Automated combat

Ashmer

Ezalor said:

Again with the "I just can't do it" attitude. I have never, until about 7 months ago, touched coding at all. I have absolutely no coding experience whatsoever prior to attempting Aetolian PK. I made many of these same posts from around that time, complaining that PK is too hard to get into, coding is too cryptic, etc. But the moment I stopped whining to myself and actually sat down to try it all fell together very quickly. Less than a month of time working on it before I could PK decently. If I can do it anyone can. I don't see how you can expect to succeed at -anything-, not just coding, if you give up before you even try or the moment you hit some adversity.

As someone who just built a system from literally scratch with no coding knowledge beyond basic computer skills, this.

 

Yes, it's been hours, and yes, I had to endure Angwe saying "Just let it happen" when he caught me out alone at a Lesser with only firstaid to protect my virtue, and yes, I still die horribly when I have write_ropes instead of writhe_ropes in my system, or demulcent is spelled wrong. But damn if it isn't satisfying having built something like that from nothing, and have it work to the degree I'm willing to have it work.

Also, after the first two/three weeks, you get a helluvalot faster. It's more like a hobby than a videogame, but I think it's a special kind of person that would play a MUD anyway, in this age of Xbox 4s and PS Ultimates.

 

Anyway. Ezalor still wtfpwns me, so I'm going to go back to coding in between bad ass RP.

Medri

Ashmer said:

As someone who just built a system from literally scratch with no coding knowledge beyond basic computer skills, this.

So how do you build a system from scratch with no coding knowledge?

Neithan

Medri said:

Ashmer said:

As someone who just built a system from literally scratch with no coding knowledge beyond basic computer skills, this.

So how do you build a system from scratch with no coding knowledge?

You put forth some effort, and learn how to do things. You don't need pre-existing knowledge to do it. You need to apply yourself, be patient and learn.

Ages

@medri A lot of wt's and the help of citymates.

Medri

Ages said:

@medri A lot of wt's and the help of citymates.

The last time I ever asked someone to help me build a system, they linked me to Mudlet's LUA manual and said 'go read this, then come back to me when you can code.'

I learn absolutely nothing from being thrown at a textbook. Right now, the most I can do is make an autobasher and autosipper and track rebounding/shield, but the task of putting together an entire system is beyond me.

Ashmer

Yeah, Google was my friend. It's hard to start teaching someone from scratch. What I did was pull from what I found on the mudlet wiki, though I've never used the mudlet manual. I took one look at that thing and went "you expect me to read what?"

 

One thing I also did was take some code someone else had started, and take it apart. I refer back to Xarisystem to see how he did things sometimes, and after multiple times reading the same function and testing things, the "what the unicorns does this do anyway?" started to dawn on me. Hell, I still had only cursory knowledge of the pairs() function in Lua when I used it to build my entire curing system, and I just tonight figured out how it actually works! Kind of!

 

So yeah, Google is good, and then when you have a bit of code started, you can ask someone the question called, "hey, what's this error?" (I'll get to that in a sec) or "why isn't this working?" or even "anything you can see about this that's off?"

 

The last thing is that, as you code, if you keep the Error window open in Mudlet, it'll tell you what you're doing wrong when you're literally just wording something incorrectly.

 

Also, the one only thing that ever produces any kind of result is action, or intentional non-action. So, if you literally just keep taking the actions consistent with having a system, you simply will have it. That's how I did mine, and suddenly I can code semi-proficiently in Lua. I even was able to code something offline on my lunch break at work without breaking it the first time, this week! (I'm sure I just haven't found where it breaks yet).

 

Oh, also:

HAVE FUN! 

Arbre

I had Atrapoema's system initially, and I kept having to dig in it to change stuff, since it was only set up for lycan and I was vampire.  After a long time of tweaking this and tweaking that, it became a cluster---- and I just said eff it and sat down and did my own.  The KLL clan is out there and a lot of help, and asking -specific- questions gets you more answers.

Daskalos

I'll admit, I miss the old days of manualling  before all this automation took over (for offense). It's a different skillset to be able to compete in the automation vs the manual natural instinct world, and no human is faster than a computer. I hated writing my offense automated. I did it, because I had to compete, but if tomorrow they took away discernment and a lot of these other things that make automation possible, it would take an adjustment back to the old way of doing things but I'd do it in a heart beat, even though it would be much, much  harder.

Setne

Medri said:

Mudlet's LUA

Lua is not an acronym.

Ashmer said:

I found on the mudlet wiki, though I've never used the mudlet manual.

The wiki is the manual, everything(should be) was transferred over from the manual to the wiki, though the manual is still available, not sure how up to date it is though.

