in the generated raws, or are there additional buildings for each large instrument?

If it is a token, will modders have access?

Can we assign dwarves to play music or do they merry-make only by their own accord?

Instruments can get [PLACED_AS_BUILDING], which allows them to be placed as other furniture.  It hasn't been generalized to other item types, since instruments didn't get turned into tools.

There aren't official musical assignments at this point.  I'm not sure how things will work later when there'll be more structure/roles in place.  Once tavern jobs are more sorted there could be something.

Did i get that right that subassemblies of an instrument have now theyr own material, thus we get a alder violine with steel strings? If so can those decoration-like subassemblies be raw-defined for objects and could we then add properties to them?

They are just for instruments now, as an improvement type which can be used in reactions.  Generalizing to everything was tempting, but too big a fish to fry for now, time-wise.  We still don't have a true multi-material item in the sense that multiple materials are recognized as equal parts of the item.  That's a hard rewrite given the pervasiveness of having a single item material throughout everything.

When, if ever, might pipe organs be on the table? Or are they just to ancillary? Would you want to wait for restructuring of the mechanism system before getting into transferal of air pressure, or what kind of things might hinder this?

There are little hand-held ones in the game (based on the real-life portative organ), and we also currently have double-a-dwarf size building ones (complete with bellows -- though I'm not sure it'll require a second operator yet, kind of depends how dances turn out), but anything requiring multiple tiles will have to wait for the restructuring yeah, and probably more.  It might be tricky to get it to understand how a water flow (for example, in the hydraulic ones) causes pipes to sound, assuming we'd need to pay attention to that level of detail, and I don't know that we'll ever have an air pressure model -- it's potentially very expensive.  With the one-tile building instruments, that's all hand-waved away, which is good for now.

If you need to have learned (or at least heard) a music style, or poetical style, before using it-- how are new styles generated?  Would a new type of strange mood create a style, or are they only created on day one of worldgen?

I haven't done any of the post-day-one stuff yet, so we'll have to see.  They get a number of styles to start, and I figured there would have to be some generated in world gen by bards etc. as it goes, so that they can intermingle (there are already variables for instance to make a poetic form based on a specific song rather than a musical form, and all other permutations -- then you could have a class of poem "sung to the tune of X" and so on).  I'm not sure what'll make it in post-wg in terms of composition and new inventions of art forms.  It's not any more difficult than doing it in world gen, really, but we never get to all of the easy stuff on the first pass.

with multiracial forts coming into play for the next release, will it be possible to define multiracial entities as well?

It'd be a large rewrite, since the "race" variable for entity definitions is scattered all over the place.  I'm going to deal with a lot of those dependencies just to get the multispecies forts running, and we'll see where we are when that work is done.  Good fortune would be a full gutting, but I'm going to skip what I can for time.  I'm not sure what a multiracial raw def would mean in terms of starting pops.  It would be easiest just to split it in half/etc. I guess, though the test myth generator's interweaved stories of various critter origins is a large wrench in this process that might be worth waiting for.

With the possibility of multiracial entities being added in the near future,will it be possible to have interracial marriages between 2 different races and maybe some sort of hybridization as well?(half-elves and half-goblins?)

Vattic brought up the hybrid quote -- it's a difficult problem.  The marriages are as easy as allowing them, but I'm not sure when it'll come up.  Once there are other critters living in your forts, there'll be a bit more pressure in that direction anyway.

How are things looking regarding save compatibility? I'm guessing it's at least tricky to inject the new stuff into existing saves.

Right now, it's still compatible.  It was too difficult to get the old saves to generate new instruments without crashing everything, so they'll be somewhat impoverished.  The dwarves might be able to make completely freeform stuff or nothing at all.  I'll know more when dwarves are using the new forms.  I'm trying to maintain the basic compatibility so I have an easier time with old bug report saves and so on, and there might be some use of the new art stuff in old forts (instrument defs are the hard part, so they might get singing forms, poetry and dance, but I can't promise it).  Other things like taverns/temples/needs etc. should work immediately in old forts, though you won't have as many varied visitors because entire classes of visitors (bards etc.) won't exist in the old worlds.

You stated that the language RAWs don't have enough information yet to actually construct poetry at this time, like for stressed feet and whatnot.  Any plans for putting in new RAW tags for this, even if you don't fill them in yourself?  I'd bet the community would love to help with data-entry work like that.  (that or someone might be able to write a program to generate the information based on rules)

Another major lack of information in the languages are words for local creatures-- it's weird that there can be no Dwarven words for creatures they encounter every day.  That said, I understand that there might be... technical issues, given that modders can add new creatures (or take them away).  Do you have any thoughts regarding how you might fix this in the future?  For example, would the creature names go in the language file or the creature file?

I haven't planned out how or when I'm going to do the linguistics additions.  It's a very large project, and for that reason it's hard to justify starting it now, though the number of features that want something are slowly increasing (yeah, to the point where I had to fudge a few parameters).

