So what else should we revert?

[Derakon]

I'll accept that I'm an outlier. I'm not really playing Angband much right now, so detailed testing isn't in the cards. I'll shut up about randart availability for a bit.

[Magnate]

OK, going way back to the original intent of the thread, here is what I (in my humble and not necessarily correct opinion) think should be reverted:

(snip)

The other part of Angband that, personally, I have no idea where it is going is with the object list. I know it can't or won't be undone, but since Angband is an inventory management game, it makes such a difference in game balance. From reading feedback the forum it seems like objects and object generation are in the middle of a complete re-write. I haven't looked at it enough to know what is going on or to have an educated opinion of the quality of the changes. But I think whatever is going on there needs to be finished before Angband can be re-balanced. Changing that affects every part of the game.

Hi Jeff - thanks for that - it was very informative. I like almost all of your suggestions - the only one I disagree with is the "Of Resistance" ego. I never find it until I've already got something better, so I would deepen Elvenkind first and then see if Resistance needed any changing.

I also think you are spot on about the help files and newbies. Angband is like any other volunteer project in that what motivates volunteers is "real" changes, and writing help is not sexy. Takk has had great plans for improving the help files since he took over (look at tickets #11-14 which are all extant), but he can't find anyone to do it. If I can ever stop obsessing with randarts and power ratings for long enough, I'd love to do it - but I'm no more motivated to do so than anyone else.

I don't understand your final query about the object list. Do you actually mean object.txt, or do you mean the object list display (']')? AFAIK neither is the subject of serious on-going work: Ewert has made changes to rarities of certain items to try and address the surfeit of wearables and the deficit of consumables (which is a lingering result of the anti-TMJ changes in 3.1.0). That's all - nothing more wholesale than addressing that one problem. (There are always other suggestions like the dragonhide shields, but none of them are wholesale rewrites either.)

[PowerDiver]

There are probably others, of course. If it's not too unpalatable, would you mind playing half a dozen standart games for comparison? If you perceive that "useful" artifacts turn up in consistently earlier and in quantity, I promise to do some further research.

I'm playing with randarts, and they definitely seem much less common. I'm playing slow and careful, waiting for CL 50 before finishing, and would have finished already otherwise. I've only seen 28 randarts, and I would have expected to see 2 or 3 times as many standarts in the same time with the same kills. Given the amount of time I have had with the Palantir, more likely 3x than 2x.

Someone needs to code a utility to generate each level of the dungeon in turn, kill all monsters on each level, and accumulate statistics. I wonder how hard that would be to do.

[Magnate]

I'm playing with randarts, and they definitely seem much less common. I'm playing slow and careful, waiting for CL 50 before finishing, and would have finished already otherwise. I've only seen 28 randarts, and I would have expected to see 2 or 3 times as many standarts in the same time with the same kills. Given the amount of time I have had with the Palantir, more likely 3x than 2x.

Someone needs to code a utility to generate each level of the dungeon in turn, kill all monsters on each level, and accumulate statistics. I wonder how hard that would be to do.

That's ticket #786 - go right ahead!

Interesting. In the old days (3.1.1) the randart rarity was very closely tied to standarts. I broke that in 3.1.2 to try and make it more closely tied to the randart's power - because we were getting some uber-randarts too common and some crappy ones too rare. But I seem to have ended up with an overall increase in rarity, which surprises me (because I don't play enough standart games).

Since the base object rarity hasn't changed, it must be the calculation of a_ptr->alloc_prob in randart.c: r2017 tweaks this slightly. If you are now drowning in randarts, let me know. (Don't forget that the recent change to drop levels will have reduced the likelihood of deep artifacts dropping, for both randarts and standarts.)

[nppangband]

There are a couple of things that are "missing": skill modifiers (per level and the starting values) and it would also be nice to have the current totals displayed (I think FA does this?).

Current NPP does this, at least until I put in the new Angband birth.c. It is a fairly simple thing to add.

[nppangband]

You don't need to do this in V, the modifiers are displayed prominently...

Modifier's are displayed, but those numbers don't mean anything to a new player.

[PowerDiver]

Since the base object rarity hasn't changed, it must be the calculation of a_ptr->alloc_prob in randart.c: r2017 tweaks this slightly. If you are now drowning in randarts, let me know. (Don't forget that the recent change to drop levels will have reduced the likelihood of deep artifacts dropping, for both randarts and standarts.)

Let me check. I am playing 1996. I'll upgrade next game.

[Derakon]

Here's what I recommend for the birth screen: replace the yellow explanatory text (e.g. "Your 'race' determines various intrinsic factors and bonuses") with the race description from birth.txt, possibly slightly truncated. Those descriptions are generally very useful for telling new players "Okay, this race is probably going to be pretty bad at spellcasting; maybe I should stick to a stronger race for now." or "hey look, this race has See Invisible, which should make dealing with those poltergeists a lot easier." Do the same for the class selection.

Honestly I don't think anything more is needed. Certainly the skills shouldn't be shown, since they're horribly opaque and even veterans don't know how they correspond to actual numbers in-game. They'll just add confusion.

