Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - KK20

Pages: [1] 2 3 ... 143
RMXP Script Database / Re: [XP] Lagless Path Finder
« on: May 23, 2017, 12:50:00 AM »
You could, but no one is going to fulfill that request.

Welcome! / Re: OI
« on: May 23, 2017, 12:49:12 AM »
Sup newbie :V:

Post topics or join our Discord server for any questions you may have.

RMXP Script Database / Re: [XP] Lagless Path Finder
« on: May 21, 2017, 12:32:06 AM »
Remove the "if" in your statement and it can be used in a Conditional Branch.

The Terrain Tag feature is not built into the script. It would need to be manually added by a scripter.

Was thinking back on this script one day and realized "I could make this a little bit better" I did. :)

Updated to version 2.0
Some refactoring and added comments. Most important change though is how the scripts are exported and imported. Originally, I had named exported files something like
where the 100 in this case means this script is located at index 100 in the script list. This was fine and all, but it made adding/removing/changing order of the exported scripts extremely tedious.

Now, I create a CSV file that lists the filenames in order. This completely solves the above problem as you can just edit the CSV file directly to add/remove/move scripts around.

Just needed a good quality-of-life update. I don't see how this script can be improved upon, so this will probably be the definitive version (hence the jump from v1.2 to v2.0).

Script Requests / Re: Mrmx-os script to can use the keyboard
« on: May 17, 2017, 12:41:40 AM »

Script Requests / Re: Mrmx-os script to can use the keyboard
« on: May 16, 2017, 09:08:18 PM »
I'll repeat again: did you install all the fonts in the Fonts folder?

Script Requests / Re: Mrmx-os script to can use the keyboard
« on: May 16, 2017, 07:52:23 PM »
i cant see nothing that i write in the textbox in the login.
That kinda sounds like a Font file wasn't installed.

Script Requests / Re: Mrmx-os script to can use the keyboard
« on: May 15, 2017, 10:40:08 PM »
What? Your English doesn't make any sense to me.

It says in the script instructions:
Code: [Select]
# Also place all weapons ID except SP damaging and absorbing ones, into the
# "ALL_WEAPONS_IDS", because if you don't, a nasty bug will prevent you from
# dealing damage with normal attack. Cool, huh ?
The line the error code is pointing to is
Code: [Select]
self.damage *= elements_correct(attacker.element_set)The error is saying the variable or method damage is currently equal to nil. The class NilClass has no defined method for the * operator.

The reason damage is nil is because you didn't configure the script correctly, as indicated above.

RMXP Script Database / Re: [XP] Blizz-ABS
« on: May 14, 2017, 12:24:06 AM »
If you can implement a legit 2 player without networking, I will honestly praise you as a god.

RMXP Script Database / Re: [XP] Blizz-ABS
« on: May 13, 2017, 12:43:16 AM »
That's for AI-controlled party members. Completely different from 2 players.

RMXP Script Database / Re: [XP] Blizz-ABS
« on: May 12, 2017, 09:58:56 PM »
It's just the way how the two scripts work together. One player "hosts" the map while others connect to it. All the event processing is handled on this host while the other clients are simulated in a sense. Since the connection is handled with TCP sockets, you won't get as smooth of a gameplay experience. Some features of Blizz-ABS are disabled as well. Honestly, it's easier to understand if you know how networked games actually worked.

If you just want to make a game for fun, then there's nothing wrong with trying it out.

RMXP Script Database / Re: [XP] Blizz-ABS
« on: May 12, 2017, 08:56:05 PM »
As far as I'm aware, no. It was a planned feature for later updates, but winkio dropped support for a few years now, so it's never going to happen most likely. Your best bet would be to use RMX-OS in conjunction, but there are going to be problems still.

Script Troubleshooting / Re: Problem with my pathfinder
« on: May 12, 2017, 03:46:31 AM »
I guess my sentence can be interpreted many ways. What I really wanted to say was something along the lines of "a square is a rectangle but a rectangle is not a square"--your script is more Dijkstra than it is A*.