In regards to the whole coding discussion, learning to code(in Lua at least), isn't that difficult. I learned the language with 0 previous experience years ago, while also in my final year of high school. While I've never coded a full curing system, I've coded a number of other things that work pretty well after a lot of testing and fixing. Learning how to code just takes patience, having good references(be they manuals or people) and just sticking with it. There is the mudlet irc channel, available through mudlet even alongside various irc clients, where a number of people are there to assist with any questions, though make sure to read the topic when you join. Find other Aetolians willing to take up a mentor role if need be.

In regards to the automated combat discussion, I'll save that for another post I hope to make tomorrow.

Medri

Setne said:

Medri said:

Mudlet's LUA

Lua is not an acronym.

See how much I know about coding?  %-(

Ashmer

@medri: C'mon, it's a common mistake. I didn't know it wasn't an acronym. The whole point Ezalor and I are trying to make is that you can do it. All that's required is a willingness to type lines of code, Google searches, and questions instead of "I can't do it," and I mean that in the most literal way possible.

Kikon

For those of you wondering how you jump straight into building a system with zero experience, the simplest way is to really get your hands on someone's system and start trying to understand what's going on. Grab yourself a copy from somewhere, a manual for the language you're trying to read, and start reverse engineering it and understanding why it works the way it does. The next thing you can do is start taking baby steps by modifying what's under the hood to make sure you know what's going on.

The first script I ever created in Aetolia was one that would stop me channeling CRYSTALHOME if someone tried to follow me while I was performing it. Think up a similar simple goal and try to create it, eventually pushing yourself into the more harder stuff like a full blown system. I personally never had a fully automated system for at least two years when I began playing.

On this line though, just to let people know if they didn't already, I recently released the framework behind my current system. It's a great resource if you need somewhere to begin understanding how a simple system works. It pretty much replicates FirstAid, so it isn't too crazy (yet!) for a beginner. You can find the thread here.

Lastly, if you have any questions about Lua, or having troubles with your code, I'm generally happy to have a look if you want to throw it on pastebin and shoot me a tell. I'm not much of a ZScript expert though.

Toz

Hi. I'm a 21 year old CSCI student, a year or so from graduating (God I hope...). I got my start by doing all of that - at 15 years old, I was tinkering with code for playing IRE games. I had no idea what I wanted to do as a career, it was just to play the game, but I got hooked on it. I started by buying a system and trying to add on little attachments. Come up with a small project, ask friends, google, rtfm, whatever it takes - within a month or two, depending on how dedicated you are? You'll find yourself answering questions more often than asking them, and you'll be taking on projects for other people sometimes. Been doing it for years now and though I'm certainly not an expert, I can hold my own when it comes to figuring out how to make stuff work. All because, 5-6 years ago, I sat down and wondered how to auto-sip on Nexus.

Kaeus

I'm 3 years into my career now all because as a kid I got hooked on programming via this game. I don't necessarily program for a job, but it led me into a lot of other options and I still use programming on the job quite a bit. The mindset behind programming can be applied to quite a bit of technology out there and does wonders for being able to understand how things work!

Ashmer

Kaeus said:

I'm 3 years into my career now all because as a kid I got hooked on programming via this game. I don't necessarily program for a job, but it led me into a lot of other options and I still use programming on the job quite a bit. The mindset behind programming can be applied to quite a bit of technology out there and does wonders for being able to understand how things work!

This. This, this, and doublethis. The way I work at work now is a system, it's crazy. IF THIS AND THIS OR (THIS AND THIS) THEN THAT, ELSEIF THIS THEN THAT SO THERE END.

Instant success in administrative endeavors.

Alexina

I'm not sure how the discussion moved from 'why do you have to spend RL months learning scripting/coding to participate in PK?' to 'yes you can!', but, at least for me personally, I want to play the game. I have tried learning to code and I've picked up some basic stuff, but I honestly don't feel like doing something I don't even like just to be able to play the game. Scripting has literally become (or rather, it's always been) a pre-requisite for PK'ing, which is a pretty big aspect of all IRE games.

A new player joining Aetolia (even assuming one familiar to muds) must overcome two very steep hurdles; gain enough credits to tri-trans, plus vision (discernment), and survival (focus), and learn to Lua. Together, they represent huge investments in time, energy, and possibly money that most games we're competing with do not require from their players. IRE's model is obviously successful enough to run as a company, but over the past few years it seems more like the focus has been monthly, shiny, new promotions than keeping the games up to date with mechanics available in other games (MMOs, particularly). Like Moirean pointed out earlier, most games have streamlined their experience to have less chores and more fun. It's generally not a bad approach to take if you want to retain your playerbase.

Kikon

I would personally like to see FirstAid improved to an almost practical replacement and leave customization or fine tuning to system builders who want that minor edge.

