Feelings about recent "development"

[jens]

Hi Timo!

Yes the game is changing fast. I have also been following Angband since the early 90's. I used to come back to the game every 2 years or so, play for a month, then get on with life again. Sometimes there was a new version out, and I was always happy with the improvements.

The last few years development has been quicker, I even started actively checking to see if there was a new version out instead of just waiting for the urge to play Angband again. I find the current situation with development is the most exiting time to be a fan of Angband. Sure, some changes I don't agree with, but all in all the game is getting better.

I agree with you that when changes happen on top of each other it gets harder to balance each individual change. But I don't really see that as a big problem, the changes will be tested and balanced over time, then new changes will unbalance the game... I think were we differ is that I believe that rapidly improving Angband, with the side effect of it being constantly somewhat unbalanced, is a good thing for the game.

To ease your mind a bit before the coming of 3.3: If you check the issue tracker roadmap, you can see that the devs have decided on a release date for 3.3 (16/7). There is also indication that they are concentrating their efforts on bugs and game balance until the release. Two months to get the release in as good a shape as possible is good enough for me, then they can start breaking game balance again :-)

[Tobias]

I also think the changes have been mostly positive.
But yea, some changes are sitting around half finished. Like the message.prf interface. Got put in overnight a year or so back, and it is still totally undocumented and some features are missing.

[takkaria]

I also think the changes have been mostly positive.
But yea, some changes are sitting around half finished. Like the message.prf interface. Got put in overnight a year or so back, and it is still totally undocumented and some features are missing.

If y'all want to continue listing unfinished / unpolished / unbalanced stuff, that would be great (more bugs to file, more things we know to fix). The devs all need to have a chat today about this thread and we'll put out a decent reply later.

[Tobias]

If y'all want to continue listing unfinished / unpolished / unbalanced stuff, that would be great (more bugs to file, more things we know to fix). The devs all need to have a chat today about this thread and we'll put out a decent reply later.

Things I can remember getting annoyed at recently:
You can't set colors for monster resists or monster dies to messages.prf. And that the file even exists is a secret known to few.
The K key is documented nowhere.
Flush keys on disturb becoming required broke a lot of my favorite macros, like m1a5'\e .
Caves are annoying to search, needs either easier lighting or a change to the run code.
!CLW and CSW are too common in the dungeon and too rare in the shop.
There is the annoying bug I reported about the Alt Gr stuff.

[takkaria]

Hello all devs and maintainer.

With all the changes game is going thru I have a feeling that vanilla angband stopped being vanilla angband in 3.2 and now people are developing a variant and vanilla angband is no longer maintained. Too many changes with too little testing and too few opinions asked. That's no longer community-based development, it is few people shared view variant development.

Takkaria, I think you are still maintainer, right? Stop this madness. For a year or so after 3.3 is released freeze the development and allow only bug fixes. It feels like Robert was the last maintainer, not you. Now things have been taken over by committee that does whatever they feel like doing.

OK, so this is a joint response from the devteam.

So the versions of Angband leading up to where we are now are amongst the most well-discussed versions we've had for a long time. The forums are hugely active, and while we make a lot of changes, we also get a lot of comment. I think we're pretty responsive as a team. Most bugs get caught before release (though we do need to start doing proper bugfix releases, so we can commit to releasing a 3.3.1 bugfix after 3.3.0).

We've had some discussion amongst ourselves (magnate, myshkin, d_m, and myself) and we don't agree that the pace of change is a problem in itself. The reason it has been a problem is because we've not been particularly co-ordinated and have not been careful enough about managing what goes in the game and at what point in the dev cycle.

For 3.4 we will be a lot more careful about what goes in and where and we're going to establish a collective vision about where we're going. We haven't done this before, so we'll try it and see how it works. We'll introduce big changes at an earlier point in the release cycle so that there are no surprises. We will also be focussing specifically on balance and getting it better.

