Room Variation Problem

[Pepisolo]

This is a bit of a copy and paste from Mantis, but it's something that needs discussion.

http://arcengames.com/mantisbt/view.php?id=18063

"One of the big things concerning me about this game at present is the lack of variety in the maps. Specifically, there is a distinct lack of a SHMUP flavour in the rooms, and I'm speaking mechanically rather than aesthetically. AVWW2 seemed to suffer from a similar problem in that although there were numerous upon numerous chunks in the game, they were largely pretty samey, weren't particularly platformy, and were rather bland. The rooms in this game must not suffer from this same blandness! The main problem seems to be that there aren't enough cool options available in the room editor, especially SHMUP-style options. So, this report is going to suggest various ideas for additions to the room-editor, inspired by SHMUPs of old, that if implemented might add some more variety and essence of SHMUP to the game. I will keep adding to this report over time.


1) Rotational Laser

Very common and presumably fairly simple to implement. This is simply a laser or a few lasers, usually attached to a circle that rotates either clockwise or anti-clockwise.

Example: https://youtu.be/gXqKsGxJL7M?t=53s [^]


2) Spawners

I think something like this is already planned. Spawners, we all know what they do, generate enemies or obstacles until they are destroyed.

Example:https://www.youtube.com/watch?v=2ct8sssUI6k&feature=youtu.be&t=4m30s [^]


3) Wall funnelers

These are enemies or destroyable obstacles that rather than simply being present in the room, they instead funnel in from openings.

Example: https://youtu.be/Ens0SDe4yNQ?t=1m32s [^]

First it's the giant snake thing on the bottom-left. Then later it's some ships from the top-right. For Starward Rogue, the concept would need to be slightly modified to prevent the player from trying to go down the opening, so instead you could have a hatch that opens and then closes immediately after the enemy or series of enemies are released.


4) Asteroid-like obstacles

There are usually lots of obstacles or enemies like this in SHMUPs. Basically, it's a big floating object that bounces around the screen. You can either choose to shoot it or dodge, but if you shoot it, it breaks up into smaller pieces. Example:

https://youtu.be/hpDOgXCYRyo?t=21s [^]

If anybody has any more suggestions please comment!"

To sum up. Despite there being numerous rooms in the game there is very little potential variation in the designs -- there just aren't enough cool map objects to use. As a result the rooms are way more boring than they should be. Any ideas on ways in which more variation could be introduced would be greatly appreciated, and is actually badly needed as far as I'm concerned.


Edit: As Misery mentions below here is also an interesting Mantis report about the idea of having some puzzle or hazard rooms.

http://arcengames.com/mantisbt/view.php?id=18082

Later edit: More ideas will be added to the first Mantis report linked at the start rather than in this thread.

[Misery]

Just to add something here, ideas for entirely new types of "normal" rooms might be useful too.  Right now, all rooms are combat rooms.  Having alot of those is fine and kind of expected, but it seems to me like there needs to be other things that can pop up too.  Bluddy had the idea of hazard or puzzle rooms that could appear, those are good possibilities.

So any ideas for full concepts like that might help as well.

One upcoming block type also is "phasing" blocks, if those havent been mentioned.  You know... like the Megaman 2 ones.  Those things.  Popping in and out.  So that's at least one object that's incoming.

[Pepisolo]

One upcoming block type also is "phasing" blocks, if those havent been mentioned.  You know... like the Megaman 2 ones.  Those things.  Popping in and out.  So that's at least one object that's incoming.

The more the better (within reason of course). I think that the level designers really need more options at their disposal. When I briefly messed around with the room editor I felt super restricted, there just wasn't enough cool stuff for me to use.

[x4000]

Whew, okay, talk about a tight spot.  1.0 release is planned for the 22nd of January, and would be pretty hard to move at this point for... various reasons.  Definitely keep the ideas flowing, and we'll see what we can do.  Hopefully Misery, Zack, and Dayton can kind of coordinate on this and we'll go from there.  I can get things in the editor after new entities (spawners, etc) are implemented in xml, and Keith can help advise on how to do things if there are questions, as well as then queuing up things to add later if they require new xml components.