F0's pathfinder script is a true implementation of A* in that the heuristics are incorporated into the node costs. It doesn't perform as fast as your script though (~0.44 seconds) due to the sorting it does to find the lowest cost (kinda similar to Epherex's heaping methods), but it does evaluate less nodes.

Funny thing though is that if I remove the Math.hypot_squared in your script (and make it pure Dijkstra), it actually performs faster, at least in regards to the demo project Epherex provided (~0.177 sec). Of course, in terms of straight line path requests, it still pales in comparison. The data types you used are key to its speed.

I tried taking a stab at the problem in making my own pathfinder. It's very rough since I was mainly experimenting, but it does get the job done. Print statements commented out.
(click to show/hide)
It performs somewhat faster than your default script (~0.22 vs ~0.27), but not as fast with the Math.hypot_squared removed.

And yeah, tests are performed assuming all tiles on a Cartesian coordinate system have an equal cost--my script would break otherwise. I'll keep looking into improving this.

Script Troubleshooting / Re: Problem with my pathfinder
« on: May 11, 2017, 08:18:13 AM »
It still doesn't explain why it doesn't perform how A* should ideally. If you change the part where it checks for distance from the current node to the target and THEN check for path cost, then it will definitely behave more like A*, but, as I mentioned in my first post, it becomes greedy. A* needs to factor in both the path cost and a heuristic together to get the absolute lowest cost. You're using two values separately with a bias (which is not that much faster than Dijkstra to be honest).

I don't know what you mean about me being wrong about Dijsktra when I never talked about it.

Script Troubleshooting / Re: Problem with my pathfinder
« on: May 11, 2017, 05:05:45 AM »
I've studied Blizzard's and your scripts. Your script evaluates about half as many tiles as Blizzard's on the demo map, but Blizzard's finishes in about a quarter of a second while your's takes about 1.75 seconds.

I believe the main offender to this time difference are your heap methods and chosen data types. You are constantly swapping items around in hashes and arrays that, when they become very large in size, get very slow very quickly. I also don't see how it is guaranteeing that you will be evaluating the lowest cost node on the next loop iteration as your update_heap skips over potential good candidates (newly inserted node checks halfway through the open list to see if it is a better cost--if yes, then swap; repeat until no longer the better cost). I really don't know why you have both an open nodes array and an open nodes hash; just choose one.

Blizzard sticks with a open nodes hash where the key-value pairing is
Code: [Select]
[x, y, cost] => directionHe then chooses a node from the hash that has the lowest cost in its key; otherwise, the node closest to the goal is chosen. This isn't really A*, but more of a modified Dijkstra. There are no heuristics really being done here. If I choose a location that is like 5 tiles away, it evaluates well over 5 tiles to find the best path (refer to my previous post's animated image where it makes a beeline to the goal in the beginning). Your script functions exactly as how I expect it to.

The TL;DR version = Your heaping is slowing everything down unnecessarily. Please take a step back and think of a better way to sort your nodes.

Script Troubleshooting / Re: Problem with my pathfinder
« on: May 10, 2017, 08:43:03 AM »
I don't see anything wrong with the code. The simple act of removing @cost from your total node cost just makes the algorithm greedy--it evaluates nodes that look like they are close to finding the path while disregarding optimization (i.e. the shortest path). By putting @cost in, the algorithm first favors nodes that are close to the goal then those that have not been evaluated a lot. At some point, the length of the current path (which might be the correct one) has a larger cost than a node that is close to the beginning.

Wikipedia has a good visualization on what you're experiencing:
(click to show/hide)

Lexima Legends IV - Chaos Project / Re: Bugs Megathread
« on: May 05, 2017, 07:58:31 PM »
Resolved. Map wrapping was left enabled accidentally.
Thank god I didn't have to do anything

Pages: [1] 2 3 ... 143