We're not going to roll back to a maintenance mode. We've taken note of your concerns (some of which we share) and we're going to work better to fix them rather than just backing off. If we switched to maintenance mode, it's hard to imagine the angband developer community being excited about that, and I would imagine the active coders would die off a lot.

[Zababa]

I was happy to read what the Angband dev team has responded. I am also very excited about the current development roadmap.

Maybe there is some experimenting going on with the game balance (not noticeable to me, I still haven't won the game), but I have the impression that the Angband's code is becoming cleaner and easier to handle, which is more important in my oppinion.

It will be much easier to tweak the game balance once the game code is written in a way that when you need to change/tweak something, there will be ideally only one place where you need to do it.

Although I haven't had a single look into Angband's code I imagine that after all the years of "maintenance" and "bugfixing", people had been adding many things to the same code framework, adding them one by one, over long periods of time. While they may have seen parallels in the micro-structure of the code, they never actually cared about coding the abstract thing in one place and refer to it in the concrete cases where it is used. Instead, the same (or similar) thing was programmed for each case individually. (So, for instance, we ended up with two kinds of behavior of spells affecting the gear: Enchanting waits for you to choose what to enchant (as does identifying), but Removing curse does not.) It's things like these, just more internal to the game, they may not have any effect on the UI at all, but all such things which have something in common, should refer to and make use of one piece of code.

I love the milestone of separating the core game code and the UI. It is a lot of work which may not have any immediately visible impact, yet it should make it significantly easier for the future, when something of the UI should get changed/unified. Porting to handheld's, portables, Pads, etc. should be mainly just some cosmetic work on the UI, no need to work through the whole code.

So far I can see that Angband's current dev team has the best intentions to make a clean and well organized code for the future. There is supposedly a lot of mess which needs to be cleaned up. Some game balance issues which may have arisen are really peanuts in comparison to the actual work these guys do. As said, a clean and well-structured code is essential for easy balancing work. Nobody will try to tweak something in the game it would require to work through 60% of the game code. But if it's just modifying the values of the arguments for some funtction in a place, it will be a piece of cake.

[Nick]

I made a flippant response to this yesterday, which is better forgotten - here's a more considered one.

The big different thing that takkaria has done as maintainer is to change from single maintainer to devteam. The big advantage is that there are now several smart, keen programmers working (apparently almost full time) on Angband.

This doesn't come without difficulties, though. These include:

• Probably only takkaria in the devteam has good understanding of the entire codebase, which leads to bugs due to unforeseen consequences of changes. These usually get cleared up fairly quickly; also, the code is gradually becoming more modular, so changing stuff locally without effects elsewhere should be easier.
• It's easier to avoid doing the boring stuff in the belief that someone else will do it; happily, almost every task is considered non-boring by someone.
• Changes can be happening on several fronts at once, which looks chaotic; moreover, sometimes these changes sometimes interact in unexpected ways.
• Smart, keen programmers tend to be batshit crazy.

There were a lot of things that needed modernising, improving, or just changing, and a lot of it is all being done at once. While I can see the difficulties with that, I thnk it's much better than having a drawn out change process. It's been chaotic for a while, and will be for a while longer, but I think the aim is to get the game into a state where it is more robust and stable.

So I agree with Timo that the game is improving, but I also think that with the amount that needs changing, fast is the right speed.

[Antoine]

It will be helpful to bear a few things in mind:
- need to think very seriously before making a gameplay change which makes the game easier (ego-DSMs, offweapon art +blows/shots)
- need to think very seriously before changing an aspect of gameplay that has been around for a long time (pointy penalty)
- need to be cautious about introducing a significant gameplay change without prior announcement or discussion (Destruction destroys arts, which i think is what set Timo off in the first place).

A.

[AnonymousHero]

It will be helpful to bear a few things in mind:
- need to think very seriously before making a gameplay change which makes the game easier (ego-DSMs, offweapon art +blows/shots)
- need to think very seriously before changing an aspect of gameplay that has been around for a long time (pointy penalty)
- need to be cautious about introducing a significant gameplay change without prior announcement or discussion (Destruction destroys arts, which i think is what set Timo off in the first place).