Honestly I took out a number of the suggestions from Zack (things like wind, etc), because I felt like they were too much work (visually in particular) for too little novelty (we already have gravity wells and repulsors, etc).  Lasers in particular also would be very hard to do, but having some very tight lines of rotating bullets would achieve the same thing and be well within our current code abilities.

Things like phasing blocks I'm unsure about.  I took those off the list, too.  We have the visuals for them already, but they would cause pathfinding recalculations every time they appeared and disappeared, which I'm unsure of the wisdom of.  I'm not sure that they add enough to warrant the extra cpu and coding-time cost.

Binding of Isaac and Rogue Legacy and various others were actually notable to me in just how few different types of things they have in their rooms, to be honest.  It was all about the enemies in them.  That's how I've been trying to approach things here.  We can certainly expand in other directions, but a lot of the various things that we might add in terms of room designs ought to be big and impressive in ways that simply adding new enemies would not achieve.

There's also the long term versus the short term: if people like the core concept enough here, then a mixture of post-1.0 free content and paid expansions can add miiiiles of stuff.  Then again, if there is little enough stuff below a certain bar, then that harms the 1.0 version to the point people might not like it as well, etc.  So there has to be some balance between adding the pre-1.0 stuff and in focusing on polishing what is there versus adding "Ever Moar," which has been a downfall of us in the past.  Generally speaking, overall we want no new major features after the 8th if we want this polished on the 22nd.  I'm out until the 4th.  Yippee...

[crazyroosterman]

Whew, okay, talk about a tight spot.  1.0 release is planned for the 22nd of January, and would be pretty hard to move at this point for... various reasons.  Definitely keep the ideas flowing, and we'll see what we can do.  Hopefully Misery, Zack, and Dayton can kind of coordinate on this and we'll go from there.  I can get things in the editor after new entities (spawners, etc) are implemented in xml, and Keith can help advise on how to do things if there are questions, as well as then queuing up things to add later if they require new xml components.

Honestly I took out a number of the suggestions from Zack (things like wind, etc), because I felt like they were too much work (visually in particular) for too little novelty (we already have gravity wells and repulsors, etc).  Lasers in particular also would be very hard to do, but having some very tight lines of rotating bullets would achieve the same thing and be well within our current code abilities.

Things like phasing blocks I'm unsure about.  I took those off the list, too.  We have the visuals for them already, but they would cause pathfinding recalculations every time they appeared and disappeared, which I'm unsure of the wisdom of.  I'm not sure that they add enough to warrant the extra cpu and coding-time cost.

Binding of Isaac and Rogue Legacy and various others were actually notable to me in just how few different types of things they have in their rooms, to be honest.  It was all about the enemies in them.  That's how I've been trying to approach things here.  We can certainly expand in other directions, but a lot of the various things that we might add in terms of room designs ought to be big and impressive in ways that simply adding new enemies would not achieve.

There's also the long term versus the short term: if people like the core concept enough here, then a mixture of post-1.0 free content and paid expansions can add miiiiles of stuff.  Then again, if there is little enough stuff below a certain bar, then that harms the 1.0 version to the point people might not like it as well, etc.  So there has to be some balance between adding the pre-1.0 stuff and in focusing on polishing what is there versus adding "Ever Moar," which has been a downfall of us in the past.  Generally speaking, overall we want no new major features after the 8th if we want this polished on the 22nd.  I'm out until the 4th.  Yippee...

so I guess that means that hazard/puzzle rooms are going to be a thing then? will puzzle rooms have the threat of damage or death in them?.

[Pepisolo]

Lasers in particular also would be very hard to do, but having some very tight lines of rotating bullets would achieve the same thing and be well within our current code abilities.

