Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dariush]

Feature creep isn't a thing in roguelikes. At any given moment more features is better than less, regardless of all other factors. Look at DF, Nethack, Dungeon Crawl. All of them are in constant development and nobody complains about feature creep. Moreover, EVERY roguelike MUST have randomized maps. That's in the very definition.

Roguelikes are centered around Dungeon Crawling through randomly-generated environments randomly stocked from a huge list of monsters and items. Some (such as ADOM) also have a static overworld and/or special levels, but even those games rely on random content in other places.

There's just no point in playing in the same map over and over again.

There is a class called Zombie in the game. It is a subclass of NonPlayerCharacter which is a subclass of Creature (PlayerCharacter is another subclass of Creature). So far those are all the creature classes there are. You can differentiate a zombie from another creature by checking whether it originated from the Zombie class.

Hm. Maybe you should scratch the zombie class and instead implement the flag I mentioned? It would allow for stuff like zombie animals and, depending on the implementation and the similarity to other classes, greatly ease the process of zombification (i.e. you'll be able to simply change one value instead of moving the object to a completely different class).

[finka]

I'd favor a Grindhouse style revenge motif(Awake in cabin, family just newly dead, player character utterly snaps and goes on frenzied bid unto destruction..or something), emphasizing far less on the mathematical details versus over the top exaggerated cause & effect in terms of physics so as to get nicely visceral and lethal on all ends.

Well, not to say we couldn't, but the roguelike "permadeath is always just around the corner; weigh your options carefully" ethos is a bit at variance with "snaps and goes on a frenzied bid unto destruction".

Regarding choice of map style and all that.  How much do you care about a consistentish backstory?  Urban is at least easy that way: the cities are where all the people are to start with.  If instead the maps you implement (now) are traditional roguelike cave / dungeon style, why would there be zombies congregating in such places?  Because that's where the citizenry decided to hide out (odd choice, I'd think) before they succumbed?  Because the zombies themselves are drawn to the deep dark places?  etc.

[Dariush]

If instead the maps you implement (now) are traditional roguelike cave / dungeon style, why would there be zombies congregating in such places?  Because that's where the citizenry decided to hide out (odd choice, I'd think) before they succumbed?  Because the zombies themselves are drawn to the deep dark places?  etc.

- Because miners;
- Because you were hunted by zombies until you stumbled into this place. Their sense of smell is worse in tight places, so they aren't able to instantly track you. Or maybe you just blocked the exit;
- Because zombies are resurrected corpses, which are typically found in the earth;
- Because it was once a city, but survivors built another one above after it succumbed to zombies, and you're stuck inside.

[Tuplis]

Feature creep isn't a thing in roguelikes. At any given moment more features is better than less, regardless of all other factors. Look at DF, Nethack, Dungeon Crawl. All of them are in constant development and nobody complains about feature creep.
...
Hm. Maybe you should scratch the zombie class and instead implement the flag I mentioned? It would allow for stuff like zombie animals and, depending on the implementation and the similarity to other classes, greatly ease the process of zombification (i.e. you'll be able to simply change one value instead of moving the object to a completely different class).

Wait, what? Feature creep isn't a thing in roguelikes? Nobody complains about feature creep?
Feature creeping has the following effects:
- Long development times
- More stress for the developer

Now, correct me if im wrong but as far as I know, these are typical attributes for roguelike development:
- Long development times (how long have nethack and DF had between releases? LOL!)
- Lots and lots of failed projects (because the developer got frustrated and quit)

I dare say that out of all game development, roguelikes are THE most susceptible to feature creep because they are in almost 100% of the cases not productized (focused in scope) properly.

Also, don't worry about zombification, etc. I've generalized to the point where I don't have to be concerned with issues like the one you pointed.

[Tuplis]

Regarding choice of map style and all that.  How much do you care about a consistentish backstory?  Urban is at least easy that way: the cities are where all the people are to start with.  If instead the maps you implement (now) are traditional roguelike cave / dungeon style, why would there be zombies congregating in such places?  Because that's where the citizenry decided to hide out (odd choice, I'd think) before they succumbed?  Because the zombies themselves are drawn to the deep dark places?  etc.

I'd say we can justify any map style whatsoever. It's more about what type of a story we're telling. For example, we could easily go for the standard cave-type dungeons that are in almost all roguelikes. Consider this scenario: The government anticipated a zombie outbreak and had huge vaults (a la fallout) built and the people were relocated there to weather out the storm. Now, conveniently for us, the quarantine goes to shit at some point and voila, we have caves full of zombies, right? Hero needs to find his way out or sth.

