New vision for Angband - Part 1

[takkaria]

Or 4. Something else happens

OK, someone needs to say it.

Nick McConnell should be the Angband maintainer.

Comments from anyone, particularly myself, encouraged.

[Nick]

1. A new takkaria steps up and produces version 4.0, and people love it.

2. V fossilises at 3.5 and Angband lives on through Sil, ToME, FA, Mist and (perhaps!) Pyrel.

3. One or more of the devteam acquire a new lease of life and produce 3.6 which, somehow, is interesting for them to work on and welcomed by people who are not shouted down by detractors. Timo welcomes this as the best Angband ever and I update my sig again.

Or 4. Something else happens

[Antoine]

I think fizzix or PowerWyrm could be good maintainers for V but i doubt either has the time.

A.

[Magnate]

I think it wouldn't be a bad outcome if V itself stopped moving and fossilised, like Moria did years ago, with variants continuing.

Noooo. How come? Is it because the codebase is old? Or the community doesn't want the change? Or the design is a bit of an evolutionary dead end?

As takkaria said, nobody with the expertise to lead the development of V has the inclination to do so. All three of your guesses are correct - the codebase is too old and crufty, and is also a bit of an evolutionary dead end in design terms (and so is the gameplay). And yes, we have, as you guessed in another post, suffered huge amounts of opprobrium for making changes that people didn't like. I don't just mean the normal balance of criticism (which is always a bit more common than praise simply because we all take a bit more time to post about what we don't like rather than what we do) - I mean totally rabid, insanely vehement and over-the-top, both in volume and tone, way out of all proportion (IMO) to whatever it was that was bugging people. It really did feel like V was some sort of sacred animal that we repeatedly violated. I guess it was.

It's been a really interesting journey (I've been on the devteam for about 4.5 years, though takkaria started it six or seven years ago) - the archives of this forum are a really good case study of how a small number of vocal detractors can drown out a much larger number of quieter folk.

It turns out that vocal people preferred the old "maintain rather than develop" model, where V changed ever so slowly over the years. We made more changes in 3.1.x over two years than over the preceding decade and a half put together, and that just wasn't right for V, though I still commend takkaria for putting a team together to do it. (There were also lots of changes in 3.0 and 3.2-3.4, relative to earlier versions.)

So we are in a situation where "the community" prefers the old maintainer model, but nobody wants to do that; the "devteam" prefers the newer model, but there's no incentive to develop Angband because people don't want what we want(ed) to do, and there's nobody who's willing to do development solely based on what forum posters want, without any direction of their own.

I think there are three possibilities:

1. A new takkaria steps up and produces version 4.0, and people love it.

2. V fossilises at 3.5 and Angband lives on through Sil, ToME, FA, Mist and (perhaps!) Pyrel.

3. One or more of the devteam acquire a new lease of life and produce 3.6 which, somehow, is interesting for them to work on and welcomed by people who are not shouted down by detractors. Timo welcomes this as the best Angband ever and I update my sig again.

[fizzix]

The damage multiplier issue can also be fixed in a number of ways; Pyrel does so by forcing the player to make a tradeoff between lots of weak hits and a few strong hits, with the average DPS being similar regardless.

Yeah, in effect this is almost the same idea as Sil with the exception that instead of lots of little hits, you get critical hits. The reason I don't like multiple blows is mainly because it feels less clean from an "action" point of view. I like the simplification of having each non-movement take 1 turn.

[Derakon]

Multiple blows creates two problems: first, they can be a huge damage multiplier, and second, they create a lot of message spam.

I fixed the second in Pyrel by condensing combat messages (as I'm sure other variants have): "You attack the Large kobold (4, 7!, miss)". The exclamation point indicates a critical hit; more exclamation points mean better crits (stacking indefinitely with a supercharge system much like speed rings use). Perhaps not quite as flavorful as "It was a good hit!" but the improved terseness is worth it.

The damage multiplier issue can also be fixed in a number of ways; Pyrel does so by forcing the player to make a tradeoff between lots of weak hits and a few strong hits, with the average DPS being similar regardless.

[fizzix]

For what it's worth, I already have a coded version with various monsters moved to speeds -5, +5 and +15. This could be ported into V with no issues, so maybe after 3.5 is out, we can do that if there's enough interest.

