Plans for 4.1 - 4.3

[Nick]

This is pretty much what is in my head for the next three major versions at the moment. It gets vaguer as the version number increases, and some of these may not happen for various reasons (failure to implement, don't work as expected, player revolt, etc). Anyway, please comment.

4.1

• Traps - basic principle is less traps, less dangerous, harder to avoid. No more specific trap locations in vaults. Magical trap detection and specific searching for traps to be removed. The player will have one chance to detect nearby traps, dependent only on perception/searching skill. The player will also get a saving throw against the effects of traps. Consider allowing objects in trapped grids and/or trapped objects.
  
• ID - ID for consumables will work as currently. For wearables, each potential property (“rune”) will be immediately recognised once it has been seen once in any game (similar to how flavours are learned). How exactly this works for artifacts, ego items, jewellery, and to-hit/to/dam/to-ac bonuses is still to be fully worked out.
  
• Player knowledge - split the player’s knowledge of the current level from the game’s knowledge, allowing things like doors opening out of sight, but the player still seeing them as shut.
  
• Dungeon generation - lots more vaults and room types. New (fairly rare) level types in the same vein as caverns and labyrinths. More use of areas (from rooms to whole levels) containing monsters all fitting some theme. Consider reduction in size of levels, but making them more dense and interesting.
  
• Monster AI - monster pathfinding will be improved, so they’re not as easy to trick. Maybe some of the stuff under 4.2 gets done here.
  
• Monster breaths - to become cone shaped spreading out from the monster, as in Oangband. Force will push players and monsters, distinguishing it from plasma.
  
• Terrain - all rubble to become passable, but blocking line of sight. Lava will probably at some point become passable, but damaging to walk through.
  
• Calculations - there are lots of individual calculations, particularly related to randarts and object and monster power, which might be adjusted for reasons
• of balance, accuracy or simplicity.
  
• Theme - look at being a little more thematically consistent (along the lines of Angels -> Ainur).

4.2

• Terrain - consider new types of terrain to allow different combinations of:
  • allowing some or all monsters and/or the player;
  • allowing line of sight;
  • allowing spells and missiles;
  • slowing or speeding movement of some or all monsters and/or the player;
  • holding objects;
  • holding traps;
  • adding atmosphere or giving information.
  
  For example passable rubble allows movement but not line of sight, a window might allow line of sight but not movement or projectiles, and a battlement might allow line of sight and projectiles but not movement.
  
• Monster list - Consider how changes to terrain and monster AI have affected existing monsters and see if they need to be adjusted or moved shallower or deeper. See if there are obvious gaps where there is no monster in an interesting “ecological niche”. See if there are too many monsters which are more or less the same.
  
• Monster AI - No intention at this stage to introduce monster mana, but yes to a thorough examination of spellcasting AI. Possible changes to how stealth works. Possible introduction of monsters being awake but not aware of the player (as in Sil). Think about group behaviour, and potential for (limited) conflict between monsters.
  
• Classes/Races - Rethink what the aims of all the classes are, especially Rogues which are more interesting in many variants. In particular, rethink current spellcasting, with a view to reducing the number of books for half-casters and allowing variation of the same books for different classes. Look at introducing some new classes, possibly from variants - candidates include Necromancers, Warrior-Mages, Druids, Assassins. Races - consider Tolkienising.

4.3

• Combat system - the current one could be tweaked, or we could introduce O-combat, or v4 combat, or Sil combat, or some combination, or something else altogether. Any change will be disliked by someone, but it’s still worth thinking about.
  
• Objects - this is very dependent on the combat system, and is impacted by all the previous changes. One obvious thing to consider is the affix system developed by Magnate for v4.

[debo]

If you add force breath pushing, you should also add chasms (like in Sil) so that you can get force-pushed off a cliff ^_^

[Ingwe Ingweron]

If you add force breath pushing, you should also add chasms (like in Sil) so that you can get force-pushed off a cliff ^_^

I'm surprised, Debo, where is the "...and Rockets"!

[Derakon]

Force is already technically distinct from plasma in that they destroy different items, but yeah, sounds good.

We need to get better help text onto the chargen screen, with the goal of directing new players at easier race/class options or at least properly setting their expectations when they roll up an elf mage. That's the only thing I can think of that needs doing and wasn't on your list.

