TFMurphy
Level: Smitemaster
Rank Points: 3118
Registered: 06-11-2007
IP: Logged

This one's more a query with regards to how sure I am, as there are *many* monsters that have strange reactions to Stalwarts when brained.

Aumtlich always want to face their target, and will only move once they are facing it (similar to Guards). However, if a Stalwart is close enough, they will turn to face the Stalwart instead.

This isn't the problem. The problem is that they continue to obey the brain, and move towards you, whilst still facing the closer Stalwart.

Now, this is a behaviour that isn't unique to Aumtlich: other enemies will treat nearby Stalwarts as targets but still move towards you (Goblins and Wraithwings as quick examples). But the turning behaviour from Aumtlich makes a much more marked and visible difference, so I felt I should bring it up to confirm things before going ahead and making use of it.

mrimer
Level: Legendary Smitemaster

Rank Points: 5106
Registered: 02-04-2003
IP: Logged

That sounds funky, for sure. But, eh, I'm personally worn out at the moment from all the game changes that we've made over the last year, and it sounds like other people are too. So let's just leave this as-is unless there's a substantial, unanimous voice to change it right away.

____________________________
Gandalf? Yes... That's what they used to call me.
Gandalf the Grey. That was my name.
I am Gandalf the White.
And I come back to you now at the turn of the tide.

---

[Last edited by mrimer at 05-26-2008 09:27 PM]

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

Note to self: check how damaging fixing this would be.

____________________________
My website

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

There is a room in Advanced Concepts based on this. Otherwise I don't think it comes up much?

Changing this seems questionable, as the basic reason it seems weird is that the brain pathmap overrides moving towards a stalwart for most monsters (though this is not the case for guards). I guess that could be a bug too, but it would have a lot more impact to change it.

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

No, the reason is weird because Aumtlich chases player (which is expected for a brained monster) but faces Stalwart. At least that's what weird to me . It should face its target in my opinion

____________________________
My website

Nuntar
Level: Smitemaster

Rank Points: 4711
Registered: 02-20-2007
IP: Logged

I built a room using this for Variations, currently in indefinite hiatus, but I would be the first to admit it's not a great room and I don't mind if it breaks. I agree with skell that it would be better for the aumtlich to face its target -- which is always Beethro if the aumtlich is brained and has a path to him, because that's how brains work.

____________________________
50th Skywatcher

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

The brain is just telling it how to move but not how to face.

Is it weird? Yeah. Broken? Not obviously. The aumtlich just isn't smart enough to think that it might want to face towards the guy the brain is pointing it at, rather than the closer guy with a sword who it wants to kill. And the brain doesn't have the ability to tell it to do that. I don't see particular reason why either of those things should be true, as opposed to the way things are currently.

Lucky Luc
Level: Smitemaster
Rank Points: 1215
Registered: 08-19-2012
IP: Logged

The change would be to face towards the player, then? I used brain + aumtlich + stalwart in Take Five: A Five Brains Hold 1N, but I think it probably wouldn't break the room, though some demos might break. In total, room search comes up with only 24 rooms that contain all elements, so it probably wouldn't be a huge deal.

I'm also not sure this is clearly broken. Lore-wise, this could be a self-defense thing, where even when they are on their way to somewhere else, they still try to stop the thing that's trying to kill them in their tracks.

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

I think I am looking at this from a fundamentally different perspective, because imagining I am a new player I can't see how the current behavior makes the slightest slither of sense or would be in any way possible to figure out beforehand.

____________________________
My website

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

The part that doesn't make sense is that the aumtlich doesn't move towards the stalwart.

Brained monsters move towards decoys (other than goblins - another anomaly to potentially discuss if we're rethinking things) and guards, who are always brained, do target stalwarts even when they have a path to the player.

But other brained monsters ignore stalwarts. (They may incidentally kill them, but they don't move towards them unless they can't make a move the brain wants them to make.) This is the weird part, and it's a lot more consequential to change.

If I went in with only basic information about what stalwarts, brains, and aumtlich do, and even add in the specialized knowledge that the brain only makes a path to the player and not stalwarts - my assumption would be that the aumtlich would move towards the stalwart.

However, given that it doesn't do that, I don't find it any weirder that it still faces towards the stalwart. From my perspective, the weird part has already happened. Once that weird aspect is established, I don't have particular expectations for whether it's going to face Beethro or the stalwart - just that it's going to face one of them as opposed to, say, deciding to face northwest regardless of where either of them are.

navithmastero
Level: Smitemaster

Rank Points: 614
Registered: 01-03-2012
IP: Logged

I don't really have a comment about this particular bug but in general I'm not necessarily sure that pretending like we're a new player is the way to go when looking at some of these borderline bugs. The new player base for DROD is relatively small. I would argue it makes far more sense to keep things consistent for current players for whom behaviour has been in place for 5, 10, sometimes even 15 years than to change things, possibly breaking demos, or even worse rooms, for the sake of the 1 or 2 new players who may find an interaction slightly odd (but not obviously buggy). Personally, I think the focus should be placed on fixing obvious bugs, rather than just interactions which feel a little odd, as there are plenty of interactions which feel odd and which certainly aren't bugs.

