[Announce] Poschengband 3.0.0 Released

[Arjen]

Btw Chris,

The Blade of Chaos of Sharpness you get when starting as a Servant of Korne doesn't have the vorpal_flag. Is that on purpose? If it is, why not make it a weapon of Armageddon?

[chris]

A minimally tested new release is available here. I only recommend upgrading if you like any of the following changes:

* Strong sensing of cursed items has been restored where appropriate. Some of the feeling names have been changed. If you turn on the Easy Destroyer ('=' '6') and enable the leave known worthy items option, then anything that pseudo-ids as 'bad' will get destroyed automatically. These options are on be default.

* Weakly sensed non-cursed items are now labelled as 'enchanted' rather than 'uncursed'.

* The autopicker ('_') now has a verb for cursed items (called 'cursed'). This includes anything known to be cursed. Here are some samples:

#Destroy cursed items that are not egos or artifacts
!cursed nameless items

#Leave cursed ego items on the floor.
~cursed ego items

#Pickup cursed ego weapons whenever dd*ds >= 30
cursed ego weapons more than 29 dice

* Fire Elementals can teleport onto deep lava. Their fiery touch will no longer destroy scrolls of destruction.

* Clouded provided an awesome new Light Town redesign. If you are put off by the wilderness in Poschengband, then give this option a try!

* The Vampire Lord in The Arena now drops a scroll of mundanity.

* Aggravate, Permanent Curse and Ancient Foul Curse are now much more rare, and should no longer generate on weak cursed ego items.

* Back to the autopicker, corpses and skeletons now have separate names to support the possessor (who only cares about corpses, after all). If you don't use the default pickpref.txt, be sure to change your lines for wanted corpses to

wanted corpses#Wanted!k!k!k
wanted skeletons#Wanted!k!k!k

For a possessor, try something like (note order matters!):

#always pickup wanted corpses
wanted corpses#Wanted!k!k!k
#leave all corpses on the floor whose monster level is >= 30
~corpses more level than 29
#destroy any others
!corpses

Of course, you will want to change the level threshold (29 above) throughout the game.

* The demigod Loremaster talent now functions the same as the Stone of Lore

[chris]

I'll consider further pseudo-id changes for a future release as I'm trying to take a break from coding for a bit.

The current weak pseudo-id behavior is historically consistent but I wouldn't mind changing it. Currently, whenever I play a priest, I simply rest for N turns and then destroy anything in my pack that did not pseudo-id in the allotted time. N can be decreased with level. Finally, 'uncursed' doesn't make much sense now that you mention it. Let's call this category 'enchanted' instead.

As for jewelry, many classes have sensing (rogues, mages, priests) and many do not (warriors, etc). It really is all or nothing with regard to jewelry as anything not average is at least an ego item.

But be warned, if I do make pseudo-id changes I'm likely to rollback the CL35 instantaneous underfoot sensing to respect class restrictions (currently, priests get upgraded to strong pseudo-id). Also, I'm considering some sort of drawback for The Stone of Lore as it really stands out at the moment. Perhaps it should only pseudo-id underfoot and regain its perilous activation? Finally, I'd like the sensing to take account of class specific favored items (probably by leaning on the information in the weapon skills tables). For example, a mage should get strong sensing on wizardstaves and maulers should have no sensing on daggers. Ninjas should not sense weapons they cannot wield. And so on.

[Avenger]

One odd thing about weak pseudo id is average items. They have no ID sense whatsoever, which is... somewhat unexpected, or at least it was the first time I encountered it as a priest in the latest MAngband.

I was rather surprised to find out that I could be sure that all uncursed items were actually good or better, and even more surprised that average items never got a feeling no matter how long you hold them.

I would personally be more comfortable with either the reintroduction of the average feeling to the weak pseudo-ID scheme, or the extension of uncursed to include average items as well... having an item that never id's at all feels very unusual to me - but that may just be me.

Additionally, what are your thoughts on extending a weak pseudo id for jewelry to classes that currently have a strong pseudo for armor/weapons?

What I've noticed is that, in PosCheng, there doesn't seem to be a good sense for jewelry, and I don't believe(though I could be mistaken) that there are any jewelry ego types. I suppose that, if both of those things are true, giving other classes a weak pseudo ID would be rather pointless, since it's effectively the same as strong pseudo ID(barring artifacts which may still ID as special - I have yet to find them, and I don't actually know if you left room for artifact jewelry the way you've overhauled the jewelry system).

[chris]

Chris, what are your thoughts on reinstating multiple levels of pseudo-id for cursed objects?

Keep in mind that not every class has high level pseudo-id. Priests/Monks/Mages etc will still pseudo-id cursed egos and artifacts as {Cursed}, so having these automatically destroyed is problematic. Having them not be automatically destroyed leaks information, effectively rendering the pseudo-id strong, at least wrt cursed items. This was the main reason for the change and I somehow managed to talk myself into the logic that cursed items mask player's knowledge as part of the curse, at least for classes with strong sensing.

