World of Warcraft Pixel Botting - A Cross-Platform Odyssey

[developerer]

I don't see what asking questions has to do with what I pointed out. Something can be learned from everything and asking questions does not mean that I consider the product perfect or good (how could I, with one video). I'm simply interested in how you've things, not looking to do them the same way necessarily.
I was also referring to the second video btw.
edit: I also didn't ask a single question about the bot, just general programming questions. I'm sure you're a way more proficient coder than me, but this is not what this is about.

[Wontuse]

I don't see what asking questions has to do with what I pointed out. Something can be learned from everything and asking questions does not mean that I consider the product perfect or good (how could I, with one video). I'm simply interested in how you've things, not looking to do them the same way necessarily.
I was also referring to the second video btw.
edit: I also didn't ask a single question about the bot, just general programming questions. I'm sure you're a way more proficient coder than me, but this is not what this is about.

Genuinely curious about this problem space and want to hear how others have implemented solutions to get from Point A to Point B. Not the pathing algorithms, but the actual character movement. I see the issue being two different problems, mapping and movement. Of those two, I'm only aware of solutions being implemented like so:

Mapping

• Navmesh
• Predefined paths
• Defined boundaries

Movement (whether it be done in process or externally through keyboard/mouse control)

• Click to move
• WASD movement/turning
• Auto run
• Mouse drags

The bot I'm writing, uses predefined paths, and auto run + mouse drags. I did try WASD movement/turning. I know it takes 2 seconds to complete a 360* rotation using a key. From there it's simple math to determine how long to press a key to turn N degrees. But this method was too slow. With mouse drags, the bot can do a near exact 180* turn in a couple milliseconds. This allows the bot to turn corners w/out the need to stop or slow movement.

I'm sure there's no magic bullet, my bot can get around Stormwind pretty well. I haven't implemented it yet, but it should also be able to go between zones w/in a continent using either flight paths, or by getting to a Node that will connect two graphs (zones).

What do others do if they want to navigate tight spaces that leave very little margin for deviation? Spiral staircases (getting to the FP in Thunder Bluff), caves, etc. Or don't they?

[mikeboy69]

.3 seconds for each cord read probably doesn't help and it's probably why you overshoot cords that are close so often. python is too slow maybe some sort of location predictor could be used based on speed and distance to next node to sort of stop a full screen read needing to be done on super close nodes that you know the .3 second cycle will not do well with. other than that, go slower or use a faster language XD

[Wontuse]

.3 seconds for each cord read probably doesn't help and it's probably why you overshoot cords that are close so often. python is too slow maybe some sort of location predictor could be used based on speed and distance to next node to sort of stop a full screen read needing to be done on super close nodes that you know the .3 second cycle will not do well with. other than that, go slower or use a faster language XD

Agreed. Python is slow, though I suspect with a better computer I could halve the read time.

So far the best way I've been able to handle passing waypoints is to constantly get a new path to the destination and based on movement speed, looking ahead N waypoints. This way it will never pass a waypoint, because it is always chasing one.

It works well enough, though I do still desire to further refine it. The only thing the bot will absolutely not be able to do is navigate in places where I would need a Z axis as well. Caves, the FP in Thunder Bluff, etc. I'm curious how other bots handle movement in these areas?

[mikeboy69]

maybe for flight masters you could get a rough estimate of Z cord by marking the flight master and comparing it to you.

[mikeboy69]

you could also have a set routine for flight masters like thunderbluff where you know the starting z point and the ending z point and fly blind based on that just using x and y + speed to determine movements and climb rate to estimate z. if stuck hearth out or log. but most people probably avoid these areas.

walking probably helps trace to predefined path up more accurately but also probably makes you look like a bot.

[Wontuse]

Okay - hear me out, is there enough people out there who want/use out of the box leveling bots? Or did I misse the boat by 10 years? The bot will work on any version of WoW, both live and private servers. I believe I can subvert any issues with detection sans being reported, or Blizz taking a screen shot and manually identifying pixels. Which can be set at 1x1 and scattered all over the screen to further make it difficult.

I think the issue with other bots, is that 1) they're detected because they modify the game through injection or 2) they all use the same pixel addon that can be hashed and checked for. No?

I may spend time coming up with a way to share the pixel addon and basic extractor code. But first, I will essentially mangle the addon so that every single download of it is unique. Unique variable names, reordered functions, different number of files, unique addon name. Worth it to others? The extractor code is simple and doesn't do much else on its own.

I'm wondering though if there is enough money on the table to pursue furthering development? I can pretty much leave the bot unattended for 8 hours in a zone. And I've been improving my pathing to go between zones via FPs or running. Really what I'm selling is access to a unique proxy (Blizz can't check for network calls if every bot uses a different endpoint) pathing data that is maintained, out of the box rotations, and near guaranteed avoidance of detection. Plus the server can notify of and allow you to respond to whispers, receive updates, maybe even to some extent control it remotely? The bot would be as simple as

• Download bot
• Install addon
• Load game
• Go AFK

Maybe the extent of avoiding detection is overkill these days? Does Blizzard even care, unless they can identify and ban tens of thousands in a go?

[nooberang]

What do you even need Dijkstra for? Lol.
The wall of text makes it sound like you've developed something amazing, the videos show otherwise.

[Wontuse]

What do you even need Dijkstra for? Lol.
The wall of text makes it sound like you've developed something amazing, the videos show otherwise.

The bot has recorded dozens of walkable paths in numerous zones such as: nearest repair/vendor, flight master, grind paths, spirit healer, paths between zones. Based on world state, the bot needs to get to a destination. If on the grind path and durability is 0, go repair. If the character dies, go to corpse, if it levels out of a zone, go to the next grind path

The videos are just following a path. Perhaps a better suited video would be the bot leveling a new char from 1-10.