____________________________
Member of the Snake Appreciation Society

One of your local HAs.

My stuff:

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

I think this and a few other "weird interaction" threads along the same lines are rubbing me the wrong way because they seem to be starting from a presumption that the behavior should be fixed, without an argument that it's actually broken (as opposed to just unintuitive).

The general expectation is that behavior will remain consistent, and people make their rooms understanding this. It's unfair to make things work differently so their rooms no longer work.

Sometimes that unfairness is a necessary price - for instance, I agree it's not reasonable to keep the weird behavior with decoys in order to preserve that one room in Zealot's Trials. But I tend to think it needs a stronger case than "it's unintuitive" to warrant paying that price, and this doesn't meet it for me.

navithmastero
Level: Smitemaster

Rank Points: 614
Registered: 01-03-2012
IP: Logged

Dragon Fogel wrote:
I think this and a few other "weird interaction" threads along the same lines are rubbing me the wrong way because they seem to be starting from a presumption that the behavior should be fixed, without an argument that it's actually broken (as opposed to just unintuitive).

The general expectation is that behavior will remain consistent, and people make their rooms understanding this. It's unfair to make things work differently so their rooms no longer work.

Sometimes that unfairness is a necessary price - for instance, I agree it's not reasonable to keep the weird behavior with decoys in order to preserve that one room in Zealot's Trials. But I tend to think it needs a stronger case than "it's unintuitive" to warrant paying that price, and this doesn't meet it for me.

+1. I too am starting to get concerned with how bugs are being handled. I will reiterate the main point from DF's post: unintuitive =/= broken. The fact that we are considering changing behaviour which has the effect of fundamentally changing the way elements interact and has the potential to break hundreds, if not thousands of demos (and what is more concerning, is the potential to break hundreds of rooms) is... pretty scary. Many people do not find serpents to be intuitive. That does not mean we should try and 'fix' snake behaviour. While it is great that DROD is being actively worked on, things shouldn't be 'fixed' just for the sake of it. I would urge those who are working on DROD to consider why certain things should be worked on. If the answer is 'because it feels a bit weird to me' it almost certainly shouldn't be changed.

That being said, I really want to stress that despite this I am still very grateful for those who are spending their valuable time working on DROD. My comments are made out of a love for this game and this community!

____________________________
Member of the Snake Appreciation Society

One of your local HAs.

My stuff:

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

Let me address a couple of comments and share my thoughts. I apologize ahead if I sound annoyed, because I kind of am but I need to get this off my head so I can resume work .

1. What is bug and what is odd is a matter of perspective - for me the behavior posted in this thread is a very obvious bug. The token thread? Obvious bug. Snakes + invisibility + brains? Very much a bug. Your mileage may vary, trying to steer the discussion away from the individual merit of a specific report towards "this is not a bug, just a quirk" is fruitless and in my mind is more of a deliberate tactic to derail an objective conversation towards emotional exchange.

2. I don't like the idea of a game being developed by veterans for veterans. That tends to create hostile environment for the newcomers and a welcome community can only do so much against problems in the game itself. I don't want to make a game for veteran DROD players because they are an endangered species that has sub 2.1 fertility. I literally do not want to make such game, if DROD community ever becomes an elitist circlejerk I am out.

3. DROD has thousands of little interactions, rules and exceptions that create an immensely complex clockwork machine. It's really, really, really easy to accidentally create things that don't make that much sense if you're not paying the strictest attention. See for yourself. And here. And that doesn't include any development from TSS initial release up to 5.1. There is only so much you can keep in your head at once. For the rest you must rely on intuition, discoverability, luck, mnemonic associations and triggers. If you ever see me oppose a rule remember that I am most likely much worse player than you are, despite statistically most likely playing DROD for longer than you. If I say it's not intuitive to a new or beginner player I am not only talking about some hypothetical newbie - I actually am talking about my own personal experience with the game. Which holds its own set of biases, I admit.

4. If you frequent chat your DROD skills are likely way above the average and that means you're also likely to be out of touch with reality of those for whom majority of the released content is challenging. I don't want to send them the message "screw you, lest the veterans get confused by some obscure edge case used in 50 rooms out of 40 thousand". I don't want to send the veterans the message "screw you, I don't care about your feelings" because I do, but see #2 and #3.

