Class/magic feature branch

[Nick]

OK, I accidentally entered a shop in fox form. Now all shops are closed. What am I supposed to do ?

Oops, that's bad. Save and reload seems to be a workaround, I'll look into that and get a fix out.

[Thraalbee]

Well, the intent as far as I understand it is to simplify testing new ideas and I for one love this. After many years thinking about it, finally it's easy enough for me to work on a new variant or possibly just a new class. I can't see how greatly simplifying for the community (including dev's) to test new stuff can be a bad thing. At least not unless the goal is to come up with an interesting new version of Angband in very short time.

[Tibarius]

agile development

The development of the 'new' classes reminds me a lot of agile development.

Meanwhile i consider it vastly inferior to the classis software development. And i think this threat is a shining example of why that is the case.

Either unwillingness or unableness to define goals and aims first leads into several development circles which require a lot more ressource investment than the other way around.

After thinking about it, i consider 'agile development' a clever method to increase total amount of required ressources without having a single person involved being able to be identified as cause for this, because it is method inherent.

So applaus to the consultant whoever came first to this wisdom and makes corporations pay a lot more money for software development than realy being required.

[Estie]

OK, I accidentally entered a shop in fox form. Now all shops are closed. What am I supposed to do ?

[Estie]

When trying to start the game, I get an error message about libpng12.dll missing.

Edit: Nm, redownloading fixed.

[PowerWyrm]

Banishment before: 1 cast, everything gone. Banishment now: lure stuff outside of vault, banish, repeat. Does that change anything? Nope, except increasing the turn count of your game...

[Ighalli]

+5 speed replaced by doubled movement speed

Does it make sense to break the speed display out to have movement listed separately when there is a bonus or penalty? Should multiple blows/shots end up showing effective melee/shooting speeds? How about spellcasting? One could imagine a necromancer spell that increases spellcasting speed while causing HP drain or bleeding.

[Nick]

There are new builds up on the build page, but not a lot of obvious change - mages get a room-lighting spell, and druids' fox form has been tweaked (worse combat, +5 speed replaced by doubled movement speed).

I have spent considerable time rewriting the effect code, though, to try and improve the readability both of the actual code and of the data files. I am also working on getting the Necromancer class ready. So I thought I'd at least get the latest code out there, but it's not all that exciting.

[Nick]

You say my playthrough was helpful - in which way ?

Mainly for the info that the 3rd book spells were all rubbish, but also the feedback on the shapes.

You intend, if I understand correctly, for shapes to not be increasingly superior ways for applying melee damage, but rather niche solutions to specific problems.
If that is so, there must exist a reason to leave a given shape, besides the global "you cant use magic while shapeshifted" that applies to all shapes. In the case of the fox, I can see the possibility for such a design if the fox damage output is actually lower than unshifted. The druid would turn into a fox for a stealth bonus while exploring and unshift for combat. While a speed bonus is thematic, it directly affects damage, so if you insist on including it, maybe remove all attack abilities from fox.

Yes, your understanding is exactly correct, and the intent was for the fox damage to be lower. I don't want to completely remove combat, but to make it noticeably worse - so you can still deal with weak enemies while exploring, but need to change back for anything serious.

Leaving the bear aside for now, what would be a reason to use pukelman ? And what would be one to stop using it ?

OK, I think the fox is more or less explained; I'll just run through the other shapes in some detail.

• Pukelman - its a statue, so should be slow but hard to damage . The way I've done this is be a minus to speed (currently -5) and stealth (-2), and then a whole bunch of buffs:
  • Sustain STR and CON
  • PStun
  • Regen
  • Hold Life
  • +4 STR and CON
  • Poison immunity
  • RShards
  • immediate healing of poison and stunning on changing
  • can't get cut, and any existing cuts get just frozen at the current level until changing back
  • Direct reduction (currently by 10) of damage by any individual attack
  