Seir

Hey, remember awhile back when some of us were asking for server-side curing? It would pretty much fix all of the issues stated in this thread.

If PK becomes a contest of who can code better, you're doing it wrong. I sincerely hope that balance discussions don't evolve into: "AMG. LUA OP. BUFF C++." 

Quite frankly, while even I learned how to "break systems", it was mostly on how to break or go around their pre-restoration. I never thought about actually breaking someone's system in order to win. I also don't even see the fun of allowing your system to completely automate your offense. This, in my eyes, completely defeats the point of fighting in the first place. PvP is a contest of skill, and maybe coding skill should be a minor aspect of that, I won't disagree with that point, but automating your offense for the sake of "winning" completely defeats the spirit of what PvP is supposed to be. Is winning that important that you don't even bother to participate yourself any longer? Food for thought.

Moirean

Yeah, I have no idea why the discussion is centering around "Lua is easy to learn!" Sure, it isn't that hard. I built some lua add-ons for WoW, just for fun, but if I had to do that kind of add-on building over and over before I could do arenas and battlegrounds, I would have dropped WoW ages ago. That is what Aetolian PK is, though - "easy" or not (and I don't really think something that takes MONTHS of work really can qualify as an "easy" prerequisite), there are barriers in the way of simply playing the game.

Again, that's Aetolia's (and IRE's) decision to have those there. A reliance on coding has long been something you find in IRE and it is something that's grown over time to become a core feature of the games. I suppose if somewhere along the way they didn't like that aspect, they'd have put the brakes on it.

Oleis

I'm not going to get too deeply involved in this conversation, because it's not entirely my place to set the tone for our combat direction. I do need to say that server-side curing is not an answer to claims and complaints in this thread. It may indeed be true that lots of players want server-side curing, but the real issue at hand is the requirement for heavily-automated offenses. It is my personal and completely not official opinion that you're right. The need for a completely automated offense (or the fact that it's even possible) is a dangerous element to the fun of a game like Aetolia. While some level of text-based automation is unavoidable in a text-based game, the deciding factor in a fight should be something intrinsic to the people involved in it. While there will always be outliers, that's the goal. How that translates into Aetolia's combat balance is a question for much time and contemplation.

@Moirean's posts in this thread allow me to make a really important distinction. As I said, some scripting is going to be inevitable, because it's -possible-. Tools like gags, highlights, and basic aliases make your interaction with the game world so much easier. But there will always be players determined to win; with the evolution of MUD technology (as old as it is) to a level of very fast, very accessible automation, we as MUD administrators have to choose how we'll deal with it. At one time, Achaea fought the tide. Flashback to 2004, when randomized asterisks in paralysis messages were an effort to make automated curing more difficult:

"It shouldn't really be any much harder unless you're lazy and relying on a full automated curing system, which isn't very smart. Just don't have a Curare trigger and if you see a bumload of asterisks then hit a little button that gets bloodroot out of the Rift and eats it." (link)

Within a couple years, those asterisks were irrelevant, because technology had advanced to the point that that kind of pattern matching was trivial. As Aetolia's combat has similarly grown and the demands of those advanced coders rose to the forefront of the conversation, our game mechanics swung to the opposite extreme: aside from the baseline difficulty of the scripting itself, functionally creating an automated PKbot has never been easier or more desirable from the perspective of the player who wants to win.

Personally, I believe that there's a sweet spot. We're trying to make defensive and utility coding as simple and streamlined as possible. What used to be a complicated, message-heavy system of writhes is now a couple of triggers. Combat Messages (as they proliferate) and afflictions in GMCP allow for the filtering of irrelevant information and highlighting of the most important bits in a way that is accessible to everyone. That's an element of scripting that I do want to encourage, but in a way that makes it dead simple on the entry-level. It would change the very nature of Aetolia to remove all of the scripting from defense and healing, and I will freely admit that I don't want to see all of it go.

Offensive scripting is not nearly as ensconced in Aetolia's nature as defensive scripting. Frankly, it's a system I'd love to shake up.

Moirean

I make jokes sometimes that IRE games should advertise themselves on programming forums and sites like dice.com. There really is a fun element that can be found in coding for MUDs - you can create very discrete, tangible coding projects for a MUD. "I want to build a tracker for esteem." 3 hours later, bam, you've made a functional one, and that's fun and satisfying. I really do enjoy that facet of playing MUDs, despite what my posts have said in this thread, and I'd hope that I'm considered among one of the 'coders' even though I do it more for fun. I opted to get Citadel instead of building my own system, because I want to spend the bulk of my game time doing other stuff like RP and org leadership, but that doesn't mean I don't know how to use my client of choice (and I've tried to share my projects via the CMUD script thread; anyone who peeks at that can even see that I've visibly improved over time!).