5. Continuing that thought, I would think that the veterans would be more understanding towards normalizing atypical rules for the sake of simplicity for themselves and because they are easier to adapt. Moreover, I don't want to touch core rules that are part of majority rooms, only things that you encounter at most dozens of times during a full playthrough of all the holds, so it's not a rule that, if changed, would have a major effect on a veteran. But for a less-skilled player may be the difference between being able to reason about it and being forced to post on H&S thread.

6. Probably won't be a surprise to many but I honestly do not care about breaking demos. Of all the consequences of changing behavior, this is of the least importance to me. I do care about your feelings about your demos being lost, but I am unable to empathize with this on a more intimate level. I also don't understand the argument, that breaking demos sucks because you need to replay the same room again - the reason is that even if you have to replay the same room, due to rules being changed you don't replay the same puzzle. I do not intend this as a personal attack on your preferences - they're just alien to me. My goal when minimizing the amount of broken demos is to save others feeling sad, reduce the amount of negative feelings sent towards me and to have less work with the release of a new version.

7. Similarly, I don't care very much about not breaking rooms. I disagree with the notion that holds should be held in amber, protected from any change. I'd prefer NOT TO break rooms (because it makes people sad and is more work for me), but there are certain classes of breakage that in my opinion can be trivially amended: introducing unintended solutions and changes that can be fixed by changes in architecture or scripting. At the same time I do recognize that it sucks if something that existed before is no longer available. I do miss that Fool's Errand trick is no more because I liked it. But the art changes. You likely do not play the holds in exactly the same context as the author wanted, but that doesn't really diminish your enjoyment? It's a very broad and complex topic about what is art and how art is experienced. I think that if the architect is there to fix the room, or the room can be trivially fixed by developers by editing the hold, then it's not a change that it's worth worrying about.

I am going off on a tangent now. There might be differences on where you and me agree on what's important or not, but remember that any change meant to make rules more coherent benefits not only the 1-2 new players that will get confused (which I disagree with, new players will already be confused by a lot of things that are intended), but also the 50% of average and worse players.

I must call something out specifically:

I'm not necessarily sure that pretending like we're a new player is the way to go when looking at some of these borderline bugs [...] for the sake of the 1 or 2 new players who may find an interaction slightly odd (but not obviously buggy)

This is an irreconcilable difference between our approaches. This is why I wrote all of the above, because I don't think I can find any argument against what you said, as anything can be defended with "that's slightly odd, not a bug" or "there won't actually be (m)any players who will be confused about this". I kind of felt that way in the previous conversations trying to defend my position which is why I didn't even attempt to defend it here. I may be out of line and completely mistaken, but I want to think I am trying to hold a bigger picture in my head - finding solutions that will benefit the game in the long run, rather than just forcing my personal opinion. Of course I am still just a human so that might look completely differently from the outside.

----

And since this was written before Fogel's post about presumptions, let me address that:

1. Broken or not is not a metric I think has merit (as I talked about in #1)
2. These threads were posted by players. Here Mike himself stated that's a behavior he'd change if not for being tired. In the tokens thread the community consensus steered towards the existing behavior being wrong. Pushing spearman into bump had no real conversation. Brains & serpents had at least a bunch of people say it's a bug.
3. That's why there wasn't much argument, because in all cases: there was official wording, or the consensus steered towards it being a bug, or there was no argument from the other side.
4. I've worked with the DROD code a lot in the past, during TSS development and 5.1 development and now 5.1.1 development. If my familiarity with the codebase and its history gave me a reason to think a given behavior is intended I'd mention it. If, during the implementation, I found any indication of that, I'd mention it, but of the four threads I mentioned I implemented the change in only one. I don't want to claim to be an authority, but also given my extensive contributions to the engine I'd appreciate at least assuming that I am acting in good faith.

----

Welp, that's a wall of text. The reason I think the way it works right now is bollocks is because my understanding of the rules is that if brain affects a monster it takes control of its pathing, making it ignore anything else except decoys in range. There is no other monster or situation that is comparable to Aumtlich facing Stalwarts while targetting player. There is no other situation where aumtlich moves in a direction they are not currently facing. There is no other situation where a monster faces something else than their target (with the exception of slayers which are governed by completely different set of rules, and serpents which also have their own separate ruleset).
I also think the puzzle potential of keeping this as-is is a great fuel for awful puzzles (because Stalwart pathmapping), and somewhat decent fuel for accidentally breaking puzzles (because it's obscure and easy to not know).
I also think it'd be unintuitive and difficult to understand for a player who encounters it, because of what I wrote above and because when you actually engage with this in play it looks as if Aumtlich is moving randomly if the distances are similar.