I'll try to come up with a different mechanism, and might just give everybody strong sensing of cursed items. After all, a priest should be able to tell if a curse is minor or really bad!

Edit: Here is my original code comment for the change:

/* CTK: Worthless and Terrible may now actually be useful. The problem is, that
 *      the game often tries to auto destroy all cursed objects, and this is rather
 *      hard to avoid satisfactorily. One approach is to change the way unidentified
 *      but sensed objects value. However, this leaks information for weakly sensing 
 *      classes (or we destroy useful objects). Given a choice between removing the 
 *      easy destroyer altogether and nerfing heavy sensing of cursed objects, I am
 *      temporarily choosing the latter pending a more creative solution. */

Edit2: One solution would be to add a new indeterminant pseudo-id sense level for weakly sensing classes and to hack the auto-destroyer to not destroy these items. So classes with weak sensing will play as the game is currently, but classes with strong sensing can regain their powers of discrimination and automatically destroy bad items (i.e., cursed but not an ego or an artifact).

Edit3: To clarify,

/* For strong sensing, we have now have (3.0.3 and later):
 *
 *                    egos         artifacts
 *                    =========    =========
 * average -> good -> excellent -> special
 *         -> bad  -> awful     -> terrible
 *
 * For weak sensing, we have:
 *
 * FEEL_NONE -> uncursed
 *           -> cursed */

[Avenger]

Chris, what are your thoughts on reinstating multiple levels of pseudo-id for cursed objects?

[chris]

Something I've been meaning to do for a while, I have done an update to t_lite.txt, it hadn't been updated for quite a while by the looks of things. Here is a diff or here is the raw file, which ever you find easier to work with.

I changed the quests around, now there are two quest chains: Thieves > Dark Elves > The Vault > The Barrows > The Cloning Pits > The Old Castle and Orc Camp > Doom Quest I > Tengus > Mimics > Doom Quest II. This splits things into Outpost/Telmora and Morivant/Angwil, omitting a few quests. All the rewards are updated to the current ones. I replaced the Thieves reward with the current Wargs reward where appropriate and replaced the reward for the mimic quest with a rod of detection because I didn't want to put the vapor quest in.

I altered the services a little, combining a few in order to make room for a few absent ones, the arena and the gambling shop, which now appear. I adjusted the prices a little to make it similar to Morivant's prices.

I redesigned the map to make sure all the services were within the town walls and took care to put the most used ones right next to the home. The tower of sorcery being a million miles away from everything was the main reason I wanted to do this.

Berserkers rejoice!

Very nice ... I confess that I seldom think about light or vanilla town modes in Poscheng, and the former has become rather out of date. Your changes look awesome! Thanks for bringing in the Arena, Gamble and Reforge options as well as The Cloning Pits quest, all of which were sadly lacking before

I pushed your changes to the current Develop branch. Next release is pending an attempt to redress some of the {cursed} criticisms mentioned above in this thread (when I finally get back into coding mode ...).

[clouded]

Something I've been meaning to do for a while, I have done an update to t_lite.txt, it hadn't been updated for quite a while by the looks of things. Here is a diff or here is the raw file, which ever you find easier to work with.

I changed the quests around, now there are two quest chains: Thieves > Dark Elves > The Vault > The Barrows > The Cloning Pits > The Old Castle and Orc Camp > Doom Quest I > Tengus > Mimics > Doom Quest II. This splits things into Outpost/Telmora and Morivant/Angwil, omitting a few quests. All the rewards are updated to the current ones. I replaced the Thieves reward with the current Wargs reward where appropriate and replaced the reward for the mimic quest with a rod of detection because I didn't want to put the vapor quest in.

I altered the services a little, combining a few in order to make room for a few absent ones, the arena and the gambling shop, which now appear. I adjusted the prices a little to make it similar to Morivant's prices.

I redesigned the map to make sure all the services were within the town walls and took care to put the most used ones right next to the home. The tower of sorcery being a million miles away from everything was the main reason I wanted to do this.

[getter77]

What about some sort of Rolling ID metric to further reckon things? IE: When you've ID'd or especially *ID'd* some random artifact or whatnot, the type of modifier(s) become known to you, at least on that character, going forward----from there you'd certainly still have a great variety to "discover" but then there'd be no retreading and it could allow for more strategic use of the full ID since situations would come up where your partial info could make a difference between discerning an unknown power versus unknown powers.

[Nick]

You absolutely do not need to *identify* every object in the game. You absolutely do not need to never miss a particular goody in order to win the game. You will never clear every level, nor clear every vault, nor slay every unique, nor kill every monster you meet, nor find every artifact nor every once in a blue moon awesome object. It's not possible to do so, and not even remotely necessary for success. Optimal play is not required (nor even possible, IMO)!

This is great stuff, and IMHO shows the essential difference between V and the Z-evolved variants (particularly the Hengband ones). V is certainly a big, chaotic-seeming game, but what you're aiming to do is impose order on that chaos - collect yourself a kit that covers every eventuality, at which point you're ready to win, and you're kind of done. In Poscheng, though, you're never done - you're always a small fish in a big pond.