[nppangband]

Hi Jeff - thanks for that - it was very informative. I like almost all of your suggestions - the only one I disagree with is the "Of Resistance" ego. I never find it until I've already got something better, so I would deepen Elvenkind first and then see if Resistance needed any changing.

Maybe I have been lucky. I always seem to have a resistance armor or a shield by 1000', and poison resistance by 1500'.

I also think you are spot on about the help files and newbies. Angband is like any other volunteer project in that what motivates volunteers is "real" changes, and writing help is not sexy. Takk has had great plans for improving the help files since he took over (look at tickets #11-14 which are all extant), but he can't find anyone to do it. If I can ever stop obsessing with randarts and power ratings for long enough, I'd love to do it - but I'm no more motivated to do so than anyone else.

Oh, tell me about it. I hate the feeling of having a release all ready to go, and then realizing.....oh, wait. I need to update the help files. That and the pref files for graphics. I always rush those & do a poor job because I am so primed to do a release.

I think it would be a fairly simple thing to code. If I ever have time I will blaze the path.

I don't understand your final query about the object list. Do you actually mean object.txt, or do you mean the object list display (']')? AFAIK neither is the subject of serious on-going work: Ewert has made changes to rarities of certain items to try and address the surfeit of wearables and the deficit of consumables (which is a lingering result of the anti-TMJ changes in 3.1.0). That's all - nothing more wholesale than addressing that one problem. (There are always other suggestions like the dragonhide shields, but none of them are wholesale rewrites either.)

Definitely not object list display. UI improvements are always a great change. I mean the object generation code, and the actual object list. People are still playing with it. Not that there are big changes, but object generation is the foundation of game balance. So small changes has a big ripple effect to game balance. I guess I was trying to say there is no need to try to balance everything else in the game until the object list and generation code is finalized.

And as for "too late to go back now" comment, I was thinking of the Angband 3.0.8~9 and beyond object list, where having everything in the dungeon safe to try and is double-edged at worst, and you barely have to identify anything any more. I haven't given it a fair chance, and the grumpy old man in me is saying "it can't work".

[Derakon]

And as for "too late to go back now" comment, I was thinking of the Angband 3.0.8~9 and beyond object list, where having everything in the dungeon safe to try and is double-edged at worst, and you barely have to identify anything any more. I haven't given it a fair chance, and the grumpy old man in me is saying "it can't work".

I was definitely in the same camp when I first heard about ID-by-use, but thinking about it, what did outright bad items add to the game? Past 400' or so you'll be IDing everything before using it anyway, so it just slowed down the very early game. Bad items just mean that either you don't use un-ID'd stuff and have to return to town sooner (to sell all the new flavors you found), or use un-ID'd stuff and accept an early penalty to stats / cursed equipment. Which, if they make a serious difference, just mean going back to town for Remove Curse, which is not exactly expensive.

I tried to think of a reason to keep the bad items beyond just flavor, and I couldn't manage it. And I don't think the bad items added enough flavor to be worth keeping, compared to the increased fluidity of gameplay without them.

[fizzix]

I also think you are spot on about the help files and newbies. Angband is like any other volunteer project in that what motivates volunteers is "real" changes, and writing help is not sexy. Takk has had great plans for improving the help files since he took over (look at tickets #11-14 which are all extant), but he can't find anyone to do it. If I can ever stop obsessing with randarts and power ratings for long enough, I'd love to do it - but I'm no more motivated to do so than anyone else.

I'm certainly willing to help write the actual text. I just don't feel comfortable working with the UI interface.

[camlost]

I tried to think of a reason to keep the bad items beyond just flavor, and I couldn't manage it. And I don't think the bad items added enough flavor to be worth keeping, compared to the increased fluidity of gameplay without them.

Most of the nasty stuff in Sangband can be thrown for some potential use. And consuming them won't (instantly) kill you.

[Magnate]

I'm certainly willing to help write the actual text. I just don't feel comfortable working with the UI interface.

Please, please do this. The help files date from 3.0.8, and many from even earlier! Please just take a text editor and send me update ones, and I'll commit them. That would be a great start - and all the html and context-sensitive stuff can come later when someone wants to code for it.

[Magnate]

Most of the nasty stuff in Sangband can be thrown for some potential use. And consuming them won't (instantly) kill you.

ID-by-use was always a non-negotiable priority of Takk's, and I think he very much liked the Sangband approach where nothing is truly useless.

@jeff: I'm not convinced the object stuff will ever be "finalised" - we've made some significant changes since 3.1.0 despite a clear dearth of consumables. When Ewert's changes (or whatever equivalent) are implemented, it'll be a bit better but not perfect. I see Angband as like one of those giant puzzles, where you push one bit and something else moves out of place over on the other side. We'll be forever trying to get them all neatly balanced without anything being out of kilter - but that's the fun part.

[fizzix]

Please, please do this. The help files date from 3.0.8, and many from even earlier! Please just take a text editor and send me update ones, and I'll commit them. That would be a great start - and all the html and context-sensitive stuff can come later when someone wants to code for it.

Ok, I'll have time this weekend, so I'll get working on it ASAP. Any area specifically to start with?