____________________________
My website

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

And to add:

I too am starting to get concerned with how bugs are being handled.

I feel hurt by this. I disagree that I gave the community a reason to think that stuff is handled wrong. Not once did I try to override anyone saying "that's how it'll be" and try to end the conversation. I have my opinions but every time I listen to the voice of the community. The only time things got even close to being handled "improperly" was the pushing doubles + token, where the code change was not merged into the main codebease, ergo NOT making it into the game. I even was convinced by the voice of community to change my opinion.

Perhaps it's the unfortunate wording of using the word "fixed" when bumping threads that I wanted to look at? I apologize for that, but in the last few days I read a few hundred bug threads and that was what I thinking - it's a bug and we use the "fix" verbiage for them.

I don'think there is any reason to get concerned by anything.

____________________________
My website

---

[Last edited by skell at 10-13-2020 09:19 PM]

hyperme
Level: Smitemaster

Rank Points: 1143
Registered: 06-23-2006
IP: Logged

I think the underlying issue here is that for a long time, protecting demos and rooms has been placed above fixing things that are objectively wrong with the game. It's the Reserved List of DROD, complete with the unnecessary propagation of harmful design patterns in order to hold up something that seemed like a good idea at the time. Worse bugs than this have not been fixed, deliberately, and that creates an obvious conflict when it comes to lesser bugs that might break things.

The thing is, of course, is that in this case skell is right. Brained Aumtlich getting distracted by Stalwarts is completely out of step with how Brained monsters generally work. It's a weird exception, and it was obviously meant to be fixed, but then fell out of memory.

Sure, you can be concerned that skell being right now could lead to skell being right again in the future. But that's the risk of actually making the game work correctly - sometimes stuff gets broken. Some of it should be broken. But that's not skell's (or my) goal. What we should all want is a game that makes sense, and isn't filled with inane edge cases, because then it stops being about interesting puzzles and starts being about memorizing the entire game ruleset. There's no shadowy conspiracy out to destroy your rooms and demos. If anything, the conspiracy is in favour of all of the terrible rooms that should be destroyed because they accidentally rely on the pulsating tumour that is the Smart Pathing algorithm and it's straight-up, factually incorrect implementation.

____________________________
[Insert witty comment here]
Qzvlkx?

navithmastero
Level: Smitemaster

Rank Points: 614
Registered: 01-03-2012
IP: Logged

Edit: I did not see hyperme's post before I posted this.

skell wrote:
And to add:

I too am starting to get concerned with how bugs are being handled.

I feel hurt by this. I disagree that I gave the community a reason to think that stuff is handled wrong. Not once did I try to override anyone saying "that's how it'll be" and try to end the conversation. I have my opinions but every time I listen to the voice of the community. The only time things got even close to being handled "improperly" was the pushing doubles + token, where the code change was not merged into the main codebease, ergo NOT making it into the game. I even was convinced by the voice of community to change my opinion.

Perhaps it's the unfortunate wording of using the word "fixed" when bumping threads that I wanted to look at? I apologize for that, but in the last few days I read a few hundred bug threads and that was what I thinking - it's a bug and we use the "fix" verbiage for them.

I don'think there is any reason to get concerned by anything.

I'm sorry for this Skell, my comment was not an attack against you or your contributions to the community. For me the use of the word 'fix' is one of the main things which made me feel like this wasn't being handled right. For me, 'fix' implies that you've already decided something is a bug and so will change it. Apologies if this is not what you mean or is more short-hand for 'this is something I will check out'. I am not a developer and am not familiar with how you do things, I'm sure this is more of a misunderstanding than anything else!

As for your comments in the previous posts, I will try and address them below:

1. I agree that this thread is maybe a bug. The token thread though? The only thing that makes it seem like it could be a bug to me is that Beethro activates tokens when pushed onto them. That seems more likely to be a bug than the other way round and if we are going to change some interaction, I think it is obvious that this is a much more favourable route (I am happy to ellaborate on this on the other thread if needs be though). I fundamentally disagree with your point on 'not a bug, just a quirk' though. When it comes to DROD behaviour, I think that emotion does play a part in deciding what we should change and what we should not change. As an example, everyone hates the way that citizens work because the way they move is unpredictable. In my opinion, if it weren't for the need to 'grandfather', we should have changed them a long time ago, because despite them actually not being broken (as far as I am aware) Edit: just seen hyperme's post above, if Citizen's are indeed broken than a more apt comparison it seems would be stalwarts or gentryii chain straightening or halph (there are a number of examples you can use in stead), they feel broken and/or entirely unintuitive. To me, some of these interactions (like the token one you mention) does not feel broken, at least to me, and I think this is important. DROD is a game, and ultimately, games are subjective, and thus there should be an element of subjectivity. This is the way I view things, although I fully understand why you disagree.

