ExilebuddyBeta #903/#905/#906 Feedback and Discussion

Just wanted to say, Pushed, that in just the week since I bought this you have made incredible improvements in the bot. It wasn't doing so good at first but now it seems to be working much more smoothly. Good job man! Try sleeping occasionally though

 

i swear pushed doesn't know what sleep is... every day i wake up and he's pushed more fixes out if the game devs were half as on top of shit as he is... well we wont think about it :x

 

agree -_- /

 

https://www.thebuddyforum.com/exile...75565-beta-quicksilver-support-autoflask.html << this link Commented pushedx

 

Seems for me, that it ignores the strongboxes, was working last #build. (Got the GlobalChestIgnore unchecked for it)

 

Push it real good.

 

For some reason for me it still isnt moving fluently, still kinda stutter walking, moving and stopping after short distances for half a sec and so forth not sure what the problem is

I've tried on a VM with dedicating between 1 and 4 cores on an i7 3770k with an ssd and same result, regardless resources I allocate, and in task manager bot is using like 5% CPU so I doubt its the resources im allocating?

Just moltenstrike and movement skill on bars, not sure if there's anything else that needs configuring to avoid stutter movement.

Combat seems to be fine, movement however is not.


edit: seems linked to the fps of the game, which I normally limit since it consumes too much cpu running it uncapped, however the bot has all the resources it needs. The previous release versions of exilebuddy had no issue with capped fps on PoE, is this ever going to be fixed or will it remain this way, I see it's part of known issues but will it be adressed? Because if it won't it's a huge resource hog to leave poe running with fps uncapped, mine goes in the hundreds...

 

Have you tried the following?

At Settings->Main,

change

MsBetweenTicks to 5

and

InputEventMsDelay to 25

?

 

Yeah, it's directly linked tot he fps of the game... if I uncap fps for PoE bot will move fluently, otherwise it stutters all the time, don't understand why, previous release versions were running fine with PoE FPS capped.

 

It's to do with Poe, when you limit fps, you limit the input rate. Or I think.

 

When you lower FPS in PoE, you indirectly slow down the rate at which the bot runs, because the bot runs based on hooking the current game frame to work. If you run at 20 FPS, the bot itself could not run any faster than that (20 ticks a second).

However, in this game, unlike most others, a lot of logic has to be performed over multiple game frames. That is, we cannot call a client function to move, use a skill, open the inventory, etc... We have to use emulated input actions to move the cursor and press keys. Performing these actions requires a frame each. When you have enough actions that need to be performed, you start losing a lot of frames, so the time between real logic execution grows further and further as your frame time is now 100s of MS long (as opposed to say 10).

As a result, you get the stutter movement and overall slow bot performance, because the bot is running extremely slow. The only way around this is to increase FPS so the bot has more frames to execute across, and ensure you have enough CPU for handling pathfinding and raycasting. Movement is more affected by this because of the style of movement being used to avoid stutter and various desync issues in this game. MoveTowards calculates a new path and only moves towards the first hop when trying to get to the destination to have something that works as smoothly as possible under normal gaming conditions.

For low FPS setups, this movement style is not ideal, because of the time between frames. Instead, the better implementation would be to calculate the path once, then find a point to move towards along that line. That would get rid of most of the slowness in movement, but creates a new set of issues when it comes to desync and the general design of this client when the game is updating as slow as it is when you run lower FPS. That is, if the game is running really slow, there's no physical way to send enough movement commands to the client fast enough to maintain smooth or steady movement. if the client's pathfinding was perfect, and desync didn't exist, it would be a non-issue, but that's not the case.

That's why post 1.2, the bot/APi will always have more issues with lower FPS than before. For now, you'll just have to make sure FPS isn't set too low.

 

Well regardless if what you're saying is true, this doesn't really make it ok for me. I thought re-writing the bot was there to make existing issues disappear and non issues be the same. But with this, it looks like you're creating issues, that weren't there before, and not solving a lot of the new ones.