The main obstacle to adding new words on the fly is the rules for their generation.  I already have simple preliminary rules in place for each of the languages for when I make their stock words, and you can also see a kind of word generator in place for kobold utterances.  It would just need to set those up with more care, and I feel pretty confident with world-gen/post-world-gen incorporation now after all of this generation, so I expect adding new words during the game would be fine.  There are related topics involving language change overall that also benefit from a system that allows change.

So maybe if we remove all the fantasy stuff,add the economy system back again and hope that the culture stuff is not broken in the next update....maybe we could use DF to simulate real life stuff?Could DF be used to simulate real life worlds?

I think there have been mods along these lines.  Nothing would stop you from having various human entities modeled on various peoples, though there isn't a raw format for specific dances etc. yet, so you wouldn't be able to align everything.  Once we have enough content in DF, we're planning to have a world gen slider(s) from real-world-no-dwarves to gray-goo-world-no-dwarves, with regular fantasy in between.  I'm not sure when the proper time to do that is -- certainly "not having dwarves" would need to be addressed in a more favorable way than "only adventure mode works", but that's a can of worms we've been successfully putting off for a decade or more.

Who exactly made those poems and who is going to make new ones?Because every single book in our worlds is made by crazy necros in their towers and its weird that only necromancer towers can  be used to store books.It would be nice to see libraries and non necromancer writers.

How are the fortress mode/adventure mode tavern activities going?I really want to be able to sing and play games with my living meat shields companions!

There are going to be a few new classes of "content creators" now that'll be active in world generation and also visiting your fortresses and so on.  You should also get a chance to be one in adventure mode, though I'll have to expand my vocabulary of negative words to describe your skill-less attempts.  There aren't new prose authors though.  It would be nice, and I was considering adding something to complement the new poetry books that are going to floating around, but there might just be poetry books and necromancer books for a bit.

The activities haven't been going since I've been working on the art form generation.  Once dances are generated, I'll have everything in place and we'll be able to get critters running around.

How much semi-random knowledge, on any number of topics, have you acquired over the years through having to research all these different subjects?

I don't know when that started exactly, but most things I learn now are somehow related to something I'll be working on in the game.  I recently got through reading several law codes/commentaries from various civilizations in preparation for that framework, which is something I never would have ventured into without DF.  It's hard to say how much, though.  I don't feel particularly well-rounded.

Is poetry and music going to stay within te current modern western paradigm? Or might we have things like freestyle, forms that blend poetry and music (like rap) or forms of oppositional poetry like flyting and rap battles?

I'm not sure how to answer the first question without more specifics.  There are aspects of the poetic forms that are difficult/impossible in the typical western languages (not just tones, but some of the positional stuff, etc.).  There are already very free poems -- the verse paragraphs don't specify most parameters.

I haven't gotten into poetry recital at all yet (aside from what we now have in songs), and that's still up in the air.  I don't want to promise any competitive/etc. forms since that's tougher work to organize in-game, though it's an important theme and we were hoping to do something.  The immediate frontier of possible features has grown larger again...

Currently there are many materials, reactions and nearly all workshops hardcoded. The glass industry might be a good example. Are there any plans to un-hardcode some of those? Also any plans to enable multi-level workshops for modding?

The glass industry will be ready to come out once the categories can take materials, if I remember.  So it's basically just sitting there now, once the wind blows that way, like many other things.  Save compat is an issue -- because the glasses are hard-coded, there'll probably need to be a save compat break.  I have long lists of small things that I can accomplish during save compat breaks, and I guess this is another one.  Another issue is having to read about glass a bit more -- I think our current crystal glass is bizarre and needs to be replaced, if I remember.

I'm not sure what multi-level means.  Like upgrades?  Maybe I forgot to paste that in from the question.  We'll probably get to the de-building-ification of workshops before we do anything like upgrades, which'll accomplish something similar through furniture.

Will banquets be an aspect of the new entertainment system?

He he he, I read that as Nintendo Entertainment System and was confused.  I imagine we'll look into culture surrounding foodiness a bit with the recipe release (which is just after this tavern one and bug-fixing).  There'll be replacements for the removed parties, but what exactly those'll be is still unclear.  They'll involve the new art forms in various ways, most likely.  Food service for groups probably won't be well-organized this time.

DF Data structure question: Could you list what the 16 historical figure flags refer to?  Based on work from the DFHack team we know that bit 3 flags a deity, 4 a force, and 8 a ghost.  The rest are unknown but I've been doing some digging and I know that all the other bits are in use (except possibly 1, 2, and 5)

Yeah, the first one is for whether associated artwork should be revealed -- don't remember if/how that works anymore.  Second is whether or not its initial equipment has been created.  Five and six are whether it has "skeletal" or "rotting" for it's deity name (should probably be in a deity profile somewhere instead).  Seven is a temp world gen flag for "I've acted".  Nine, ten, eleven are whether skin/meat/bones have been destroyed in its death, for corpse generation in tombs etc.  Lots of exceptional cases there that don't work probably.  Twelve through fifteen are copies of the quest-type flags -- brag on kill, kill quest, chat worthy and flashes.  Sixteen being on means to never cull the hf.