2. While I agree with the sentiment I think one of the things that is very important to note is that veterans are exactly the people who keep DROD accessible for new players. Veterans are the people who reply to the H&S forums and veterans are, more often than not, the people who build the holds and run the contests for people to play. You yourself are an example of this - you're a veteran putting in his time for the game that you love. I build holds and frequent chat and make videos not just because I love this game and I love this community but to give something back. I am engaging with the bug fixes you are doing because I feel passionate about this game and just like you want DROD to be the best it can for new players. I think that by potentially pissing off veterans you are also potentially worsening the experience for new players. I'm not saying that this is absolutely the case, but just something to keep in mind.

3. Forgive me if I'm misunderstanding this point but isn't this exactly the type of thing that you said you were opposed to in point 1? Unintuitive is a subjective quality.

4. Likewise, see my above points.

5. This point feels a bit circular to me. If veterans only see it a dozen times in their play through of all the holds (tm), then presumably newbies aren't gonna see it at all, or if they do, they are statistically very likely to see those interactions by the time they would then consider themselves to be a veteran, by which point, it's moot.

6. I don't really care too much about breaking demos because it's a lot easier to replay a room than to rebuild a room, but again, breaking demos can have the effect of pissing people off. See #2

7. Honestly, this kinda does feel a bit like a 'screw you veterans'. As someone who has released a bunch of holds with plenty of USs, I get really frustrated when I see a room break because of something I haven't considered. Now think about a room breaking because a behaviour which had one post about it in 'Bugs' at the time of the hold's release but was widely considered by other members of the DROD community as fair game now means that the room is easily breakable (or even worse, rendered impossible), this would make me annoyed! The art does change, but the rules should (in most cases) stay the same. I agree that there is some level of flexibility around this as there are some things which have been exploited by architects which are clearly buggy but in a lot of cases, to me, they seem like edge cases and saying that you don't care about how this will affect other people's projects which they have put out for other people to enjoy is also a little hurtful. Not to mention that if a newbie goes through and plays a hold who's rooms have broken or which are unfinishable because of rule changes also makes the experience less enjoyable for them. Also see #2.

I hope this explains my feelings more on the subject of bugs but please feel free to ask me to ellaborate on any of my points above!

____________________________
Member of the Snake Appreciation Society

One of your local HAs.

My stuff:

---

[Last edited by navithmastero at 10-14-2020 11:22 AM]

Lucky Luc
Level: Smitemaster
Rank Points: 1215
Registered: 08-19-2012
IP: Logged

hyperme wrote:
Brained Aumtlich getting distracted by Stalwarts is completely out of step with how Brained monsters generally work.

I feel the need to point out that this is not an objective truth. Aside from guards, Aumtlich are the only monsters in the game that can even do turning moves. In fact, it's perfectly in line with how goblins still actively avoid tiles close to stalwart swords when brain-mapping towards the player.

EDIT: Maybe as a side note, about this whole meta-discussion that was started now. One thing I want to mention about the new-player-perspective is that I personally love discovering interesting interactions in games. Yes, they shouldn't be counter-intuitive, but I think just because I couldn't predict it beforehand it doesn't necessarily mean a behavior is unintuitive. And yeah, it's easy for me to say that now, but I think if I was shown a room with those three elements, one of my first reactions would be "hm, I have no idea how these interact. Let's find out!" In fact, I'm pretty sure that's how I came up with the mentioned room in the first place.

EDIT2: Just another quick note: Skell, I absolutely admire the work you've put into DROD not just the last couple of weeks and I'm pretty sure everyone who has been around for a bit does. If you're getting annoyed because you put in a lot of effort and all feedback you get is disagreement over certain proposed changes, I would understand that (I certainly would). So just to make that clear: I think you're doing an amazing job and I'm very grateful for it. It would be a shame if you took this thread as "the community doesn't appreciate me", because that's most certainly not what it is.

---

[Last edited by Lucky Luc at 10-13-2020 10:23 PM]

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

I repeat: the weird part of the behavior, to me, is that the aumtlich moves towards the player when a stalwart is closer.

This is how most brained monsters behave with stalwarts (and multiple targets in general).

However:
-When a brained monster (other than a goblin) is near a decoy, it is drawn to the decoy instead.
-Guards follow the brain pathmap, but prioritize nearby stalwarts.

So this is inconsistent and confusing. It will be confusing either way. Even if we make the aumtlich turn towards the player character all the time, the player will still be asking "why doesn't the aumtlich go after the stalwart".