I'm not saying this is the one we definitely want to choose, I'm just saying we should pick a style that's most likely to be popular and fit the story around that.

Personally I'd lean towards some rural setting (forests, fields, sparse housing) as my original, intuitive idea was. City has been done (Rogue Survivor, which I respect immensely and would rather not copy) and obviously 90% of the roguelikes explore caves so this would just be a dungeon crawler with another skin.

[getter77]

I'd favor a Grindhouse style revenge motif(Awake in cabin, family just newly dead, player character utterly snaps and goes on frenzied bid unto destruction..or something), emphasizing far less on the mathematical details versus over the top exaggerated cause & effect in terms of physics so as to get nicely visceral and lethal on all ends.

Well, not to say we couldn't, but the roguelike "permadeath is always just around the corner; weigh your options carefully" ethos is a bit at variance with "snaps and goes on a frenzied bid unto destruction".

Technically even permadeath isn't of the utmost incredible importance, but the "around the corner" part certainly isn't a given----windshield killin' is a long established trope.  Somewhat relevant, there was a competition entry some months back involving a Dwarven Berserker descending into a mine in such a way that death was inevitable and in the end it all came down to how far down you could make it and how many Orcs and such you could take with you.

I see having this more stylized/pomp/swagger to be a good route to take it as opposed to simulationist as the latter generates depth in the sense of filling out a rubric that is kinda locked down in peoples minds at this point between Cataclysm, Rogue Survivor, etc---yet it strikes me that this project, one meant to evolve in fits of punctuated equilibrium and general organic growth, would do well to keep the player guessing and to shoot for "the experience".



As to the feature creep part:  Yes folks, trust me, it is a thing in Roguelikes in a big way....the "Talkie" phenomenon being the biggest graveyard.  The like of Dwarf fortress working as it does is moreso the exception beget of a confluence of disparate circumstances, not the rule.   Creeping ain't a bad thing though when paired with released early/often, but can still get out of hand before you get a foundation proper.

[Tuplis]

I see having this more stylized/pomp/swagger to be a good route to take it as opposed to simulationist as the latter generates depth in the sense of filling out a rubric that is kinda locked down in peoples minds at this point between Cataclysm, Rogue Survivor, etc---yet it strikes me that this project, one meant to evolve in fits of punctuated equilibrium and general organic growth, would do well to keep the player guessing and to shoot for "the experience".

For one, I'm interested in what you are proposing. What do you feel would be the next few steps implementation wise to take this game to the direction you're describing? If we simulate that atmosphere by maybe making a map or two (quickly done) and implementing line of sight / guns, we would get some usable features even if we decide to ditch that route. Please describe in some detail what kind of a scenario you'd like to try, getter77.

I definitely want there to be a factor (besides the zombie apocalypse which is very popular at the moment) that differentiates this game from all the other roguelikes out there so giving this game a little "arcade" feel might be a good idea, actually.

What is everyone thinking?

[Dariush]

What is everyone thinking?

"Mmmm... I sure would like some ice cream right now..."

[Tuplis]

A new version is out:

-Implemented line of sight (Bresenham)
-Implemented sight range (5) for the player

Next I think I'll configure the xml builder to also build items.

[getter77]

I see having this more stylized/pomp/swagger to be a good route to take it as opposed to simulationist as the latter generates depth in the sense of filling out a rubric that is kinda locked down in peoples minds at this point between Cataclysm, Rogue Survivor, etc---yet it strikes me that this project, one meant to evolve in fits of punctuated equilibrium and general organic growth, would do well to keep the player guessing and to shoot for "the experience".

For one, I'm interested in what you are proposing. What do you feel would be the next few steps implementation wise to take this game to the direction you're describing? If we simulate that atmosphere by maybe making a map or two (quickly done) and implementing line of sight / guns, we would get some usable features even if we decide to ditch that route. Please describe in some detail what kind of a scenario you'd like to try, getter77.

I definitely want there to be a factor (besides the zombie apocalypse which is very popular at the moment) that differentiates this game from all the other roguelikes out there so giving this game a little "arcade" feel might be a good idea, actually.

What is everyone thinking?

In terms of concrete mechanics..hmm..

-Physics, particularly of the dramatically exaggerated sort.  Most base form:  Knockback and Momentum/Weight/Collision.  Zombie bashed by player with a Sledgehammer, through a window/glass sliding door, smashing at last into a nearby tree to either then crumple in a broken heap, get impaled on a branch, or fell the tree from the impact depending on the state of the tree, circumstances,etc.