@multiple blows:
I would get rid of them. It's hard and wonky to balance. I would either replace them with a critical hit mechanic like in Sil, or just eliminate them altogether. They probably would be replaced with special abilities like a 30% chance of shield bashing, etc. I would have all actions take one movement turn, except for character movement which is what speed would take care of. This is more than just rescaling V though, this is more in line with "the game I'd like to write if I only had the time/ability"

[Timo Pietilä]

Well I think the player divides the monster list in his head into slow/normal/fast/very fast so when he sees a monster he generally knows their speed and whether they are slower or faster than @

Also since at the moment each monster is a multiple of 10 it is easy to calculate how often you get an extra move.

But if each monster was +5, +8, +11 etc. then I think it would be hard for the player to remember and visualise relative speeds.

+5 doesn't happen, but +8 and +11 can already. Base speed is multiples of 10, but there is variance. Not sure how that is calculated, because usually it's just base speed. I usually notice the difference when I'm running away from largish group of orcs. There are always one or two that are faster than rest, and also few that are slower than rest.

[TJS]

I think it's probably easier to start with something like Sil and build on it than it is to strip Angband down and rebuild it. Sil to a large extent has already done the stripping. Plus it has 4GAI, and roaming monsters, and a great stealth system. I don't know though, I haven't looked into it in much depth.

I spent last night reading up about more of the Sil mechanics and I'm starting to see where you're coming from. It has solved the searching problem as someone mentioned earlier and has done other things like detecting invisible monsters much better. Experience being reduced for fighing the same monsters is a good idea too. Plus the AI is way better.

The code is crusty even though it's in pretty good shape all things considered. If I was going to continue development I'd re-introduce Lua, introduce JavaScript or switch to Pyrel.

Had a look at the code early on and it looks pretty good to me.

It was in Lua at some point?

I think there's some changes that could be made (pretty well-trodden ground by now I think, in this thread and others) to V and it remain V but no-one with the ability has the enthusiasm right now. If you have a strong sense of game design you're not going to continue V, you'll do your own thing. Unless you're a masochist.

You could be right. Actualy this thread was an attempt to see if it is possible to make big gameplay changes on V and what people thought about the idea.

Did you get a lot of hassle for certain changes over the years?

Hah.

[TJS]

Neither of these seem like rule-fudging to me. Some rooms are lit in the dungeon and some aren't - is this rule fudging?

Actually you're right and I thought after I'd posted it that it wasn't really rule fudging at all. More in the category of rule complexity which I'd argue was unnecessary and overly complex.

I don't think early completely pre-lit levels are a good idea because it changes one of the fundamental gameplay mechanics of the game (the light system) and I don't think the benefits from it are worthwhile. Plus there are several drawbacks, such as confusing new players who would deem light sources unnecessary and also introducing a jarring gameplay change at the transition between lit and unlit levels.

It would change the internal consistency of the map generation too I'd argue. At the moment dungeons are generated in a semi realistic fashion with some rooms lit and some not and corridors unlit. You could imagine that's how a dungeon complex would actually be. If the shallowest levels were all lit up it for some reason it would break that logic (to me anyway).

Another problem I have with it is that it is a method for trying to make early levels easier rather than trying to fix something with the actual light mechanic itself (which I think works very well) and I always think it is better to treat the symptoms rather than the cause. In this case it is the fact that the player is so weak at the start in comparison to after the first few levels.

The player starts with no escapes, hardly any healing, a 1 square light radius and pretty ropey equipment and for anything except a warrior hardly any hitpoints (or mana). If you stumble upon the wrong monster early on you've had it without having any chance of avoiding the situation.

This was going to be in my next post actually

What's the difference between knowing some rooms from the beginning of the level vs some not?

For me personally if I go down to a brand new dungeon and then it was suddenly all mapped out, it would feel like a jarring experience almost as if I was robbed of the chance to explore it for myself. It would seem odd too that you could know some levels out of an almost infinite network of dungeons that are impossible to ever return to. Again as with the pre-lit levels I think that it is trying to change the wrong thing to fix a problem.

I guess the point I'm getting at is that in the end rules should primarily be about good gameplay, and secondarily should be understandable and simple. But if you have strong opinions about what constitutes a fudge then you're opinionated enough to write a variant.

I agree with you about gameplay first, I just think that suddenly changing unbroken mechanics to try and fix a different problem isn't the way to go about it.

Yes I probably am opionated to write a variant

[TJS]

Does this really matter that much? As a player, I often go from 0 to 2 to 5 to 9 to 13 to 17 to 20 to 23 to .... speed -- even if the monster has a fixed 0, 10, 20, or 30, it doesn't really help me know how and when they'll 2x turn me. Especially since afaicr speed isn't linear anyways?

Well I think the player divides the monster list in his head into slow/normal/fast/very fast so when he sees a monster he generally knows their speed and whether they are slower or faster than @

Also since at the moment each monster is a multiple of 10 it is easy to calculate how often you get an extra move.