And if we keep things as they are - the player is still going to be asking that. The fact that the aumtlich is trying to face towards the stalwart is also going to be a question, but at least the player who observes it can say "I guess that's happening because the stalwart's closer".

Neither outcome is predictable to a completely unfamiliar player because the fact that the aumtlich will not move towards the stalwart isn't predictable to them in the first place. If it happens to them, they will then try to understand why the aumtlich is facing the way it is, and both options can be understood in terms of what logic they're running on, even if the broader question of "why are they using that logic" isn't clear.

If the game were in development and this were something that hadn't been implemented yet, I probably wouldn't have a strong preference. I might try to make an argument for "face the closer target" on the basis of "I think it's cleanest overall if brains only affect movement, not turning or anything else" but I don't think I'd be that invested in it.

But with this being how things currently work, and rooms have been made based on this being how things currently work, I think it's best to leave it the way it is. The confusing part of it is not the turning, it's the moving, and that would be even more drastic to change.

Rabscuttle
Level: Smitemaster

Rank Points: 2485
Registered: 09-10-2004
IP: Logged

This behaviour also applies to timeclones - an aumtlich pathed past a timeclone will stare at it as he passes

My take is that the current behaviour makes sense. The brain is only telling monsters where to go, not who they should be targeting. The brain may seem to be leading them away from who they think the target is, but brains often do that in rooms with windy passages.

So a goblin will follow the brain to Beethro unless there's a weapon (or lack of weapon) close by.
A wraithwing will use the closest enemy to decide whether to be brave and go with or be scared and go against the brain.
Aumtlich behaviour is consistent with these.

---

[Last edited by Rabscuttle at 10-13-2020 10:20 PM]

Nuntar
Level: Smitemaster

Rank Points: 4711
Registered: 02-20-2007
IP: Logged

I want to write posts, both here and on the doubles/token thread, with my thoughts on both issues. Since things have gotten a bit heated, I just want to say that I am really grateful for the hard work that's going into improving DROD at the moment, and I just want to give my thoughts in case that can help us steer towards the best decisions.

skell, one thing you said is incorrect, and I'm just pointing this out so that everyone has easy access to the relevant information:

skell wrote: There is no other monster or situation that is comparable to Aumtlich facing Stalwarts while targetting player.

There are two other comparable situations, as TFMurphy stated in the opening post. Brained goblins will target the player, but will avoid a nearby Stalwart's sword, even though ordinarily a goblin will only avoid the sword of its current target. Brained wraithwings will target the player, but will stay 5 tiles away from a nearby Stalwart.

The coincidence of all three examples does make me suspect that this was deliberately coded -- but that's just speculation, and isn't really relevant to the current question: is it best to change the behaviour or not?

I'll start with some general thoughts.

DROD is already a game for veterans. Sure, there are entry-level holds, and the better ones are very enjoyable (otherwise few of us would have stuck around to experience the veteran content!); but most of the interest and pleasure comes from the way the space of possibilities within DROD rooms continues to provide new challenges long after we've reached the point where something like the three-tar-mother room is a walk in the park.

Brained behaviour (in general) is an especially good example, because it's extremely complex -- the Advanced Brain Movement level in Advanced Concepts is by far the largest level in the hold, as well as the most exposition-heavy, and it was made in the TCB era, so anyone working with brains now has even more interactions to think about. However, the level of complexity is manageable -- a single room won't include every interaction, and the ones that come up more frequently will become more familiar to players. Recently I watched navithmastero's LP of the brains level in "She'll Carry a Really Big Sword", which is a great example of a level that would be totally opaque to new players, but every room seems fair and is a really beautiful puzzle with an elegant solution.

However, there are definitely quirks in the behaviour that don't make sense. The brain tells enemies where Beethro is, so they target only him: that's fine. But then why are brained monsters fooled by decoys at all? Why is there an exception for brained goblins? (Most likely, because these quirks were added when there were far fewer elements in the game, so the amount of stuff players had to learn was more manageable than it is now, so it was felt to be okay to throw in weird gimmicky exceptions.)

So, should we smooth the learning curve by getting rid of some of these quirks? This is where the argument from existing content comes in. I don't know why you would say:

I disagree with the notion that holds should be held in amber, protected from any change. I'd prefer NOT TO break rooms (because it makes people sad and is more work for me), but there are certain classes of breakage that in my opinion can be trivially amended: introducing unintended solutions and changes that can be fixed by changes in architecture or scripting.

...because we can't change the architecture or scripting of existing holds. It used not to be allowed ever, and it's true that that has been relaxed a bit lately, but it's still something that is almost never done. Historically, when rooms were broken because of behaviour changes, those rooms had to be made unscorable or replaced with blank rooms.