[Vattic]

Cheers Toady.

[Robosaur]

Is it possible for there to eventually be capabilities to apply creature variations through syndromes in the future? I think it'd add a bit of interestingness to were-creatures, if they still somewhat resembled the creature they once were. An ettin-werewolf being huge and having two heads, for instance.

[Pidgeot]

Any plans to extend the 256 ASCII-Tiles in the future? Probably to 512 or even unlimited tiles for optical differences. As there are so many tiles having multi-uses, it is quite confusing sometimes. I am not referring to adding graphics, just implementing other tiles not included in the ASCII e.g. horizontal brackets/letters or something like that. Same regarding colours. Any plans to increase them from 16 to e.g. 256 colours or even 24-bit? Some parts of the code like the color-definiton tokens seem to support them already. Or will that cause major recoding?
...
I am not referring to graphics, which for sure will cause major and long time dedicated work. Just extending it from 256 tiles to some more. Additional tiles could be derived from e.g. UNICODE. I am no specialist regarding copyright. But as most of the UNICODE-tiles are from various recent/historic alphabets, I can't believe they are more restricted than ASCII.

Besides, moving to Unicode would require rewriting a large portion of the display code that relies on 8-bit display tiles.

Even a basic addition of another character page is not a rewrite I feel comfortable doing.  The DF part is fine -- it'd take a while, but I could do the DF side of it once I knew what data to push where.  However, I have no idea how to set up additional texture atlases while still keeping the speed up in OpenGL, assuming it wouldn't require a completely different approach, so I can't really start.  It's way on the back burner for this reason.  The game could probably use another thousand characters at this point to remove fundamental overlaps.  That's potentially independent of the item graphics problem, though they both run into the texture atlas issue (if you want more than a few item pictures).  Item pictures are also complicated by material considerations -- those aren't prohibitive, but take us into graphics territory that I don't want to get mired in.

This has probably already crossed your mind, or been suggested before, but... at least in principle, it should "just" be a matter of using a 3D texture, which is essentially just multiple 2D textures placed next to each other in memory. They work just like 2D textures, except you use a third coordinate to specify which of the loaded textures to use.

This approach would require all of the involved textures to have the same width and height... but if the tiles have to be the same size anyway, that gets handled by having the same amount of tiles on each texture. You could also allow tilesets with an arbitrary number of tiles if you padded smaller textures by adding blank areas.

Actually, doesn't this problem already exist with creature graphics? They're in separate textures too, so I'm assuming you're switching out the atlas for those - is the performance hit too large to use the same approach for multi-tileset renering? (If it is, using a 3D texture would allow you to handle creature graphics by just adding them as extra layers to the 3D texture...)

[mifki]

Even a basic addition of another character page is not a rewrite I feel comfortable doing.  The DF part is fine -- it'd take a while, but I could do the DF side of it once I knew what data to push where.  However, I have no idea how to set up additional texture atlases while still keeping the speed up in OpenGL, assuming it wouldn't require a completely different approach, so I can't really start.  It's way on the back burner for this reason.  The game could probably use another thousand characters at this point to remove fundamental overlaps.  That's potentially independent of the item graphics problem, though they both run into the texture atlas issue (if you want more than a few item pictures).  Item pictures are also complicated by material considerations -- those aren't prohibitive, but take us into graphics territory that I don't want to get mired in.

I don't see any problems with OpenGL part here. Anyway DF rendering code works with one texture only and loads all the tiles into it, getting pointers to some data for the loaded tiles. Then renderer just finds that data by tile number for each changed tile, and does its job. I don't think any graphics card exists now with maximum texture size < 2048x2048, that gives us 16k 16x16px tiles - more than enough.

Screen buffer currently uses 4 bytes for each screen tile - tile number, fg, bg, brightness. It doesn't seem to be very difficult to change it so that tile number is `short` instead of `char` (assuming that you use functions from the graphics lib and don't access these bytes manually all over the place). Then the old code will work as before but you'll be able to pass tile numbers >256 where you want and load them from raws. Other options are possible as well like making better use of the brightness byte - currently only one bit is used, but another several bits can be used to alternate tileset as the first bit alternates set of colours (again, it has nothing to do with OpenGL textures - if I'm not mistaken, it should be only several lines to change in the graphics code). And yes, Unicode characters can be used so that you don't have a problem with the actual tile images you mentioned earlier.

Another interesting question is with colours - here we don't have a problem with data types, and I think I saw a piece of code in DF that seems to approximate colour token values to 16 available colours. Being able to load more than 16 colours from the config file and use them for approximation seems to be much easier than all that tileset stuff. I don't care that much about colours though.

Of course it's all just a theoretical talk in case it helps you to understand graphics stuff, it all depends on your priorities.