|
Post by Marisa on Sept 2, 2018 20:08:13 GMT
I mean, some observations I made that could make this game better and/or are making this game worse and I have probably been banging on about these since when I played this back in 2011-2013.
----
Spawnkilling: Seems to be a fundamental tenet of hallway. If you could somehow make hallway hard to spawnkill in then you'd probably also be capable of finding brain cells in a bureaucrat.
AoS 1.x used to have a horrible spawncamping problem where the spawns were bunched together and you could easily farm them with the Commando's MG, which was really only useful for that back in protocol v122. A suggestion was made that if you spread the spawns out, it becomes very difficult to spawncamp effectively.
The most notable use of Iceball was by Jagex when they were looking for better spawn points. It actually started out when I was curious to see how well it'd perform on Shauny's laptop, and then he found that a map viewer I had there was actually useful for this. Tom ended up using it later and somehow managed to fix Arctic Base's spawns, which is something I would consider to be a miracle considering how bad the cover is on that map.
The irony is, while you still could spawncamp with spread-out spawns by having at least one person per spawn area, it theoretically dampens the effect that squads have when applied to this situation.
Now if someone could come up with a way to make hallway not so prone to this shit I'd be impressed.
----
Weapon spread: AKA "Why I Hate 0.75 In One Weapon Structure Field".
Rifle should always be balls-accurate.
The SMG being balls-accurate would be interesting. You could actually balance this. I recall 1.x had an SMG that was actually useful at range and yet wasn't an overpowered piece of shit - it actually required a fair amount of skill to use. The way 1.x balanced this was to have each bullet do the equivalent of about 15 damage per hit. There was more for the head but eh, I'd argue against making headshots special for the SMG.
The Shotgun being balls-accurate would almost make it into another rifle. Almost. The reload pattern is different. Oddly enough during 0.55 beta testing (I may still have at least one of the two versions of this I had somewhere, IIRC it was really only tested within the <@> clan so lul opsec), which eventually resulted in 0.58, the shotgun had a more realistic spread. This eventually resulted in the 0.60 "sniper shotty".
----
I'd like to cover squads and how to balance those out (short ver: go back to /follow and enforce a strict tree) but this train ride's almost over.
Also I'd like to elaborate on some ideas of how to balance the shotgun at some point, that sounds like fun.
|
|
|
Post by Marisa on Sept 3, 2018 6:28:11 GMT
Eh, might as well do another post. At least before someone says "go home you're drunk" (oddly enough I'm in the process of going home). It might be useful. It might not. Whatever. I don't really care that much about the utility if it's at least interesting.
----
Unbreakable-floor maps (most notably pinpoint, hallway, and goonassault)
When everything takes place on the unbreakable layer you lose the ability to trench. I think this is why you have people who say that custom maps ruined AoS.
I have a couple of cheapass map variants lying around which IIRC were done via some Python script I may or may not still have. One is "normandieplus10", which allows for more tunnelling and it helps w/ team balance, but you can still trench in the normal version so that's kinda off the topic of this thing, and "pinpointplus2" which allows for at least 2-high trenching.
Now, years ago I had a terrible idea for a map I'd call "bridgepoint", even though there exists some other lame-arse map called that which literally nobody cares about so I could actually get away with calling it the same thing. The idea was to pretty much take bridgewars or whatever the proper form of the name was, and replace the bridge with one resembling pinpoint.
That then got me thinking about how thick the bridge should be and I eventually started thinking about pillars to put on it. If it's wafer-thin then it gets a bit ridiculous. But if you make it thicker... hey presto, you can trench again and all is fine with the world. I suspect the optimal thickness would be either 3 or 4 if you want to reduce the amount of tunnelling that can happen. Shove in a dickload of 2x2 pillars and you've got yourself a bridge that's hopefully less likely to fall down than anticbridge. (Seriously, fuck anticbridge. Never admining that shit again.)
I think I got somewhat sidetracked here. Oh well. Let's say a bit more and then wrap up.
The Minit Assault server had a few variants of goonassault on it. My personal favourites were the ones where the middle section was raised up 2 or 3 higher (there was also one with pre-cut tunnels but I forget how high that one is, and I also don't remember if I preferred the 2-high or 3-high one though I think I preferred the 3-high one).
TheGrandmaster IIRC did a couple of variants of maps, Assault2 and Pinpoint2, which are a lot "smoother" and stuff looks like islands and whatnot. They're kinda hard to give justice to with a textual description so check them out some time if you can get them. Pinpoint2 doesn't just raise things up though, it also makes a few rivers cutting straight through it.
I suspect that a raised-by-2 treatment could... OK, I redact that before I finish it, I suspect it wouldn't make that much sense on Hallway as everyone's always high up and pot-shotting people. Unless, of course, one raises the spawns 3-high and builds a roof... But you'd want to elevate the roof a bit higher anyway.
----
Squads
Yeah alright I might as well fill this one in.
The reason this came about was because /follow was a thing, and some people made the observation that two people who used /follow on each other could behave like what would happen nowadays if you had two people in a squad.
