ExilebuddyBeta Developers Testing/Feedback Thread

Hey dude, ok.
1 - it's working as intended, it's meant to keep you alive so it has to move, only way to avoid moving so much, UP your dps, 1-2 shot kills and you will see the ball rolling.
2- it's not a tank bot, write your own routine to do what you want, don't expect the generic routing to be universal and do everything, it won't.
3. You gotta re-read the boss farm. it's meant for act 1 only and write your own for act 2-3, it's actually pretty simple.. Like really, really simple.
But yea, thanks for the feed back, just keep asking questions, someone will help.

 

1 Feature request that is commonly used by legit farmers.

Requirements :
-1 DPS Account
-1 Full Magic Find account with culling

Basically they would have to group up. DPS Account would clear all the way to the boss and dropd the boss down below 10%, open a portal go to town. Magic Find account enters portal and culls it without even taking damage due to the 20 second protection. Basically it is just the normal boss farming route but instead of killing it you would stop below 10% and open a portal, send a defined signal to the other ExileBuddy process so it takes the portal and kill the boss.

Feature Request number 2:

Use Leap Slam/Lightning Warp as movement skill ignoring the rest of the monsters(unless it is something that we want to kill) during those runs.

 

This is pretty much the issue with all our generic logic and this game; it will never work in all cases. Basically, you need area-specific logic to avoid issues like these, or be ready to do a ton of calculations to make sure you can actually do things like kite or avoid surround. For now, those issues are just going to be issues in the stuff we provide because there's no way to handle it properly for all scenarios and builds. The idea though is for people to have custom tailored CRs that take advantage of the information we provide in the API to work leaps and bounds better than what we provide (since we can't spend time specializing stuff like that).

The auto-flask plugin was provided for users to change it if they need to, but flask logic like that is something you'd want to put in a custom CR. The bot does not know what type of damage you are taking, just that you are taking damage based on health changes. This is due to the way the game works, and there are no known ways that I can think of to determine actual damage type breakdowns without hooking packets/client (which we can't). As a result, you would need to code logic into your custom CR that says, if you see fire mobs coming at you, and they are using their explode attack, then use this flask to help avoid the problem. As per the previous issue, this is really specialized logic for this game, that users have to cater to as our stuff is generic.

I'll need a full log please. It looks like some memory got corrupted, which can happen, but that error in itself can't be used to figure out what happened.

So thanks for the feedback. Our model (Buddybots) for development is to focus mainly on the API for the game, and providing the bot setup for people to be able to do just about anything they want. The stuff we provide will always be pretty generic, and not even fully feature complete, because our focus is being able to provide a botting solution that does work out of the box, but has the possibility for users to do amazing stuff we otherwise wouldn't have the time for.

Poe/EB is a bit of an exception compared to the other bots/games in the sense that, this bot is always changing, the API is always having to get reworked to keep up with the client, and the fundamental game changes in ways that breaks a lot of logic, whereas you don't see these things from other games but maybe once a year or so for large content updates. That's the big reason why this bot is still undergoing rewrites and doesn't have a dev community built up. There's nothing we can really do about that though, since we're dependent on the game itself, so if the game keeps changing, we have to keep changing and try to keep up.

 

This is 100% something the community would have to develop, just so you're not waiting for us to do it.

You can actually implement this by changing the BossFarmExplorePoi code. At the bottom of Execute, the movement code is:

Code:

            var res = await Coroutines.ClickToMoveTowards(position);
            if (res != Coroutines.ClickToMoveTowardsError.None)
            {
                Log.DebugFormat("[Execute] ClickToMoveTowards returned {0} for {1}.", res, position);
                _exploreLocation = null;
            }

To generate a path to the destination, the code looks like:

Code:

                _path = new PathfindingCommand(LokiPoe.Me.Position, position, 8);
                if (!ExilePather.FindPath(ref _path))
                {
                    // Handle path generation failed
                }

Now your path is in _path.Path, so you can create logic to figure out the next best place to cast the Leap Slam skill or Lightning Warp.

As long as you handle when the skill fails to cast, and just do normal movement instead, it should work out, but you'd have to give it a try.

To change which combat targets get included, you just have to modify the CombatOnInclusionCalcuation in BossFarmBot.cs.

 

hi, got an problem when I work on my dominus plugin, the boss "Imperator Stantinus Bitterblade" cannot get by GetObjectByName() function.

 

The name you are using is not correct then. What you should do is run the code:

Code:

foreach(var monster in LokiPoe.ObjectManager.GetObjectsByType<Monster>())
{
	if(monster.Rarity == Rarity.Unique)
		Log.DebugFormat("{0}", monster.Name);
}

And you'll be able to get the exact name. You can do it in the Dev tab or just throw it into a routine or plugin or something.