-Weapons that center on their utmost function, not so much chipping away at die rolls.   A Sledgehammer favors literal crushing and sending things flying, an Axe dismemberment of all sorts, and so forth.  "El Ray the RL" might be a bit of a stretch, but if you've seen the movie you can generally figure it from there.  Positioning, timing, resources at hand employed, and the circumstances at the time being the defining aspect of combat.

-In other words, as it was in Grindhouse/Planet Terror, turn the convention of "survival" upside down a tad via liberal application of Power Fantasy such that you still have to survive outlandish horrors, but will start out far from being the usual weakling versus the massing hordes of whatever while getting better still.  You can only get so over the top when approached mathematically, but logically and thematically?  Imagination and scenario are the only things to deal with outside of the tricky coding.

[Cities_Burn]

Limb damage system is always good 

[Dariush]

Since Google Docs isn't a good dialogue platform, let's take the noise/hearing discussion here.

You yourself said that every feature should be broken down into the smallest possible components. It would be hard for creatures to react to noise that isn't there, wouldn't it?

In any case, here's how I propose it should be done: Every action has a loudness rating (0 for silent actions, gets progressively higher for louder actions) and a type - shot, branch crunch, zombie moan, human moan, fighting, etc. At the time of an action, a sound with the initial loudness is generated at the tile where the action happened, which then propagates into every direction, losing 1 loudness per tile, until it reaches, let's say, -4. Obstacles such as walls give a penalty to the sound that passes through them. If the sound of different loudness reaches the same tile by two different ways (for example, after going around the wall), the higher value is taken. Creatures have a hearing modifier that is expressed in integers between, for example, -5 and 5. This modifier is added to the sound perceived by this creature. If it is above 0, the creature heard the sound. If it is below 5, the creature isn't certain about the origin and slowly goes to investigate. If it is above 5, the creature searches for the origin of the sound at a run. Additionally there's a set of exceptions: for example, a certain type of zombies may be repelled by a wailing siren, which attracts some other kind more strongly. Player receives different messages depending on perceived loudness and type of the sound, optionally he is unable to recognize the type if the sound is below a certain threshold. Just for kicks, we may add a random queit ambient sounds that a good listener may mistake for zombie footsteps.

[Tuplis]

Since Google Docs isn't a good dialogue platform, let's take the noise/hearing discussion here.

You yourself said that every feature should be broken down into the smallest possible components. It would be hard for creatures to react to noise that isn't there, wouldn't it?
...
<Technical details>

This is what I said:

Every new backlog item needs to have end-user value.
...
When adding a new item, consider whether you can break it down to independent functionalities out of which at least one has end-user value in itself. Iterate this process until you can no longer break the feature down.

Making sounds does not have intrinsic end-user value unless a monster can react to them. Therefore, making sounds is not a backlog item. You made it a backlog item and as a result, we're having a non-implementer (that's you) listing down implementation details.

I explained the concept once already: You're supposed to implement large underlying systems (making sounds) after you can justify the effort with backlog items that use the system and have end-user value (hearing, smell, general ai). If we include backlog items that have have no end-user value, we will get stuck developing "features" that will not be available because the content part (ie. reacting to hearing and smell impulses that are available) is missing. First you decide what you want (that's the job the community has), then how you want it(that's my job).



edit: just released a new version with ranged weapons! Works with the xmlbuilder so you can play around with it (pistol.xml) check the new keybinds from the help file.

[qwertyuiopas]

The player also gets to use sound, though. So, for two features that could justify sounds: Player hearing (listen closely at a door, and you may hear a zombie moaning on the other side. Or around corners. Or perhaps a distant wolf (in an open above ground setting) that signifies both relative lack of zombies, and a possible risk of encountering more dangerous creatures), and "sleeping" opponents (could just be distracted or inactive) (wake when they hear something loud enough)

It would also allow for more stealth-type gameplay, as you could throw a rock at a distant wall, drawing attention in that area, and then quietly pass through another route, where the now-distracted zombie would have noticed you.

But as a simple "reasons to introduce this feature", sleeping opponents and an additional sense for the player can be added long before there is any sort of decent AI for enemies, and both of them would have effects clearly visible to anyone playing.

[Dariush]

So, you took a full post to bash my heretical view of your pa-a-aradi-i-igm, but don't so much as review the technical details I proposed. Nice going, there. I can see bright future for this game.