But if each monster was +5, +8, +11 etc. then I think it would be hard for the player to remember and visualise relative speeds.

[TJS]

Actually I think the combat formulas in Sil are more complicated (with the exception of the Angband critical hit calculation which is a gigantic mess...and fine...I guess the blows calculation too.)

This mainly arises from Sil having different numbers for to-hit (evasion) and damage received (protection) whereas angband either uses the same number for both (monsters attacking player) or does not include any protection (player attacking monster).

Here are all the things that would need to change.

• Player stats range 1-30 (instead of 3-40)
• Player HP ranges from 1-200 instead of 1-1200
• Monster HP ranges from 1-300 instead of 1-20000
• Player SP ranges from 1-100
• Player spell damage changes
• Weapon damage dice changes
• Monster damage changes
• Monster spell damage changes
• Monster breaths completely rescaled
• Player level maxing at 50 is still ok, I guess.
• XP should be more linear, yet not vulnerable to grinding (hard)
• AC needs to range from 1-30
• Probably need to rethink multiple blows per turn
• Complete rebalance of archery damage
• Rescaling of stat-gain stat-loss effects/potions

and probably more.

Cool thanks for the information. Thinking about it some things could be done one at a time, like rebalancing AC first for example.

What are your thoughts on multiple blows by the way?

[Derakon]

For what it's worth, in the Variant I Want To Write™ Once Pyrel Is Finished™ (), there'd be a wide range of action categories whose speed could be modified. Off the top of my head:

* Walking
* Melee attacks (blows/round)
* Missile attacks (shots/round)
* Throwing items
* Casting spells
* Using items
* Tunneling
* Swapping equipment

And these could be broken down further. In particular, some spells could be faster to cast than others, and you could find e.g. gear that made casting any teleportation spell faster.

Most of these stats would remain unmodified for most characters. But they would provide interesting ways to differentiate special equipment. You might be tempted to try out an Amulet of Wands (+2.0 wand use speed) and go with a wand-heavy "build", for example. I haven't yet figured out how energy costs would be messaged to the player in a clean and consistent way, but given how far off in the future this is (for me anyway), I'm not too worried about it.

[takkaria]

Even if all the most obvious changes desirable to Angband are already done well in Sil then isn't that a good reason to adopt some of those changes?

Could you expand more on why it would be better starting from Sil?

I think it's probably easier to start with something like Sil and build on it than it is to strip Angband down and rebuild it. Sil to a large extent has already done the stripping. Plus it has 4GAI, and roaming monsters, and a great stealth system. I don't know though, I haven't looked into it in much depth.

I think it wouldn't be a bad outcome if V itself stopped moving and fossilised, like Moria did years ago, with variants continuing.

Noooo. How come? Is it because the codebase is old? Or the community doesn't want the change? Or the design is a bit of an evolutionary dead end?

The code is crusty even though it's in pretty good shape all things considered. If I was going to continue development I'd re-introduce Lua, introduce JavaScript or switch to Pyrel. I think there's some changes that could be made (pretty well-trodden ground by now I think, in this thread and others) to V and it remain V but no-one with the ability has the enthusiasm right now. If you have a strong sense of game design you're not going to continue V, you'll do your own thing. Unless you're a masochist.

Or is it like when Roger Waters left Pink Floyd and didn't want the others to continue to use the name?

Hah.

[takkaria]

Agreed with the knowledge differences, although I'm not sure what the answer is. I would be against the game rules being fudged with some early dungeons getting mapped out and lit beforehand though (this is a good example of the type of exception to the game rules I was talking about before).

When you say movement do you mean actually moving the @ around the map or do you include other actions such as drinking potions? I'm not sure I like the idea of different actions taking up different amount of time (another example of the kind of rule fudging I was referencing earlier).

Neither of these seem like rule-fudging to me. Some rooms are lit in the dungeon and some aren't - is this rule fudging? What's the difference between knowing some rooms from the beginning of the level vs some not? I think the thing to strive for is rules that are easy to express. e.g. "Some things modify your attack speed, others modify your movement speed" seems like just as good a rule as "Some things make everything you do happen faster, other things let you attack faster". And "Some levels have some rooms which are pre-lit and/or pre-known" doesn't seem obviously more fudgy than "some levels have some rooms which are pre-lit".

I guess the point I'm getting at is that in the end rules should primarily be about good gameplay, and secondarily should be understandable and simple. But if you have strong opinions about what constitutes a fudge then you're opinionated enough to write a variant.