[Avenger]

Sound reasoning certainly makes it easier to swallow. Thanks for the post, Chris.

Trying out the keep-the-cursed-stuff tactic, but I only get {cursed} not the other types, new update?

I have also noticed this - even cursed randarts still pseudo-ID as cursed, rather then terrible.

In a variant where cursed egos can occasionally be worth using(my current character is using a cursed Morningstar at the moment), differing pseudo-ID indicators would be quite helpful.

As Arjen stated, keeping {cursed} on autodestroy simplifies things a lot, and prevents you from bringing a bag full of crap up in the hope that it might be worth using. Why not give us worthless and terrible back, so we can keep basic cursed stuff on auto-k, and preserve the egos in case they turn out to be worthwhile? Those of us that don't want those, either, can put {worthless} on auto-k, and of course arts will not be destructible, so those will still stand out no matter what.

[Arjen]

Well, I still like poschengband. Just takes a lot of time so not playing a lot lately.

Trying out the keep-the-cursed-stuff tactic, but I only get {cursed} not the other types, new update?

And still, this is *your* game as you are the developer, so you can actually do as you like.

And i like the hidden powers too, and I also only *id* in town. So I still think you are on the right way concerning the development of poschengband. Just that I might keep the auto destroy on {cursed} if I keep disliking it. Altho the increased packsize helps.

[chris]

chris: is there something you find distasteful about the way Vanilla handles things right now? Which is to say, Identify is *Identify*. Functionally it sounds like you're moving towards this anyway by making scrolls of *Identify* more common.

Yes, I do indeed find this change objectionable.

Regarding object identification, the historical system is not perfect. The mechanics of pseudo-id are simply crazy when object drop rates are high. Spending large amounts of time pack shuffling and then resting for long periods of time in the dungeon is just nuts, and some classes needed to rest for entire game days before pseudo-id would kick in. So I previously changed the pseudo-id mechanic very heavily (It's now much quicker and becomes instaneous and, more importantly, happens "underfoot" once the player reaches CL35). Variants like Zangband that ratcheted up the object drop rate have been forced to make regular Identify much more common, nay, universal, to compsenate. This is OK, and pseudo-id is only really important in the very early game when gold is scarce, and later on as a means of directing the application of plentiful identification (alas, carrying more that 20-50 identify is impractical).

So, regular identify is plentiful, and needs to be so. It is also cheap. Sure it is subject to fiery assaults of frustration, but this is seldom crippling.

Having an increased number of hidden powers was an intentional game design decision for this release, albeit one that some players may not like. Personally, I have always found hidden powers to be enormously enjoyable in games like Hengband and something that is amazingly conducive to replayability. Having played Hengband as much as I have, over a period of 10 years along with hundreds of winners, I still find the game extremely fun. I can't say the same for Vanilla.

Having a game which favors object variability (and hence replayability and, at least for me, fun), Identify is simply too common and too cheap to reveal all object attributes at once. Having the player unsure of what an object does leads to strategic decision making. Even before the changes that I have made, this was a strong component of every game I played in Hengband, especially, as Avenger mentions, after clearing a large vault. There is some skill and a good deal of luck in deciding what to keep and what to discard after such a situation, and simply removing this aspect from the game entirely is one instance of simply making the game easier in response to complaints of tedium. It is taking the easy way out. Having ubiquitous Identify that reveals objects completely is too much, in my opinion. To overstate my objections for purposes of rhetoric, why not just have objects spawn completely identified to save players the tedium? The same arguments to removing *identify* apply here as well (in a matter of degree).

You absolutely do not need to *identify* every object in the game. You absolutely do not need to never miss a particular goody in order to win the game. You will never clear every level, nor clear every vault, nor slay every unique, nor kill every monster you meet, nor find every artifact nor every once in a blue moon awesome object. It's not possible to do so, and not even remotely necessary for success. Optimal play is not required (nor even possible, IMO)!

Personally I never use *identify* at all within the dungeon except on rare instances when I stumble upon a stack of scrolls. I also rarely find myself in situations of object overload except in the early game when the player is literally showered with all manner of flavors of mushrooms, potions, wands, rods, staves, scrolls, etc, but this is not an issue with *identify*. There is, of course, the occasional large vault, but that only happens a few times in a game.

All of this is, of course, just my opinion. It is a lot of work in making a variant (even after starting with a mature, well designed game like Hengband) and as such, something that I will only do if I personally find the game fun.

My sincerest apologies if I wrecked the game at all recently. But this is, after all, just a game. I appreciate feedback (eventually) and often do make changes in response to it (even if I don't mention it at first ... Perma-curse, for example, has been made more rare). But I probably won't make changes that I, rightly or wrongly, disagree with.

[Derakon]

So the question then becomes, is this an important distinction? Is it important that *ID* be expensive?

Basically I'm trying to get you to analyze the game design decisions instead of just accepting them as fiat.

[Arjen]

Right, so, given the existence of *ID* in piles, what is the purpose of having both ID and *ID*?

2400 au difference. Buying 50 ids is doable at the start, 50 *id* is expensive