 

Sorry if I sound rude saying this but its not my intention, just giving my opinion. At this moment this bot is all about "community needs to develop it" it has been release so long ago yet it can't still identify an item so it decides if we want to keep it or we want to stash it. Meanwhile it already has been detected/currency flaged in two different occasions.

I really like Buddy team but I think they are working on a wrong goal on this bot. You wont have a large community like DemonBuddy or HonorBuddy doing all the homework for you to sell to us. On other bots the best performing/efficient/feature enabler functions were made by Community such as Trinity, Shuffle, RaidBot, etc.

At the moment ExileBuddy is an year 2014 potential bot with an year 1999 ability.

- Bot xp/h is 4 to 5 times slower than a real player farming an area.
- Mostly because it traces back a lot and it picks bad targets to fight with, i.e., lonely monsters
- Curse logic is based on monster rarity when there should be a "group" option aswell in order to curse a lot of monsters at once
- Bot flicker items a lot, it is probably one of the reasons why people get flagged.
- BETA is a testing stage to test how the features work together on a large group of individuals with different setups, not an unit testing to keep adding features every release and add more API. Why dont you work with the current API creating MUSTHAVE plugins/botbases/customclasses before complex API is added?
- What is the point of having an awesome API if an AutoIt bot farms faster, more efficiently, filters loot?

ExileBuddy has 0 reference to BETA on the Buddy website yet it shows :

- Efficient Town runs and Stash
What is an Efficient town run? Stash? Efficient? Dropping all the loot all mixed up? without even sorting/identifying/filtering?

- Inteligent Loot Filters
How is it intelligent if you don't identify it? You just loot it all good or not. Before we had ItemScore on DemonBuddy there were base filtering and it worked like a charm before a more advanced filter was applied.

- Easy to setup
This one is funny. It is easy to setup but if you want the bot to perform as good as an autoit bot you have to develop a plugin or something to make use of it.

I dont want to have a BMW on the garage, I prefer to have an Opel on the road.

If you were developing something that it is not marketed yet, for personal use, for future market, as a coder myself I understand the way you are following. You want to have all the foundations before building
As a costumer it is just not acceptable that this bot doesn't have must have features which if I've ever build a bot to market it would be a must to have them.

Everytime that there is a suggestion to improve the bot you are selling the answer is, it can be made by community.

It has been 1 year since closed beta on ExileBuddy and we still dont have a reasonable combat logic, we still don't have item filtering, we don't have map farming.

As far the combat logic goes, wether you have no idea how to play efficiently and kill monster fast or you just don't invest any time doing it. I have tons of ideas but everytime I talk about one I'm just redirected to the community develop talk.
Why does the bot stop to kill every 1/2 monsters? As a player I kill while I run, agroing those 1/2 monsters creating a big AoE pack then killing it. We may think, maybe if our character is too weak we are not able to tank those hits that make is gather the monster. You are right, thats why we add a threshold, we tank the first hits, if we get damaged too much, we will be more careful on the next pulls, if we barely get any damage, we can play more agressive and facetanking more stuff in order to clear faster.

I can even go depeer and having each monstertype also has a danger threshold, same goes for mods, if we encounter a zombie we're not gonna play as careful as if we encounter Lightning Thorns Arc caster. Having something like getDangerThreshhold(monster) function, where we would see the basetype of monster, its damage type against our resistances, etc etc. THEN we COMMUNITY would tweak those values based on experiences, thats community feedback and work to improve the bot.

As far as Item Filtering goes, we would have multiple presets and then we would calculate the itemscore for every preset we had and we would return the scores for the highest presets instead of just having 1 innacurate preset that would sum all stats. This idea comes to my mind based on a Football Manager tool, called FMScout. It would scout the players and give them ratings from 0 to 100%, obviously it would do it in a smart way. It would do it for each position so for example Cristiano Ronaldo is 96% as Central Forward, 95% as Winger, 60% as Left Back. whenever I wanted to see Cristiano Ronaldo score I would be able to see all his positions and also which position is he better.

This is just a thing I'm taking out of my chest because Path of Exile is a game that I enjoy a lot, and I also enjoy Buddy a lot and it disappoints me seeing a Buddy named product with such low standards.

 

It does identify items. Isn't that working for you?

 

Indeed, it identifies and you can filter also . . .

But i agree with a lot of things that tozedao said (not all thou) . . .

Pushed thou has been doing a great work and is always available to help either in the forum or Skype, maybe Buddy Team should put an extra dev with pushed to speed things up a little . . .

 

This is my opinion too. I totally agree with that.

 

well, i got the name from dev tab by this code...

Code:

foreach (var @object in LokiPoe.ObjectManager.Objects)
            {
                Log.DebugFormat("[{1}] {0}", @object.Name, @object.Id);
            }

 

I've noticed that the beta doesn't interact with the strange glyph wall anymore in mud flats which it used to do in release, I'm not a programmer so what do I need to add to get it to do that?

