What's happening with V4?

[TJS]

How is V4 going? Has the experiment been a success?

Are there any versions about I could have a go on (the autobuilder link in the faq doesn't work) ?

[Magnate]

How is V4 going? Has the experiment been a success?

Are there any versions about I could have a go on (the autobuilder link in the faq doesn't work) ?

Ah, two different issues here. First is that buildbot is down, which is not specific to v4. It's owned by d_m, who's been super busy with RL for the last few months - the rest of the devteam do have access to it, but none of us is quite confident enough to get it back up. (It's never gone down before so it's not a familiar process.)

Second is that development has pretty much stalled completely. We're roughly at v4's first birthday, and in the last six months there have been almost no changes at all other than those ported across from V (bugfixes and UI improvements). Fizzix put in some immunity changes for testing about a month ago, but that's it.

The main reason development has stalled is Pyrel. The three main contributors to v4 have been me (item generation), Derakon (combat) and CunningGabe (traps). (There are other changes but those are the three big ones that make it play very differently from V.) Derakon and I are both working on Pyrel. (Gabe comes and goes but has not been around for a few months.)

It looks as if Pyrel will keep going - it now has a good chunk of the active devteam members working on it - which means it's unlikely that we'll come back to working on v4, at least for a while. Whether that constitutes success is left as an exercise for the reader.

[TJS]

Ah I see so V4 has been pretty much abandoned. Does this mean none of you work on Vanilla anymore?

How far did you manage to get? Has the rune system been implemented for example? I was looking forward to trying that.

What's this Pyrel business then? Can I play it?

[Derakon]

Vanilla doesn't really need lots of active development. Fizzix is broadly in charge of making certain that nothing goes badly wrong with it, and in ensuring that whatever subtle improvements it needs get done (e.g. tweaking pack sizes). I don't think Vanilla development is going to really stop, but it certainly has slowed down, as it does from time to time. That's fine; we've had some pretty active development these past few years and it's not a bad thing to let it rest and figure out what actually really needs to be done.

v4 has rune-based ID, and it works pretty well. It's not absolutely finished, since ultimately I believe the idea was that if you knew all the runes on an item then it would be automatically identified -- this doesn't currently happen because pluses are not learnable without using the item or using Identify. But the fundamental concept has proven itself workable.

Pyrel is a ground-up rewrite of Angband in Python. It's not playable yet, but we're making steady progress on it. I'd say that in order to have a meaningfully playable game we'd need to do the following:

* Finish up item allocation
* Give brains to the monsters
* Code in effects for all the monster spells, item effects, and player spells

That leaves us with the current rather slapdash dungeons, but it'd be enough to make a playable game. It'd be a far cry from Angband proper though. And it's going to take a long time before Pyrel will be ready to take its place in the Angband pantheon.

[TJS]

Interesting.

What are the goals for pyrel? Is it going to be the new Vanilla or more like V4?

[Derakon]

Ideally the new Vanilla, though it's basing itself off of v4's current codebase. Pyrel is not going to try to exactly match the gameplay of either Vanilla or v4; if nothing else, some aspects of gameplay are clearly implementation warts (like the fact that you can "see" doors get opened even if you don't have line-of-sight with them).

The hope is that the changes in v4 become uncontroversial enough that they become "official". Of course, Vanilla itself isn't going anywhere, and I wouldn't be surprised if it continues to get updates even after Pyrel is fully playable.

The main motivation here is that Vanilla's codebase is pretty old -- the first version came out over 20 years ago, and while it's received some significant overhauls in that time, it's really showing its age. Making modifications is harder than it really needs to be, which means that variants aren't as numerous as they could be (and, let's be honest here, variants are what keep the community thriving). So a new design that strives to keep things modular and is written in a more friendly language than oldschool C should be a big help.

[Magnate]

Ideally the new Vanilla, though it's basing itself off of v4's current codebase. Pyrel is not going to try to exactly match the gameplay of either Vanilla or v4; if nothing else, some aspects of gameplay are clearly implementation warts (like the fact that you can "see" doors get opened even if you don't have line-of-sight with them).

The hope is that the changes in v4 become uncontroversial enough that they become "official".

Given the stupendous controversy over V development prior to the creation of v4 (indeed, the whole reason for v4), I have my doubts about this.

But yes, Pyrel is being developed with v4's combat system and v4's item generation system, not V's. The latter can be tuned to give a good approximation of the V system (you'll find the same items at the same frequencies), but combat can't really, so it will differ from V in at least one fundamental way.

[TJS]

I think this is a good idea, I've always wanted to make my own variant, but been put off by the old codebase.

Regarding changes and controversy, I don't think that you should try to stick too closely to the current Vanilla. Sounds like you're putting in huge amounts of effort so you should be able to make some gameplay changes, since that is the most fun part about remaking something.

Having said that changing everything for the sake of it can be a recipe for disaster.

What are the combat changes exactly?

I never understood quite why people complained so much about changes, especially since the older versions of the game are still available.

By the way are you looking for people to help with this new version? My day job is a games designer/gameplay programmer, although I'm probably not the best programmer going when it comes to engine stuff.

[Derakon]

What are the combat changes exactly?