I think one of the biggest problems -- and this opinion is based on only half an hour's messing around with the editor so apologies if I'm completely wrong -- is that even though the bullet xml system is awesome and super flexible, to the extent that you could potentially design a rotating laser enemy that would perform pretty much as well as a hard-coded map object, there doesn't appear to be any way to specifically place such an enemy in a room. Since you can only seed enemies via their category -- as far as I can see -- then there's really no point in anyone designing a rotational laser as a pseudo room object because you simply can't place it where you want. However, if those enemy categories had a sub-list of all the enemies within that category which could specifically be chosen to be seeded then that would open up a whole host of new options for room designers.

For example, let's say a new enemy category was created called "Enemy -- Pseudo Room Object" with a sub-list of every enemy within that category available for placement, then designers could create a rotational laser enemy and then actually specifically place it with a particular room in mind. With this flexibility designers could also create rooms intended for very specific enemy types. This seeding only by category thing seems like it could be restricting the creativity of the designers. I'm pretty sure that Misery could make a rotational pseudo room object in no time at all, for example, but there's really no point if there is no control over its placement. Hey, if I'm wrong about the current capabilities of the room-editor (very well could be) then.....doh!

[Misery]

Lasers in particular also would be very hard to do, but having some very tight lines of rotating bullets would achieve the same thing and be well within our current code abilities.

I think one of the biggest problems -- and this opinion is based on only half an hour's messing around with the editor so apologies if I'm completely wrong -- is that even though the bullet xml system is awesome and super flexible, to the extent that you could potentially design a rotating laser enemy that would perform pretty much as well as a hard-coded map object, there doesn't appear to be any way to specifically place such an enemy in a room. Since you can only seed enemies via their category -- as far as I can see -- then there's really no point in anyone designing a rotational laser as a pseudo room object because you simply can't place it where you want. However, if those enemy categories had a sub-list of all the enemies within that category which could specifically be chosen to be seeded then that would open up a whole host of new options for room designers.

For example, let's say a new enemy category was created called "Enemy -- Pseudo Room Object" with a sub-list of every enemy within that category available for placement, then designers could create a rotational laser enemy and then actually specifically place it with a particular room in mind. With this flexibility designers could also create rooms intended for very specific enemy types. This seeding only by category thing seems like it could be restricting the creativity of the designers. I'm pretty sure that Misery could make a rotational pseudo room object in no time at all, for example, but there's really no point if there is no control over its placement. Hey, if I'm wrong about the current capabilities of the room-editor (very well could be) then.....doh!

Yes, this could work.  Just have specific entities that arent really "enemies" but are instead meant to be placed as part of rooms.

The current engine really can allow for some very creative and interesting things.

Granted, there's still definitely issues.  But it's a start.

[Draco18s]

I think one of the biggest problems -- and this opinion is based on only half an hour's messing around with the editor so apologies if I'm completely wrong -- is that even though the bullet xml system is awesome and super flexible, to the extent that you could potentially design a rotating laser enemy that would perform pretty much as well as a hard-coded map object, there doesn't appear to be any way to specifically place such an enemy in a room.

"Lasers" using the bullet XML system should be really easy.  Lots of projectiles that move very quickly with a "source" that rotates (source would be a bullet that doesn't have any speed and doesn't get removed when collided with).

The obscura patterns from TLF came pretty close in places.

I don't know anything about the enemy spawning lists, though!

[Misery]

I think one of the biggest problems -- and this opinion is based on only half an hour's messing around with the editor so apologies if I'm completely wrong -- is that even though the bullet xml system is awesome and super flexible, to the extent that you could potentially design a rotating laser enemy that would perform pretty much as well as a hard-coded map object, there doesn't appear to be any way to specifically place such an enemy in a room.

"Lasers" using the bullet XML system should be really easy.  Lots of projectiles that move very quickly with a "source" that rotates (source would be a bullet that doesn't have any speed and doesn't get removed when collided with).

The obscura patterns from TLF came pretty close in places.

I don't know anything about the enemy spawning lists, though!

Yeah, these things are pretty easily done actually.  They're the sort of things that are nice and quick to make.

The bit about the editor and such though sure aint my doing, heh.  I think either Chris or Keith made that.  Hopefully a function like that wouldnt be too hard to add.