It also seems like it didnt care about getting all 3 glyphs before taking the transition area and it only got 2, then it was just trying to enter the passage without activating the wall, I've tried with me getting the 3rd glyph and same result.

I've tried something rudimentary like adding it to the chest list hoping it would interact with it but it doesn't work.

 

it's a known bug, will be fixed soon. Has to do with PoIs.

 

I'll double check, but you might be seeing a very specific game behavior.

If the boss is alive and running around, it should return his object, without any issues. If you see him moving around and stuff, and the API isn't returning the object, then that is a bug, but I don't see how that is possible, as that's how you got his name.

If you kill him, it's possible his corpse is exploded, and as a result, he no longer exists. Or, if you kill him in a way that makes his corpse explode, the same is true. The game deletes the object, so you have to code around that by keeping track of when you see him, and when you no longer see him, trying to distinguish him being out of range, or he actually no longer exists.

Can you check to see if that is the issue you are experiencing?

 

Beta has no quest logic at all. The actual quest API was going to be worked into the bot, in place of what Release had, but some things came up that stopped all progress on quests/maps in the beta. What's happening now is another set of changes to Beta in preparation of the upcoming 1.2 patch.

 

Really? After all I've said its all you have to say? What is the point of identifying it and just dropping it randomly on stash?

It is so easy for GGG to detect this bot because of the non-human behavior, no one drops loot randomly on stash like bot does, no one drops every single rare+ item on stash, no one is level 75+ and the only thing it does is Library/Merveil/Docks. It is a group of specific behaviors that all together flag us as botters.

You are expecting community to develop core features of a bot.

Community should develop things like:
-CC's for specific builds
-vendor specific recipes like chaos,chance, glassblowers bauble,
-Craft/Roll maps,
-Auto equip gear
-Auto craft sockets so when you upgrade your gear you can use all the gems again
-ItemScore
-Plugin that makes multibot viable as party
-A plugin that "burns" a specific currency until you want the specific mod.

These are Examples of community develop towards the bot.
You are developing a bot for a community that is 5% of Diablo 3 community, yet you are expecting more from the community than DemonBuddy has ever gotten.

Where is the map support? That should be a feature that should be on a bot that enters the market if >80% of PoE players are actually doing that.
Where is the questing support?This game has 26 quests. Twenty Six, Forsaken Masters is coming in 12 days and will introduce random missions, we have 26 static quests ingame and after 1 year the bot doesn't fully support them.
Where is the passive skill tree point alocation? We can't afk or we will just get a bunch of passive points to allocate making us looking botish.

I stay without using the bot for a while, 2 months I use it again and I still face the same bad performance of it.

Walks around to areas that have no monsters making the xp/h look like shit.
If I use it on a magic find character, 2 hours later stops botting due to being full.
Poorly decides when to fight and when to group a bunch of monsters, fights every lonely monster that there is.

In terms of performance an amazing bot plays better than a human, a good bot plays as good as a good player, and an average bot plays as good as an average player, this bot plays worse and slowly than a bad player, I can choose a friend of mine who has never played PoE before and he will clear areas faster than this bot if I explain him how to do it for 1 minute.

pushedx is an amazing coder, and he really knows what to do in terms of PoE.
I know him since 0x33 and Silkroad times.

But I don't know who is taking decisions about what to develop next and where to spend resources, whoever that guy is, he has no idea what this game is about. I suggest you play it because it is not as easy as D3, you have a lot of knowledge to acquire in terms of farming/wealth generating/combat mechanics.

 

There's two classifications of work that goes on with a bot project like ours:
The first is work that only, we, the devs, can do. This is stuff relating to the actual API for the game, the bot framework, and the core GUI setup.
The second, is work that anyone who can code can do. This is stuff like changing around the logic in the bots/crs/plugins we provide, or adding new more comprehensive features using the API.

My job here, as this stuff really is my job, and not just a figure of speech, is to focus primarily on the first type of work. That is, if I spent all my time doing the second type of work, the API would never see improvements, new client features would not be available, and in general, development as a whole for everyone else would be limited by whatever setup we currently had at the time.

That is not a good thing, and you can see what the project would be like if it were actually run that way by looking at Release. Development with Release is near impossible for people because of how it was setup. Since everyone else here (talking about EB community) is not a buddy dev, no one could actually fix those issues, so you'd be stuck with the design forever. If that were the case, the project would pretty much be dead.