[wobbly]

Look at introducing some new classes, possibly from variants - candidates include Necromancers, Warrior-Mages, Druids, Assassins, Hounds.

Good, looking forward to it.

Races - consider Tolkienising.

I'm not sure whether to agree or disagree. Sil does such a good job of staying thematically accurate. With Angband I'm not sure whether it's better to clean up the theme or to just drop the pretense & remove the most obvious references. FA I guess is not too bad in this area... well sorta... I mean it does alright... I guess...

[Nomad]

I'd like to see the list of ego weapons/armour expanded to try and add some variety to pre-artefact-level equipment choices. (For instance, shields/armour with a combo of two base resists and a higher resist, egos that grant a single protection or higher resist, mixed blessing egos that boost one stat but reduce another - nothing super-powerful, but things that can fill gaps in resistance coverage and make equipment choices more of a decision than current no-brainer progressions like single resist to resistance to elvenkind.) Rune-based ID would definitely make it easier to have more and more flexible ego types without the hassle of the ID game.

• Dungeon generation - lots more vaults and room types. New (fairly rare) level types in the same vein as caverns and labyrinths. More use of areas (from rooms to whole levels) containing monsters all fitting some theme. Consider reduction in size of levels, but making them more dense and interesting.

I have Many Thoughts On This, as I have rambled at length on numerous occasions. To try and summarise my views on what would make for an ideal system (defined mainly as one in which I can indulge my urge to design tons of new rooms without having to futz with the code):

- Pits layouts should be moved out to edit files
- The format of room_template.txt, vault.txt and the hypothetical pit layout file should ideally be harmonised so they all use the same symbols for the same things, and you can make rooms combining features of all three
- Layouts should be possible to use both specifically and flexibly, so that you can specify when a room is generated that it should contain given monster and treasure types, but you're not rigidly locked into using those same types every single time it's generated (for this, I favour combining a generic template with a monster/treasure profile from another file, similar to how pits currently work)
- It should be possible to add/remove/customise all of these rooms types, and new level profiles or themes, entirely via the edit files without having to touch the code

Separate from those more general goals, I also like the idea of unique-specific vaults/rooms (e.g. a mushroom field containing Fang and Grip, a dwarf hall with Mim and his sons, or a maze with Baphomet the Minotaur Lord at the centre) that occur at their native depth, can only appear once per game, and are found in destroyed condition if all the resident uniques are already dead by the time you get there.

• Classes/Races - Rethink what the aims of all the classes are, especially Rogues which are more interesting in many variants. In particular, rethink current spellcasting, with a view to reducing the number of books for half-casters and allowing variation of the same books for different classes. Look at introducing some new classes, possibly from variants - candidates include Necromancers, Warrior-Mages, Druids, Assassins. Races - consider Tolkienising.

I'd really like to see more race-specific (and possibly class-specific) special abilities, like Dwarf treasure-sensing and Gnome wand-identification - I feel like those do a lot more to make playing a different race feel like a different experience than just stat bonuses or inherent resists.

[debo]

I'm surprised, Debo, where is the "...and Rockets"!

That's basically implicit in all my posts at this point, as you've demonstrated

[Whelk]

This all sounds real exciting. I'm a flavor geek, so I love the idea of Tolkienising a little more - something as simple as changing some creature names would be fun. I'd be glad to help offer suggestions with this, though I wouldn't be surprised if Nick already had most things figured out already, since he seems to be more of a Tolkien nerd than I am.

The more flavorful dungeon layouts is something that I would really look forward to as well - it's one of my favorite things about playing FA and Stone Soup: you can come across rooms that are actually interesting instead of just endless randomized kludges of corridors. They don't have to have special features or anything - just an obvious theme or aesthetic design to a small room can make the experience so much more fun and makes it feel more like you're "exploring".

I'd really like to see more race-specific (and possibly class-specific) special abilities, like Dwarf treasure-sensing and Gnome wand-identification - I feel like those do a lot more to make playing a different race feel like a different experience than just stat bonuses or inherent resists.

Don't forget the best one - Hobbit mushroom identification! I've always loved that perk.

[Cold_Heart]

I would also propose randomly generated ego items somewhere along the way (as an option - preset egos(trickery/devotion/westerness etc)/present egos + random egos/all egos random).

Oh and also removal of fuzzy detection.

[Nomad]

Don't forget the best one - Hobbit mushroom identification! I've always loved that perk.