We can talk about how overpowered this shit is or we can attempt to push a star-shaped peg into a flat-design-cloud-shaped hole.
(( Also why the fuck does this site peg a whole fucking core I'm on a laptop and I'm on battery that is some bullshit right there. Typing this in vim now. ))
So anyway, we somehow ended up with squads under the notion that Triplefox had an idea and probably wanted to try it because it would be interesting. Fast forward about a year and it turns out it has the wonderful power of making people even more likely to leave in certain situations.
And that's where the problem ultimately lies. Not so much that it's overpowered, but rather it stops being fun.
The never-ending "ALL /SQUAD 1" chant doesn't help here. Let's propose a model.
Suppose that we have a full 16 vs 16 server. Now suppose that the whole entire Blue team is all in squad 1 - oh sorry, I mean /squad 1. And suppose that the whole entire Green team is in the best squad, namely /squad none. (This is the best squad to be in by the way.) Now suppose that Blue get into the Green spawn and start reaping...
You can't clear that crap out fast enough. Green starts dissipating. Then Blue notices that there's nothing left to click on. Blue starts dissipating. (If anyone wants to write a simulation for this then be my guest. My model's kinda naïve here.)
If anything this presents an argument for having a cap on the number of people you can have in a squad. I don't believe in a universal limit for this - a squad of 3 when there's only 3 people to fight against does tend to be hard to get rid of. You'd have to scale it by player count and I can see a nice fun source of bugs happening here.
An easier solution is to cap it at 2. Even then you get the bullshit where one can hide away and be a portable field spawn point. I'd argue that it would be much more appropriate to only apply it to the players who are in their team's spawn, and then it becomes a purely organisational thing.
Anyone got any ideas on how to make this better?
----
And that's that round of rants now that I'm finally out of the tunnel that confines my internet connection to being nonexistent.
Maybe next time I'll start toying with the idea of a spreadless shotgun.
|
|
|
Post by Scoper on Sept 3, 2018 7:46:38 GMT
Spreadless shotgun would be cool, it'd kill everyone with 1 shot but the shooting/reload would have to be longer
|
|
|
Post by Marisa on Sept 3, 2018 20:08:55 GMT
Still not motivated to do the spreadless shotgun rant, sadly.
----
Greener pastures (Oh shit these rants are getting longer.)
Remember Minecraft? No, not that one, the one that ran in a browser.
After however many years it was (I think about 5? Possibly more? Probably not less?), the original version was retired and removed from the Minecraft website. But never despair; ClassiCube is here! Or something like that.
But basically, if you want to play Minecraft Classic (I swear this isn't an ad): 1. Go to ClassiCube. 2. Register an account. 3. Download ClassicalSharp (readers take note: This is a custom client.) 4. Join a server and play on it.
How many custom Minecraft Classic servers are there? Well, for starters I wrote one in 2009... and I'm *guessing* there's about 30 right now. No, not server instances, but server software.
And then there's Ace of Spades.
Well, there's a site called Build and Shoot which actually managed to get the game onto life support and set up a forum and then... didn't really go much further.
MC Classic has ClassicalSharp. We have OpenSpades, which was originally an outside effort.
Sounds like we're doing alright, huh?
Except we aren't. The MC Classic community has actually extended the protocol and embraced the protocol extensions that have been made. There's a standardised set of extra block types (includes actual functioning ladders and ropes), custom block skins, and I believe one can define their own types if I'm not mistaken.
We're stuck on 0.75 for the most part. 0.76 is an option but not one server owners really bother with seeing as it's just too hard... although I now have a script which does all the hard work for you but I'm not sure how best to drop it.
At least the majority appear to use OpenSpades (I'd estimate 60%, though it could be closer to 70%, I'm not willing to bet on 80% though), and the majority of script kiddies tend to use the vanilla client, so if this game's got a future it's either OpenSpades or something else (VXW anyone?).
However, we've still only really settled on something that can do unmodified 0.75 or unmodified 0.76. From my experience, if you try to push for anything else, you get pushed to a forum that nobody ever reads, unless, say, one of your friends manages to take over your subreddit because you care about it about as much as you care about the rest of the game...
Anyway, player counts seem to vary from about maybe 100 down to maybe about 20. When I left in ~2013, 100 was roughly the minimum. Of course, 1.x seemed to be feeding players to 0.x, but that's pretty much gone now so I expect the player counts to dwindle to a lower equilibrium.
At this point if anyone wants to make some progress, here's an idea:
1. Come up with a good solution to extend the protocol and query extensions. I'd use 0.76 as a baseline. 2. Write a script which converts the server into a server which services this protocol. I've got one which turns a 0.75 server into a 0.76 server and it even works on piqueserver (not extensively tested) - we use this script on our Goon Haven server. 3. Start an OpenSpades branch which adds support for the extended protocol. 4. Come up with an extension which enhances some existing script, but is backwards-compatible with 0.76, just feels a lot better when you have the enhancement. 5. Do a pull request for OpenSpades.
Stick with backwards-compatible stuff until you've got enough traction and enough users to break compatibility.
----
Anyhow, train ride's over, and I'm tempted to have a go at that protocol extension query thing.
|
|