  This last one is incredibly powerful, and I think the main art here is going to be balancing that against the speed reduction.
  The use case is fighting multiple medium-strength enemies - trolls, middling hounds, any infestation. The big weakness is against things that deal a lot of damage at once, so big breathers and cursers. I tested it a little, with the notable things being dying to a breath from an 11-headed hydra, and having a long slow losing melee battle against Uvatha.
• Bear - pretty simple, good at bashing stuff. PFear, plus to STR, CON, to-hit, to-dam and melee skill, and negative to stealth, INT and disarming. Main reason for changing back is inability to heal and pick up items.
• Eagle (the new one) - the eagles in middle earth were in practice mainly useful for getting people out of sticky situations (Gandalf, Bilbo, Sam, Frodo, Dwarves, Fingolfin). Mostly this is achieved by teleporting in Angband, but there are some cases where you don't want to do that - the big one being when there's gear you want to pick up. So eagles get survival skills - PBlind, PConf, PFear, SI, FA and complete immunity to traps - and quadrupled movement speed (this works like extra shots - moving a step only costs a quarter energy). So eagle form becomes viable in some cases for running away instead of teleporting, and also for raiding vaults.

Come to think of it, the correct thing to do with the fox is probably to give it extra movement speed instead of regular speed. Note that with the shapes I'm trying to satisfy both gameplay and thematic goals, and so there's an extra balancing act there.

Let me know if there are obvious flaws in that, or if anything isn't clear. My hope is that druids will be played in at least two possible ways (pure spellcaster or bear-form melee fighter, or some combination) depending on gear and player choice.

[Estie]

I appreciate the diplomacy and genuineness, but would be happier still for a maybe less diplomatic but more informative response. You say my playthrough was helpful - in which way ?

You intend, if I understand correctly, for shapes to not be increasingly superior ways for applying melee damage, but rather niche solutions to specific problems.
If that is so, there must exist a reason to leave a given shape, besides the global "you cant use magic while shapeshifted" that applies to all shapes. In the case of the fox, I can see the possibility for such a design if the fox damage output is actually lower than unshifted. The druid would turn into a fox for a stealth bonus while exploring and unshift for combat. While a speed bonus is thematic, it directly affects damage, so if you insist on including it, maybe remove all attack abilities from fox.

Leaving the bear aside for now, what would be a reason to use pukelman ? And what would be one to stop using it ?

If the reason for pukelman and fox is not superior damage, then you need to make sure that those forms actually do less damage than whatever you intend for the default druid combat - probably the spells (?). That is not "just balance", it has to be in the right ballpark otherwise the niche concept breaks apart.


Currently, I dont have much time for Angband either, but I will try and stay on track as much as possible.

[Ingwe Ingweron]

Tell me if you think that's wrong, or if you think I'm doing my bit badly. And again, apologies for treating you as an "unintelligent guinea pig" - not wanting to play favorites, but I value your input greatly.

Nick, this has to be one of the most diplomatic and genuine posts I've seen. Thank you for all your hard work, and I thank Estie too, as well as everyone who contributes to the strange and wonderful survival of this game for decades.

Small aside, I do wish we could get a bug-fix for some of the 4.1.2 issues (damage info problems), but I know they aren't as sexy and fun as all the new class stuff.

[Nick]

Or are you having completely different plans ? What am I supposed to test ?

I am sorry, but my impression currently is that you have no clue what you are doing. Shapes for different purposes is a plan (albeit one I doubt will turn out well - but thats another matter), but it requires that not one shape is superior in every way. And giving one shape an extra attack and speed, while reducing speed of a higher level form is not a mistake that should be found out by testing - something like that can be rooted out long before, in planning. For that kind of pre-alpha planning, it would be better to get presented with the ideas, so we can point out unreasonable things, before you go and program everything.

I am happy to play guinea pig, but only if I am treated like an intelligent guinea pig. Getting fed cryptic lines like "All classes are going to be challenge classes" - just keep playing - puts me off, especially when it turns out that I just did an utterly pointless playthrough by missing one vital piece of information that could have been avoided if we had communicated.

I'm sorry, this has largely been poor communication on my part (including stuff that sounds funny in my head when I'm under-caffeinated sometimes not translating well ).

My starting point for this branch was the previous discussion about classes and magic. Given that discussion, I thought the best way to progress was to put out actual new classes, because

• It would give me an idea of what was possible and what wasn't
• It would get me to do any underlying code changes that were needed (as opposed to the changes to data files)
• It would give us something clear and definite to talk about, and avoid misunderstandings that could arise from talking in generalities about what was being planned.