You can download v4 if you want to check them out, but in brief, to-hit and to-damage bonuses are replaced by "finesse" and "prowess". The former governs how many attacks you get, and the latter governs how much damage they do. Different weapons have different biases towards those stats ("balance" and "heft"), so e.g. a prowess-oriented character is unlikely to want to use a dagger (because he doesn't have the finesse to get lots of hits, and his mighty blows are not especially noticeable with it), while a finesse-oriented character is unlikely to want to use a warhammer (it's not well-balanced enough to get many blows, and he lacks the prowess to get powerful blows). Ideally the system should be balanced such that each "style" gets equivalent damage, with the finesse characters being better against agile enemies and the prowess characters being better against heavily-armored enemies.

I never understood quite why people complained so much about changes, especially since the older versions of the game are still available.

When you've been playing a game for over ten years, you tend to have opinions on how it's meant to work. While you're correct that the older versions are still available, the common response to that argument that I've seen is that they like 80% of what the new version has done, but the remaining 20% is a dealbreaker.

Fortunately, with Pyrel it ought to be easier to make a "variant" that just has the 80% you like!

By the way are you looking for people to help with this new version? My day job is a games designer/gameplay programmer, although I'm probably not the best programmer going when it comes to engine stuff.

Absolutely! Pyrel's an open-source project, and contributions are always welcome. We've been discussing development in the Variants forum (search for "Pyrel dev log") and on IRC (irc.freenode.net, #angband-dev). The repository is here. If you're looking for a place to get started, there's a to-do list on the wiki here. I'm also working on a high-level overview of the code structure here, though it's as-yet incomplete.

[Magnate]

Regarding changes and controversy, I don't think that you should try to stick too closely to the current Vanilla. Sounds like you're putting in huge amounts of effort so you should be able to make some gameplay changes, since that is the most fun part about remaking something.

Having said that changing everything for the sake of it can be a recipe for disaster.

Yes. Pyrel is very much Derakon's project - with help from increasing numbers of the rest of us - but my impression is that he's keen to create something as close as possible to V *to start with*. If we can prove that the code can re-create the V experience, yet still support much easier and fuller customisation and extension, we'll have achieved his goals, I think.

What are the combat changes exactly?

Weapons have two new statistics: balance and heft. The former is how easy the weapon is to attack with (and therefore influences blows), and the latter is how hard it hits (and therefore influences damage).

Characters have two new statistics: Finesse and Prowess. The former is how good a swashbuckling duelist they are, finding gaps in enemy defences - this determines to-hit and blows. The latter is how good they are at driving hard into vulnerable places, and determines damage.

So the blows calculation is now simply Finesse * balance (multiplied by some scalar to keep blows in the 1-5 range) and Prowess acts as a damage multiplier for the weapon's dice.

A weapon's "to-hit" and "to-dam" values now add to the wielder's Finesse and Prowess, respectively.

EDIT: I forgot to say that monster AC has been split into evasion (which makes them harder to hit) and absorption (which makes them harder to damage). So Finesse helps vs evasion and Prowess helps vs. absorption.

You can see that although this is both elegant and simple, it's very, very different from V's system. (The goal was to get much closer to O-combat but avoid the nasty diminishing returns in that.)

I never understood quite why people complained so much about changes, especially since the older versions of the game are still available.

Angband has a very long history and a lot of people feel a sense of ownership of it. We created v4 precisely to get away from this and avoid damaging the thing they know and love.

By the way are you looking for people to help with this new version? My day job is a games designer/gameplay programmer, although I'm probably not the best programmer going when it comes to engine stuff.

All help would be hugely appreciated, especially if you already know any Python. Please get an account on Bitbucket if you don't already have one, and fork derakon/pyrel. If you use IRC, come to #angband-dev at irc.freenode.net and say hello!

[Mikko Lehtinen]

Angband has a very long history and a lot of people feel a sense of ownership of it.

It's important for the developers to not internalize this horrible sense of responsibility to a paralyzing extent.

Don't think "Do these changes offend someone", but rather "Do these changes offend my own vision of Angband". It's enough if you devs agree to do something, even if it would make some other people angry.

It's alright to cause controversies from time to time. That's the only way to create the best possible Angband.

[Antoine]

Don't think "Do these changes offend someone", but rather "Do these changes offend my own vision of Angband". It's enough if you devs agree to do something, even if it would make some other people angry.

That's the thinking that led to 3.2. I like the new thinking better...

A.

[buzzkill]

Angband has a very long history and a lot of people feel a sense of ownership of it.

This line reminded of a documentary I recently watched. I can't remember the name of it (it's on Netflix), but it was about George Lucas and (predominately) Star Wars. It's not just about the dissatisfaction with the prequels, but also the "enhanced" original movies. Hardcore fans hate the enhanced remakes because they feel some "ownership" over the original versions (which Lucas, won't re-release on DVD, only fragile VHS copies exist) and don't care for many of the changes.

[Mikko Lehtinen]

That's the thinking that led to 3.2. I like the new thinking better...

Note that I was assuming that most of the devs are traditionalists to a degree, and that they must agree among themselves. It's impossible to make everyone happy, and paralyzing to aim for that.

[fizzix]

Note that I was assuming that most of the devs are traditionalists to a degree, and that they must agree among themselves. It's impossible to make everyone happy, and paralyzing to aim for that.

It's remarkably difficult to get agreement from all the devs.