Now, since I don't want that to happen, I choose to spend time fixing various issues with the API, adding new client feature support, and trying to get things setup in a way that promotes bot development for this game, in the interest of keeping the project going, and setting up the opportunity for other devs to come in and do stuff with EB/PoE that hasn't been done yet. The past several months, starting with the announcements of the new Beta, was done specifically for this reason. If you read though this entire thread, look though the patch notes and changes, and check out the new guides, I think anyone should be able to see that, pretty easily I might add. Beta is leaps and bounds in a better position than Release is for development and allowing people to do the things they want to do in this game, and us not getting in the way.

The API we provide is the single most important aspect of any product Buddy offers. Without it, most of what you see, and what people compare everything to, would not have been possible. That is not to say the bot we provide isn't equally important, because the API doesn't sell the bot to the masses on its own. However, the bot is built up from the API, so the two go hand in hand. I make the choice whether to focus on API specific stuff or logic specific stuff, based on the current state of game, and the general feedback of the forums.

The past few months have been amazing in terms of production and progress, because for the first time ever, we didn't have to deal with weekly patches that change a bunch of stuff. What most people seem to forget, and ignore in their comments about the state of EB, is just how this game works, and has changed, ever since OBT. If this game would have had monthly balance patches, with the typical bug fixes, and few client improvements time to time, like most games, then I think a lot of the frustration towards the project would be more than well justified, as there would be no reason for the bot to be in the current state it is in.

That's simply not the case though. Since users aren't devs, you guys don't get to see just how much we have to do behind the scenes in a game like this, compared to other games that follow a lot more traditional development pattern. It's not your fault though, but when making arguments against the bot, if you ignore the fact how much this game has actually changed over the past year, then well, there's nothing we can really say, because your missing the whole reason why things aren't the way you want them to be.

So, everyone is free to voice their dissatisfaction or frustrations with the bot. I agree with a few of the points, but that doesn't change the way things are. Path of Exile is an ever changing game. It requires constant work to keep the API up to date and support the latest features. GGG hates bots and RMT, they don't make things easy for us. Development in this game is not fast and easy either. It takes a lot of time and effort, and as a result, the amount of time and work put into the bot will never been truly seen. The stuff we provide for the game for users will always be pretty basic. Unless GGG decides to never update the game again, and the client never changes, there's no reason for me specifically to focus on that type of work, when, if the project was setup the way it needs to be to exist in the PoE environment, other people could do that.

Buddy has a large community, and a lot of resources at their disposal. When EB can reach a point where there is little of type #1 work to be done, then type #2 can easily take place. We have a lot of options to help motivate people to get into dev and contribute to the community. However, before that can happen, the project needs to be setup so it's possible. That's why I have been working long and hard ever since taking over EB dev the way I have, in order to position the project in the best way possible. What we had before with EB turned out to not work well with the game. Tough luck for us, that's the nature of bot development, so we're doing everything possible to get it back on track and where it needs to be.

The new beta will have its own new thread soon that goes over specifically what has changed. It will start out like this beta did at first, and come only with the core API for people to test using some pre-made scripts to make sure everything works on other PCs. From there, the regular bot logic will be added in. However, since the big 1.2 update is coming, and we won't know all the actual changes taking place, the most important thing right now is continuing API updates and adjusting the bot/routine/plugin setup of EB so once the API is as good as it can be for the current client version, we can start focusing on improvements and features to core stuff.

The original plan for Beta was to add quest/map logic before the 1.2 update. I had pretty much 3 weeks after the simple BossFarmBot stuff was done. However, plans changed due to unforeseen (but fortunate) events, and the new beta API is top priority right now. So, I am sorry to keep people waiting in terms of not having automated map farming support still, but looking at the big picture, and the past history of the project, I believe the current course of action is for the best, both short and long term.

 

All I have to say after that is that if Buddyteam doesn't have anyone doing the second type of work, and they can't expect community to do it, all your first work just looks ungrateful.

ExileBuddy needs to be valuble for both costumer only and costumer/developer. I wouldn't mind paying an extra 25€ if I had someone from the BuddyTeam doing the second type of work, because that is what allows costumers to choose this bot VS others.

Beside the "Buddy" there is still no reason to choose this bot over the others.

Imagine someone who has no idea what Buddy is, he just plays Path of Exile and he wants to buy a bot, he looks on the market, he looks on the current ability of the bot and the bans and he won't pick ExileBuddy unfortunely.

 

Well, I've seen what ExileBuddy can do, and it was pretty good and my income was better then I ever expected. And I haven't even tried beta yet...

So when someone looks at the current ability of the bot and decides to leave, it's his fault, not the bot's in my opinion.

 

There really should be someone else doing the stuff that pushed doesn't have time to do, at least assign someone for 2-3 months that can add in the quest/map logic while you still work on what's left on your side of things, don't think it's too much to ask from the buddy team and it would go a long way, regardless I have faith in the direction push is taking things, even if it's taking longer but at the same time it would please a lot of people who have payed money for the bot and probably get even more that want to buy it on board