Maybe I should have put out a thorough description of each class first, with my reasoning for what the spells are and how it was supposed to play. I didn't do that partly because I thought it was better to allow people to look at it and see what they thought, and partly because that would have left me less time for actually making changes. I have definitely made basic mistakes in some of what I've put out; I expected that I would, but I didn't make that clear. Other factors include

• I don't have nearly as much time to work on this as I would like, so I'm erring on the side of releasing early and
• This is actually an enormous amount of fun, so I'm sometimes getting carried away with ideas and failing to see the negatives.

I am really grateful for all the feedback I'm getting from you and others - your druid playthrough was very helpful - but I should have made it clearer that I am nowhere near done with these new classes yet. While it looks all nice and shiny and official when you see "Druid" pop up on the birth screen, it's in its infancy. We have a long way to go. Lots of things will change.

My feeling is that me putting out changes, people here pointing out the good and bad points, and then future changes being informed by the resulting discussion is the best way to make progress. Tell me if you think that's wrong, or if you think I'm doing my bit badly. And again, apologies for treating you as an "unintelligent guinea pig" - not wanting to play favorites, but I value your input greatly.

[Derakon]

I liked the fact that you could banish vaults. We can agree to disagree. But that was not my point. Banishment was the top selling point for mages. They were arguably the weakest class even so, but by reducing their best spell to rubble, they simply fall off completely.

When I play mages, by the time I get Banishment I'm generally already ready to win, so I never really thought about it much. I'm not saying you're wrong, just that we all have different perspectives about what's important about the game. Which is why it's so important to have feedback. So thanks for giving your two cents.

"All classes are going to be challenge classes" - is that supposed to be funny ?

I assume Nick meant that the game ought to be difficult, and ideally roughly equally difficult for each class. Obviously there's going to be variations -- some classes are better early and some late, some have more difficulty with certain monsters or less with others, and of course there's always the question of whether you chose a race that's well-suited to the class. But it wouldn't surprise me to see warriors get whacked with the nerf bat once Nick's done bringing everyone else down a bit.

So yes, it was meant to be funny, but he's also serious.

[Estie]

I liked the fact that you could banish vaults. We can agree to disagree. But that was not my point. Banishment was the top selling point for mages. They were arguably the weakest class even so, but by reducing their best spell to rubble, they simply fall off completely.

"All classes are going to be challenge classes" - is that supposed to be funny ?

What you call "just balance" affects gameplay. A mage can choose to either use melee or archery or spells. Changing the power of one way within reason doesnt affect play, but if you dont give a damn at all, it creates weird monstrosities that end up being played not at all the way the endresult would be. When I played the druid, I used melee in unshifted form for a long period - is that intended ? Are you going to reduce warrior damage by 1/2 to make spells better ? If so, why dont you do it now, or, if it is a difficult task, why not simply increase spell damage to the amount where the relation with melee damage is right ? THAT would be far more important for any testing; I can play a melee mage, but to what end ?

Or are you having completely different plans ? What am I supposed to test ?

I am sorry, but my impression currently is that you have no clue what you are doing. Shapes for different purposes is a plan (albeit one I doubt will turn out well - but thats another matter), but it requires that not one shape is superior in every way. And giving one shape an extra attack and speed, while reducing speed of a higher level form is not a mistake that should be found out by testing - something like that can be rooted out long before, in planning. For that kind of pre-alpha planning, it would be better to get presented with the ideas, so we can point out unreasonable things, before you go and program everything.

I am happy to play guinea pig, but only if I am treated like an intelligent guinea pig. Getting fed cryptic lines like "All classes are going to be challenge classes" - just keep playing - puts me off, especially when it turns out that I just did an utterly pointless playthrough by missing one vital piece of information that could have been avoided if we had communicated.

[Nick]

All of the classes have their own distinct methods of interacting with the dungeon when they initially encounter it going all the way back up the variant inheritance chain. The warrior blunders into things which can hurt him like traps and monsters because he only has a radius 1 torch but he has the HP to survive if he rests between encounters. The mage is craftier and lights his way using magic to kill his enemies from a distance. The rogue sneaks around in the dark but unlike the warrior he finds his way using detection and picks his fights using stealth. Ranger is a little more idiomatic, it's the warrior/mage hybrid with bow skill attatched. The priest uses his faith to survive the early dungeon and is rewarded with awesome power and endurance later on. However he lacks some of the diversity of the mages thoroughly complete toolkit.

