Depth vs. Complexity

[Nick]

If you're improving the game, what matter that someone complains?

I mean I realise that complaints and feedback can be a good barometer if you're not sure whether a change is an improvement. But there seems to be pretty good evidence that you get complaints even in the case of improvements, so it's important not to be held in thrall by them.

Moreover, I'd be really, really surprised if the devteam was being held in thrall by dos350. Also delighted.

[Mikko Lehtinen]

I recall Magnate saying something to the effect that the real problem is disagreement among the dev team. The dev team is a microcosm of the larger community?

[Derakon]

I recall Magnate saying something to the effect that the real problem is disagreement among the dev team. The dev team is a microcosm of the larger community?

That's not been my experience. Each individual dev member tends to work on what they find interesting, and when that work produces changes that can be integrated into Vanilla (not always the case!), it gets integrated. The main issue with change to Vanilla tends to be differentiating between upgrades and sidegrades, i.e. changes which are unambiguously improvements, and ones which some consider improvements and others don't. The vast majority of changes (practically everything that isn't a UI improvement) fall under the latter category, and the community arguing over where Vanilla should go is what produces most of the paralysis.

Magnate is free to come in and disagree with me, of course.

[Magnate]

That's not been my experience. Each individual dev member tends to work on what they find interesting, and when that work produces changes that can be integrated into Vanilla (not always the case!), it gets integrated. The main issue with change to Vanilla tends to be differentiating between upgrades and sidegrades, i.e. changes which are unambiguously improvements, and ones which some consider improvements and others don't. The vast majority of changes (practically everything that isn't a UI improvement) fall under the latter category, and the community arguing over where Vanilla should go is what produces most of the paralysis.

Magnate is free to come in and disagree with me, of course.

Not this time - I agree with all that. I don't know precisely which post Mikko is recalling, but I have said two things on this theme which I believe are both true:

(i) there's often little consensus in the devteam on what the top priorities are. This shouldn't surprise anybody, as it's impossible for us to have this conversation without being mindful of what we'd enjoy working on, and it's lame to say "this is a real priority but I'm not interested in doing it" - so we don't say that. So we each acknowledge that our priorities are what we want to work on, and there's never more than an agreed wishlist of objective priorities. This is partly why stat linearisation's never been done, even though it's been on the agreed wishlist for a long time - nobody fancies it.

(ii) there's often little consensus about the best solution to a problem. Again, the best solution is the one somebody bothers to code and test and improve until it's worthy of a pull request. Again, we rarely say "I'd do it like this ... but I'm not going to" - though of course we all share ideas in case someone wants to pick them up.

[fizzix]

This is partly why stat linearisation's never been done, even though it's been on the agreed wishlist for a long time - nobody fancies it.

If you mean ditching the 18/xx scheme, then yeah, I agree. If you meant making HP and SP scale more linearly with stat gains then I disagree. To the point where I have worked on rebalancing it for my own play version and have just not bothered to even suggest porting it to V because I just don't want to hassle with justifying the change.

[dos350]

The only D&D version with 18/XX was AD&D. Even there, it was only used with Strength.

did u ever play baldurs gate?

u are wrong sir, go read ur dnd game masters guide for 89 or wait u dont hav 1

also

im not complaining about improvements

im complain 4 planned destruction of a once pure game

soz no rage ty

[Mikko Lehtinen]

u are wrong sir, go read ur dnd game masters guide for 89 or wait u dont hav 1

Here's a quote from Gary Gygax:

Well,

At first blush I decided that 18 was the maximum for a human, but then to make fighters more viable, and because the concpt of degrees of strength in the 18 cap followed logically, I used the percentile measurement. As for strength over 18, any such ability is superhuman and must be magically endowed in my view. The 18/% did give the fighter a real boost

To the best of my recollection, I have never suggested percentile breakdown for stats other than strength.

Cheers,
Gary

[Magnate]

If you mean ditching the 18/xx scheme, then yeah, I agree. If you meant making HP and SP scale more linearly with stat gains then I disagree. To the point where I have worked on rebalancing it for my own play version and have just not bothered to even suggest porting it to V because I just don't want to hassle with justifying the change.

I meant the 18/xx stuff. Understood on the other.

[Magnate]

Here's a quote from Gary Gygax:

Well that's very interesting, because in a 1E AD&D book called Unearthed Arcana he defines the Cavalier class (and redefines the Paladin as a subclass of Cavalier), which most certainly does use percentile stats for things other than STR. IIRC the Cavalier gained 2d10 percentage points per clev on his/her percentile stats, which meant an increase of a whole point every few levels.

[Thraalbee]

u are wrong sir, [...] planned destruction of a once pure game [...]

All change is toxic. To one part of the population.

But it is equally true that for the others, life *without* change is toxic. This group finds that change and progress/improvements go together, even while realizing that change will actually be toxic at times. But for us, the average is positive and there is always a next release or the next thing. Sure, it may take a while - I'm still waiting for the anti-nuke device to turn up on the shelves in my supermarket - but that's the way I am wired. To me, change is the andidote to boredom and badness

[cheerleaders:] Go dev team, go!