(Also, unintended solutions are usually not trivial to fix. There are some quick-fix cases, but in complex rooms, finding ways to fix unintended solutions without breaking things usually ends up being far more work than the original building.)

Still, I think it's important to consider the balance for each case separately. In the case of the double/token interaction, I believe strongly (and I'll post on that thread so as not to let it take over this one) that the scale of the damage to existing content is too great. In the case of the serpent/invisibility bug, it seems likely that the damage would be very small, if any, so I would personally prefer it to be fixed. This one... I don't know. It's complicated by the fact that it involves several quirks and not just one. Also, Dragon Fogel and others know the existing content a lot better than I do.

I can't think of a good way to wrap this post up so I'll bow out for now.

____________________________
50th Skywatcher

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

File: Monsters Examples.png (363.8 KB)
Downloaded 283 times.
License: Public Domain

@navithmastero - I appreciate your reply to my big post but I didn't post it to discuss it - I wanted to share a piece of my brain because I felt the whole discussion around the topics was getting unnecessarily heated and I felt there is an invisible wall I did not know how to get through. I've read your comments but I won't continue that conversation, I am sorry. All I can really add there is my perspective is likely to be different because I treat the game as a product which I develop, and my skill in DROD is average.

There are two other comparable situations, as TFMurphy stated in the opening post. Brained goblins will target the player, but will avoid a nearby Stalwart's sword, even though ordinarily a goblin will only avoid the sword of its current target. Brained wraithwings will target the player, but will stay 5 tiles away from a nearby Stalwart.

Thank you! Those are not interactions I was aware of so let me dissect them:



- A) Goblin will step SW, because it targets the Eastern Stalwarts, thus sees only its sword, and wants to go towards the player.
- B) Wraithwing will step N all the way until it gets to the distance of 5 from Stalwart, because it targets Stalwart and uses it as a calculation for distance from its target, but then it moves towards the player. Towards, but because distance to its target is <5 it actually runs away from the player, but not from the stalwart.
- C) Wraithwing will never move until you get close to it, because it Targets Stalwart for the calculation of distance, so it stays on the safe distance forever.
- D) Wraithwing will run into the Stalwart's sword, because it targets Stalwart and uses him to calculate distance, but then it makes the move towards the player. Towards, but because distance to its target is <5 it actually runs away from the player. It's the same as B) just even more bizarre.

The reason why this happens is because Aumtlich, Goblin and Wraithwing have separate steps for finding their target and moving towards it. Why? Because they need to process the target somehow before making their move. This function does not take being brained into consideration, which is why they target Stalwart that is closer even if the brain detects the player.
But when it comes time to move, there is this code which overwrites where the monster is supposed to go. Take a note that if the target was a decoy, it would ignore Beethro completely.

What does it tell us? A couple of things:
1. You could say it's perfectly consistent - a brained monster behaves like they would but their feet are controlled by brain and all the monster can do is either decide to run away or not run away. It has its issues though - if that was the case brained goblins should not avoid swords, because that's more complicated than "move forward" or "move backwards".
2. We could expand #1 to say that monsters also have some ability to direct the brain's movement, but then why wraithwings are so stupid?
3. There might be some more complicated answer that I can't come up with, but I think even #1 is already a stretch.
4. Alternatively you could argue that the code for GetTarget() suggests that the current behavior is expected. I think it's just an oversight. Below is how the code looked before Stalwarts were introduced, make of that what you wish.



For me all signs point towards "this is not intentional" and "this is confusing and hard to understand". And while I agree with Nuntar's statement that intention is not all, I think the behavior of Aumtlich is confusing, difficulty to understand and not very interesting (ditto for Wraithwing). My opinion is - change it while we still can, but I am not the final arbiter - Mike is.

And you're not wrong Nuntar with not being able to change published holds - that's something I disagreed with since I can remember and disagree still. There is no external reason to not be able to update old holds to work with new changes and I think it's the high time to relax those restrictions and allow improving/fixing old works.

There are much more that could be squeezed into this post but I'll stop here, I can't really add much more about the analysis of this specific case without having to invoke the same arguments over and over again. Just two final thoughts
- "Still, I think it's important to consider the balance for each case separately." - I agree wholeheartedly
- Don't fret about some evil party coming and changing rules while ignoring protests. It didn't happen, it's not happening, it won't happen.

____________________________
My website

---

[Last edited by skell at 10-14-2020 11:27 AM]

Dragon Fogel
Level: Smitemaster
Rank Points: 2453
Registered: 06-21-2014
IP: Logged

Again, I stress that the part that's confusing and hard to understand is that they don't move towards the stalwart. In fact, if you look at the initial post, that's the behavior TFMurphy is confused about - he has no problem with the aumtlich facing the stalwart.