This stuff gets obscured by high level play and tactics to bypass stages of a lengthy game, but they're the fundementals. Mages need light.

All right, I can buy that (although I do think it makes sense to give rangers nature magic).

Light seems to have two somewhat distinct features in Angband. One is as an element, usable by monsters and the player. The other is as an information tool, constantly telling you the contents of all lighted areas in line of sight, Clearly, the first would fall into the domain of the priest, and it seems fair for that to be exclusive. Giving the information tool only to priest seems to be the issue here, though. I agree that it is odd for the class with the most dungeon control not to have an important early/midgame tool. In my opinion, giving mages the ability to make squares visible somehow would be important. If keeping mages from using light is important, Tome4 has an interesting spell called Arcane Eye that sees monsters (in the dark) in a certain radius around wherever you place it.

Thanks you for saying that. Light in Angband doesn't just have two distinct features, it is actually functionally two distinct elements, Light and Weak Light. Light is barely used - a few breathers, the Gil-Galad activation, and amusingly a trap (kudos to takkaria for that one) - but is an element just like others that damages all monsters and the player. Weak Light simply lights up floors and walls, and in projected form hurts light-sensitive monsters (like stone-to-mud hurts some monsters).

Solution here looks so obvious I don't know why I didn't see it before. Mages (and probably only them) can have some Weak Light spells - priests get actual Light spells.

As for copying realms from Oangband, as much as I like the variant, gameplay is not particularly distinct by class in it. Necromancers and druids do little more than fill gaps between the mage and the priest, necromancers being magey but with Dispel Evil, and druids being priesty but with elemental bolt spells (and for whatever reason, TO at clvl 15 and Detection at clvl 21, both in town books). They are not particularly distinct, nor do they seem to follow any consistent patterns in terms of gameplay or theme. They have some very interesting spells in the later books, but these seem to be allocated more or less at random.

I've actually found this too. In constructing the new classes I am certainly mining O (or more precisely FA) for ideas, but not following slavishly. I'm working on necromancers at the moment, and at this stage they're looking ... different

All I can see are new classes created by removing stuff from the former casters.
The casters had the lowest damage in the game, balanced by the highest utility from their books. The utility has been reduced below even what the old hybrids used to have (because banishment got nerfed - why this change ? I strongly disapprove), but the damage is still the same low level.

I'm not convinced they have less utility (I may be cheating here by including giving them back some weak light spells). For example, Dimension Door is a more powerful teleport spell than they had before. Leaving light spells aside, what do you see as the big losses? I'm guessing Haste Self and Rift, and maybe Word of Destruction?

With the banishment nerf, are you talking about the fact it doesn't affect vault monsters any more? I think that's an improvement - banishing everything and then walking in and picking up the loot doesn't seem like fun to me. Note that I've also made Destruction not affect vaults - and in return, I (haven't yet but) will stop it destroying artifacts.

I am not one to nitpick balance, but a assumed at least a sanity check would be made. I cant see that happening at all. With the druid shapes, I completely missed that the 1st one, fox, actually had a damage boost (about the amount I suggested for the last late game shape), but the later shapes completely fall off compared to the fox. And the reaction is to nerf the fox, too.

The idea for the shapes is for them to be useful in certain situations, not a simple "I'm always better in this form" - so they all need to have advantages and disadvantages. So designing them individually (a process which is happening by me putting something out and then everyone telling me it's either useless or too powerful ) is going to require (a) a concept that has situations where it's useful (fast and stealthy but weak, or slow but hard to hurt) and (b) a lot of balancing. At this point we're early in the balancing stage.

So are mages all going to be challenge classes ? If so, why bother making so many of them ?

Plan is for all the classes to be challenge classes

Seriously, it's balance again. This will take time, and the current aim is to get some rough classes in place, tweak them for balance for a bit, and then start doing the monster list changes, which will require more balancing then between player and monsters. I don't know if you've noticed the current class experience penalties

Why have so many is because they should feel different to play, and Angband has an amazing array of possible effects at its disposal which I think are really underused (you only have to look at how creative some variants get). Note that I'm not suggesting Angband should go into wacky variant territory, but I think the difference between Vanilla and Wacky should be balance and robustness, not creativity.

tldr - it's a work in progress. Keep complaining, it helps