However "beta" this bot may be, I was way more satisfied with the old release version and didn't have many complaints. And I'm not talking about plugins or extra combat routines here, but basic bot functionality. I use a lot of PoE bots, not just yours, and for whatever reasons, justified or not, this version of buddy is way below the standards of the old release versions, especially when it comes to functionality.

What I mean is, regardless of whatever imrrovements you plan on having, if this version won't at least be just as good as the old release, then the whole re-write was pretty much pointless, or maybe your point of view differs from mine when it comes to developing a bot.

My criteria are: Functionality, stability, efficiency, and non-detectability, in that order, right now we're left with which of these? Limited functionality?

I've had patience with the entire project, but it's running thin, because I at least expected to have what was there in the old release versions of the bot, right now it's not even that.

I am by no means a whiner, but the state this bot is in, needs A LOT of work before it even comes close to let's say "market standards".

I hope it gets there, but time is also a factor here, get someone to help out with the project, I feel that one man is not enough for all that needs to be done here.

My 2 cents =), gl with the bot

 

Erhem. Here we go.
It's ok for me, it seems ok for other people. Let's just say we're mostly happy ( by the amount of whine when 1.2 came out, it's dimmed down quite a lot). If most people are happy > if one person is not happy.
A lot of issues have dissapeared.
Bot getting stuck on barrels and other shit, gone.
Bot dying a lot, gone.
Bot flagging by some stupid reason, gone(in my opinion).
This is from a sample size of 80 concurrent bots running a lot of hours over 10 machines, used to be 4 but I upped the FPS for better performance and less "bot-like" things.
Issues will always appeared. The bot as it was before 1.2 release worked well(kinda) for before 1.2 release, after 1.2 release, it did not work at all.
The bot is planned to be developed for future improvements. Last release was planed for last release.
Now unless GGG stops putting out releases, issues will not dissapear.
You state that it's creating more issues, please give examples, that weren't there before. Then not solving a lot of the new ones, this is kinda funny.
Solving a lot of new issues, then creating new issues. How does that work?

Using a lot of bots, there are only 3 bots that I know publicly, 1 is dead, the second requires VM ( takes more cpu, no point in FPS limiting), crashes more then it works. The previous version had 3-4 months work on it? when re-writing something, give it 3-4 months then compare, not one month latter.

This version is missing, interacting with objects for quests like mud flats, and that's about it. For functionality, it goes to plugins and CR, I really think you are mis-interperting what the bot has, basic functionality is back in.. like even better then release.. It has autologin, release didn't have that. You can easily move around skills/etc, it auto puts mouse move, it casts ALL auras and shuffles it around. I don't know what more you need. It can do maps now too, I think. Oh, it has chat functionality.

Push told me to try this

Code:

using Loki.Utilities;
using log4net;
using Loki.Game;
using Loki.Bot;

public class RuntimeCode
{
	private static readonly ILog Log = Logger.GetLoggerInstanceForType();

	public void Execute()
	{
		using(LokiPoe.AcquireFrame())
		{
			LokiPoe.Type("/age");


			foreach(var msg in LokiPoe.InGameState.ChatMessages)
			{
				Log.DebugFormat("{0}", msg);
			}
		}
	}
}

Put that into the dev tab and run it. read the output.
Shit like that makes QOL much much better.

I don't know bout you, but being detected is top priority on my list, and I'm pretty sure everyone else's. What's the point of everything if you are just getting detected? Try making and managing 40 new bots, shit's hard yo.

Bot currently exceeds release in "functionality"
It's market standards, you are comparing it to other bots, this bot currently IS the market standard.

Never, ever rush projects, NEVER. Shit goes wrong yo, and you get a lot of very angry people.

 

i got to agree with you on that darkbluefirefly, the bot work very well right now and all it miss for me is corrupted area and calling a portal to town after run, all the rest work as well as release if not better (didnt get the bot stuck once since i test it). and if you start to compare with other bots... then THERE IS NO COMPARAISON TO DO, exilebuddy work flawlessly and others are always buggy and slow and crash all the time, i got 2000chicken after i left the bot alone when i watched over 50run, and it happened more than once with other bot, wich is unnacceptable from my point of view. this bot will NEVER DO SUCH THING. It can even run dominus!!! and its working perfectly.