/Mathias

[takkaria]

If you mean ditching the 18/xx scheme, then yeah, I agree. If you meant making HP and SP scale more linearly with stat gains then I disagree. To the point where I have worked on rebalancing it for my own play version and have just not bothered to even suggest porting it to V because I just don't want to hassle with justifying the change.

I'd be really interested in seeing that patch.

Now I'm not maintainer anymore (yay!) I'm a lot less conservative about V. I think V will either change, or stand still and die. V should definitely be nicking stuff that works from the amazing new variants/games that are coming out - I've not played Sil but it sounds like it's got a fair bit to borrow from, for example.

When I started making V patches, before I became maintainer, it was to address common criticisms. Actually, the first year or so of my maintainership had me cut-and-pasting every interesting idea from r.g.r.a or the forums into a series of huge text files, analyzing them and trying to work out changes that would address the underlying problems behind the complaints/suggestions. And I tried to make small changes that would address wider issues.

I don't know how wise that was in the end. But it seems that if there is a long enough gap between V releases - say 12-18 months before 3.5 is even started on - then it would be good to analyse the complaints that arise after it's bedded in. One notable example that pops up repeatedly is that the dungeon is too long for the amount of content. And another thing people have noted is that Angband gets a fair bit of its interest from bizarre and uninuitive mechanics (cf. this thread), and this is being slowly undermined by more mechanics being exposed in the UI, or the depth is getting shallower as those mechanics get simplified.

One example of this latter case is curses and identify. The identify minigame was boring and scummy. And curses were just random bad things that could happen to you if you wielded items you picked up. And I wonder now if it was even a great idea to change all of this without having something else to fill the gap in difficulity/annoyance.

Anyway, these criticisms aren't going to go away. I think V from 3.1 to now is half-way through reform, and it's in a bit of a no-mans land. I think it can try to go back, roughly, to the balance of older versions - maybe 2.8.3 or 2.9.0 - and kind of be an enhanced-UI version of that with a few new tricks but without really futzing around with the mechanics too much, and maybe have a final release after which work is stopped and it's left to fossilise. Or it can try to finish off the reform job - sort out the mechanics so that they're more intuitive, so they can combine in more interesting ways, and generally fill out the gaps left by previous changes. Maybe both - maybe V should be finalised and left, and the work in v4, or other ideas about how V could move forward, should just go in variants from here on, with no particular claim to be Angband.

As a side note, I also think that a load of items I added in an attempt to add interesting content were ultimately junk and should be removed (the dog/mouse/cat stuff, maybe some of the mushrooms). Also thematically weird.

OK, that's my brain splurge over. Make of it what you will.

[fizzix]

I'd be really interested in seeing that patch.

I have a playable fork of 3.4 with a bunch of changes in it. I haven't tinkered on it in a while, so I should probably just push it to github. I'll do that this week I think, once I look over the changelist again.

[Nick]

Anyway, these criticisms aren't going to go away. I think V from 3.1 to now is half-way through reform, and it's in a bit of a no-mans land. I think it can try to go back, roughly, to the balance of older versions - maybe 2.8.3 or 2.9.0 - and kind of be an enhanced-UI version of that with a few new tricks but without really futzing around with the mechanics too much, and maybe have a final release after which work is stopped and it's left to fossilise. Or it can try to finish off the reform job - sort out the mechanics so that they're more intuitive, so they can combine in more interesting ways, and generally fill out the gaps left by previous changes. Maybe both - maybe V should be finalised and left, and the work in v4, or other ideas about how V could move forward, should just go in variants from here on, with no particular claim to be Angband.

Thank you for posting this - it is just packed with insight.

Angband is uniquely positioned as a roguelike, and probably as a game. It is not developmentally dead like the other older roguelikes (Nethack, ADOM, Moria etc), and nor is it in the first flush of youth (Brogue, etc); the only one with a somewhat comparable development trajectory is Crawl. It also has an amazing variant culture, and a community that is constantly pushing the current boundaries.

A lot of us have a vaguely formed idea of a perfect state that Angband might reach, but those ideas are all different, and even for individuals change over time. Your idea of community development was emblematic of the Angband's history, really - brilliant, productive, chaotic. And while there is a lot of half-finished stuff in the 3.4, I think it's a better game overall than 3.0.6 (and to test this, I'll run a 3.0.6 comp next...)

So currently we have 3.4.1 (becalmed), v4 (abandoned), Pyrel (embryonic), and the usual raft of variants (and Sil, which is really more new game than variant). I don't know what happens next, but I'd kind of like to rule out fossilising as the only option

[Antoine]

My belief is that the true spirit of Angband lives on only in Poschengband

A.

[Magnate]

My belief is that the true spirit of Angband lives on only in Poschengband.

Wot no smiley? I can't work out whether that's gnomic, provocative or just stupid.

I agree with much of what Nick said, though it's tragically ironic that Sil shows us all what could be achieved without any of the codebase improvements made in the last four years of collective maintenance. Much of my own motivation (and I think that of some other devs) was to make the game more accessible to new developers and variant writers, as well as improving it for players. As takkaria points out, these aims weren't necessarily congruent. I think this is why I prefer to work on Pyrel, which doesn't have any players who can complain.