I think you may be underestimating how easy it is to predict whether change X "makes the game easier" or simply "adds flavour". There are some obvious ones, like the +blows you mention, but if my understanding is correct this was the indirect result of a broken randart power evaluation calculation -- so it's not like someone decided to explicitly add off-weapon blows for the heck of it.

[Antoine]

I think you may be underestimating how easy it is to predict whether change X "makes the game easier" or simply "adds flavour". There are some obvious ones, like the +blows you mention, but if my understanding is correct this was the indirect result of a broken randart power evaluation calculation -- so it's not like someone decided to explicitly add off-weapon blows for the heck of it.

I'm sure you're right that it is not easy to determine whether a change makes the game easier (this in itself is a reason for caution, right?)

Re off-weapon blows, I think we are talking about different incidents - I was referring to the change to various fixed arts (like Haradrim), which was intentional but has now been largely reverted.

A.

[Derakon]

As far as I know, Haradrim's had the extra blow on it ever since it was first created, back before Takkaria took over. It's been unbalanced for years.

[Antoine]

As far as I know, Haradrim's had the extra blow on it ever since it was first created, back before Takkaria took over. It's been unbalanced for years.

My mistake sorry - was it the extra shot that was added (and reverted) more recently?

Anyway, my point is that some kinds of changes are more palatable than others. (Of course you well understand that already.)

A.

[Timo Pietilä]

Things I can remember getting annoyed at recently:
You can't set colors for monster resists or monster dies to messages.prf. And that the file even exists is a secret known to few.
The K key is documented nowhere.
Flush keys on disturb becoming required broke a lot of my favorite macros, like m1a5'\e .
Caves are annoying to search, needs either easier lighting or a change to the run code.
!CLW and CSW are too common in the dungeon and too rare in the shop.
There is the annoying bug I reported about the Alt Gr stuff.

(without thinking much).

Editing visuals (saving and editing at different places)
Also editing visuals doesn't allow editing all visuals, for example
# Permanent rock (dark red : solid block)
#F:60:all:4:127
F:60:dark:0x12:0x02
F:60:lit:0x16:0x02
F:60:bright:0x12:0x02

was possible only from pref-files (blue permanent walls instead of red, and walls that are highlighted by torch lite lighter shade of blue than rest).

Keymap 5 doesn't do anything unless you have numlock on.
I think the overhead screen blinking with monster that changes color is in view is still unfixed.
Documentation. Especially inscriptions, keymaps and pref-file syntax.
Is it just me or did the dungeon got smaller with new dungeon generation? A lot smaller? Which means more boring levels.
Healing potions not in the temple, but in the alchemist. Heroism still in temple, and not in alchemist.
Looking at the artifact.txt and object.txt I see that Elessar is made useless (Trickery is better in every way) and many of the artifacts have been weakened (rings of power, gauntlets), rendering especially top artifact gauntlets inferior to good ego-gauntlets and rings of power inferior to damage/speed rings. OTOH Haradrim still offers extra blow and many weapon artifacts speed, and Paurnimmen still has weapon branding.

(haven't played dev version much because my macros don't work properly in there).

[Timo Pietilä]

As far as I know, Haradrim's had the extra blow on it ever since it was first created, back before Takkaria took over. It's been unbalanced for years.

No, first incarnation (2.9.4) had only

F:ACTIVATE |
F:STR | CON | SUST_STR | SUST_CON | HIDE_TYPE |
F:RES_FEAR | RES_BLIND | RES_POIS | AGGRAVATE
A:BERSERKER:50:0


And no combat bonuses.
combat bonuses, blows and shots have been inserted as recently as in version 3.1.2 (not yet in 3.1.1, but in at 3.1.2).

[Timo Pietilä]

OK, so this is a joint response from the devteam.

I guessed that you wont agree with me. It was worth saying I think, maybe it did raise some concern about what is happening. If it did, then I consider this as success.

Back to figuring out how to clear -more- prompts with keymaps....