That one is actually particularly great because it has comparatively tiny in-game advantage but is such a neat little flavour quirk. More simple little things like that would be great - I feel like Half-orcs and Half-trolls should be able to recognise weapons of Slay Orc and Slay Troll, for instance. Actually, rune-based ID would combine very well with these kinds of special abilities, because different races/classes could start the game already knowing one or two thematically appropriate runes: priests might recognise Blessed items and paladins Slay Undead, for instance.

[Nick]

I
- Pits layouts should be moved out to edit files
- The format of room_template.txt, vault.txt and the hypothetical pit layout file should ideally be harmonised so they all use the same symbols for the same things, and you can make rooms combining features of all three
- Layouts should be possible to use both specifically and flexibly, so that you can specify when a room is generated that it should contain given monster and treasure types, but you're not rigidly locked into using those same types every single time it's generated (for this, I favour combining a generic template with a monster/treasure profile from another file, similar to how pits currently work)
- It should be possible to add/remove/customise all of these rooms types, and new level profiles or themes, entirely via the edit files without having to touch the code

This is all very much along the lines of what I'm thinking. The harmonising of symbols is I think almost completely done. I also like the idea of having different theme variants of the same room - possibly just by listing the relevant monsters/objects in the room edit file.

[Nick]

This all sounds real exciting. I'm a flavor geek, so I love the idea of Tolkienising a little more - something as simple as changing some creature names would be fun. I'd be glad to help offer suggestions with this, though I wouldn't be surprised if Nick already had most things figured out already, since he seems to be more of a Tolkien nerd than I am.

I do already have some ideas, which may even start to creep into 4.0 in a completely gameplay-neutral fashion

[Nomad]

This is all very much along the lines of what I'm thinking. The harmonising of symbols is I think almost completely done. I also like the idea of having different theme variants of the same room - possibly just by listing the relevant monsters/objects in the room edit file.

It would be good to be able to have multiple possible object tvals for a given room, perhaps with a series of object lines - for instance, you could have a 'mining camp' layout where the allowed objects are diggers, lights and flasks of oil, and then apply monster themes that make all the 'miners' orcs, kobolds, hobbits etc., whether randomly or according to an overall level theme.

And in fact, if you're eventually including new terrains, you could even make that equally flexible by theme. For instance, you could have a layout like:

D:##############
D:#............#
D:#..;;;;;;;;..#
D:#.;;;;;;;;;;.#
D:#..;;;;;;;;..#
D:#............#
D:##############
T:lava
T:water
T:swamp

Where ; is your all-purpose 'terrain' symbol that can be generated as any one of the specified terrain themes, making the same room serve triple duty as a pool, a lava pit or a swamp. Combine with, say, three possible monster themes and three object themes, and suddenly from one layout design you've got 27 possible rooms.

I'm in favour of introducing that kind of thematic flexibility wherever possible, because it both makes the dungeon less predictable and vastly increases the bang for your buck in terms of number of layouts required. I mean, I've got about 150 special rooms designed already - if they all had 27 possible variations, that would generate ~4000 different rooms from the same-sized edit file.

[Nomad]

- Pits layouts should be moved out to edit files
- The format of room_template.txt, vault.txt and the hypothetical pit layout file should ideally be harmonised so they all use the same symbols for the same things, and you can make rooms combining features of all three

On further consideration, I'd slightly amend/expand on this suggestion:

* Keep room_template.txt purely for basic unthemed rooms with random monsters and objects at random positions. (And maybe get rid of the current rather bodged 'scatter some stuff around this point' mechanic in favour of header lines specifying % chance of objects/monsters/traps in each open square, or something.)

* Keep vault.txt for current-style vaults where the positions and OOD-ness of monsters/treasures are specified but the types are entirely random.

* Make the third file themed_rooms, for all rooms with fixed-position monsters/treasures/terrains and lists of permitted themes for filling them. Pits would be a subset, and so would the tval-specified treasure rooms currently in room_template.

[Timo Pietilä]

This is pretty much what is in my head for the next three major versions at the moment.

Any plans for the curses? Sticky curse is bad game design and difficult to fix, my idea for that is "semi-sticky", you can take off the item, but you can't drop it. If you try it goes to inventory instead.

Items could have (some, lesser) effect while in inventory as well. Like aggravating item works like big malus to stealth or like that.