Also, brained goblins are messy and confusing. They ignore decoys, they'll respect a "wait" order (which otherwise only monsters with Direct movement will do), and they completely ignore the brain's orders when they're two or three squares away from you (I forget which). So I really don't feel they're illustrative of anything general about brains.

Basically, my expectations were broken when the aumtlich continued to follow the brain path instead of prioritizing the closer target. Which way it faces is then more of "so what's it going to do" than anything I have expectations for at that point. Given that, I support not breaking rooms where this is relevant.

Nuntar
Level: Smitemaster

Rank Points: 4711
Registered: 02-20-2007
IP: Logged

Dragon Fogel wrote:
Again, I stress that the part that's confusing and hard to understand is that they don't move towards the stalwart. In fact, if you look at the initial post, that's the behavior TFMurphy is confused about - he has no problem with the aumtlich facing the stalwart.

That really isn't how I read the post. TFMurphy says: "The problem is that they continue to obey the brain, and move towards you, whilst still facing the closer Stalwart." -- in other words, it's the combination of those two contradictory things that's a problem.

Brained aumtlich move prioritise the player over stalwarts and timeclones because all brained monsters do, because monsters trust the brain over their own instincts in that situation. It might be unexpected behaviour when you haven't encountered it before, but it's one single consistent rule to learn. Since it applies to all brained monsters, it affects too many rooms to consider changing it.

Also, if we change the facing behaviour so the aumtlich doesn't face the stalwart/timeclone, that would make it easier for new players to understand that the aumtlich is targeting the player and learn the behaviour.

____________________________
50th Skywatcher

bbb
Level: Master Delver
Rank Points: 215
Registered: 10-07-2013
IP: Logged

I think I'll join Dragon Fogel in saying the rules shouldn't be changed just because they are unintuitive. I don't think no demo breaking changes should be made, but we should think very carefully before making changes which break (either by making trivial or unsolvable) old rooms, which were made over more than 15 years.

Even if each such change only breaks a few rooms, it is likely that cumulatively many excellent holds will no longer be playable. I think that the minimum which has to be done before such a change is done, is to check how many demos for which rooms might break.

The criteria for rooms to be checked could be any room where at least 20 or 25% of the demos break. (The exact number and ratio doesn't matter.) The result of such a check can be one of the following:

1. Some demos fail because of a technicality, such as requiring an extra wait move, or skipping unnecessary actions. This result shouldn't block a change, and shouldn't require much discussion.

2. Demos break which can't be fixed trivially, but the room can still be solved with the intended solution and doesn't obviously break. Such a change should require discussion, but should be permitted.

3. Rooms break such that they become trivial. Should not be done without serious discussion, but shouldn't be completely prohibited. See point 4 for additional considerations.

4. Rooms break and become unsolvable or unreasonably hard. Should not be done without serious discussion. A decision can be made to accept that change without further action only if the hold is considered to be a poor hold which no one will miss, e.g., Bad Evil Restaurant. Even in this case, the author should be notified if possible to fix the room. If the broken room is in a "serious" hold, official or not, the room has to be fixed before accepting the change. The fix should preferably be one which does not change the room behavior, e.g., fixing scripts, or simulating the old behavior using scripting. The next option is to modify the room so it works with the new rules. Changes should be slight, such as modifying timers and such. Another option, which would open a new can of worms, is to continue supporting the old behavior with the engine version tag in the hold. If the original room authors are still around, they should be consulted, but changes can be made even without them. However, the fact that an interaction appears in several (good) rooms is often a fair indication that the interaction should not be changed. Even if it should have been implemented differently in the first place.

Rabscuttle
Level: Smitemaster

Rank Points: 2485
Registered: 09-10-2004
IP: Logged

A couple more datapoints:

If a brained goblin paths next to an unarmed stalwart, then he'll eat him. Although if he's one more square away, he'll ignore them. i.e. in case A) above if you put the stalwarts on oremites (and force arrows), the goblin will walk down the middle. But if you move the goblin left or right one, he'll attack the stalwart.

-

If a roach-like is unable to make a brained move because the squares are blocked, it will make an unbrained move towards the closest enemy, rather than the brain's target.

-

The latter to me is consistent with the idea that the brain only says "go this way" rather than "go this way to get this specific enemy"

skell
Level: Legendary Smitemaster

Rank Points: 3757
Registered: 12-28-2004
IP: Logged

My opinion on the current behavior remains unchanged, but the impact of this is low, majority of community disagrees, I don't want to deal with broken demos or, potentially, rooms, so I'd say we're keeping things as they are right now.

Even the absolutely atrocious wraithwing behavior .

____________________________
My website