Regarding the technology improvements, I think that is something pretty exciting. I've been browsing through old (and new) posts and articles about MUDs a lot lately and there are some pretty cool things and ideas currently out there. Handheld technology has improved to the point where you can functionally play a MUD through something like a phone or a tablet and MXP and HTML5 have made it a lot easier for websites/apps/clients to present really useful and interactive stuff like click functionality, sound triggers, editable content, local storage. I am fairly intrigued to see the sort of changes this means for MUDs - there's a potentially much bigger base of users, many of them casual, if the market can be tapped in the right way. But this means, by necessity of the things handhelds/tablets can easily do, that stuff needs to be more simple and intuitive. I think there's a ton of room for growth here - you can't very easily play a MMO from your phone or tablet, but I have a new Carnifex raving about how he's down at the lakeside on a tablet via wifi happily bashing 3rock. I think if someone can find the right way to present it, this really could be a new golden age for MUDs. At the least, an influx of new players.

I'm not a dev, but I can totally imagine stuff to help this along. For example, I'd love to see an app alongside the web client that lets players create to-do lists or important notes they want to remember (with maybe even some sort of webclient option to just click/select to add useful IG hints to that list), as the sheer amount of stuff to recall is one thing new players chatter about a lot; integrating that sort of stuff right into the client would be like, bam, we got your back. Silly thematic stuff seems (seems! I'm not a dev so I dunno!) possible, like little audio toasts when you get a level. It seems like it might even be possible to easily and quickly set stuff like firstaid sip/moss levels by just pulling something along a slider from 1-100% or even something fancy like the ability to change their attack priorities through a list they can update. 

Anyways, I'm just rambling about nifty little stuff that would make MUDs less odd and more easy to intuitively interact with as a new user. The thing is, most of that stuff - including all the awesome stuff we've already had added, like the MXP orghelp options, the cool new web client, etc - just goes out the window when someone wants to dive into actual PK. The intro process keeps getting better and better, but then there's this wall of, "Yeah, that was fun. If you really want to play, though, you're going to have to sit and do something completely different." Maybe that's not even something that can be changed, and maybe this is simply part of what IRE MUDs are - certainly, removing the scripting game would dramatically change stuff for players, I think, and that might not be a good thing, unless a way was found to do it in a seamless way. Ie, not just introducing roadblocks like the old curare asterisk attempt, but stuff like tactical choices (I know, I sound like a broken record at this point) that reward high level analysis of the entire fight. Sure, that stuff MIGHT be able to still be coded, but it's a lot more complex than just "if enemy_rebounding=1, raze, else stab em" and that sort of decision-making is the fun type. Without stuff like that, though, it's simply smart and practical to use AI for your offense...but it's killing the overall PK scene.

Look back to how the introduction of firstaid made basic curing and team fighting something anyone could do. When that went in, we saw a lot more people getting involved in combat. Our mid-tier combat completely blossomed - since that point, however, many of those people have gone on to purchase systems (why would they not? They got a taste for combat and now they want to do it better). The unfortunate result of this however is that we have basically no more mid-tier. People who are newish to PK cure at the nearly same level that high-tier people do, but offenses for someone just learning how to fight are far less capable. Because so much of PK is based around afflicting, this means that a new Syssin or Templar is basically going up against a Kaeus- or Ezalor- or Daskalos- level curing every time they fight, while their offense may consist of fixed aliases to hit xentio and curare or slike and gecko. This creates a brick wall that has really killed off our mid-tier, it makes systems feel like a must-have for new players, and it makes learning how to code an AI/smart affliction offense essentially mandatory. 

I do think MUDs are in a really cool spot right now from the stuff I'd been reading about; I just worry that the heavy scripting we see so much of is a detriment to any new player recruitment and momentum that may be gained by tapping into the wider market opened up by tablets/handhelds and the growth of the casual gaming audience. There definitely IS a bigger market there for text games and people seem happy to play games that have little to no graphics - look at Fallen London - but that same type of new MUD player likely isn't going to want to invest effort in a tangential thing like learning lua or zscript. I know that I didn't at first, and I spent my first few years of MUD play using webclients/telnet, internal aliases and manually curing. That sort of gameplay simply isn't possible today, though.

Anyways, I've rambled and I can't really read all I've written because the post expand box is dead (so sorry if it jumped around!), so gonna stop now. 

Haven

Oleis said:

I'm not going to get too deeply involved in this conversation, because it's not entirely my place to set the tone for our combat direction. I do need to say that server-side curing is not an answer to claims and complaints in this thread. It may indeed be true that lots of players want server-side curing, but the real issue at hand is the requirement for heavily-automated offenses.

@Oleis: I completely disagree. One reason being that heavily-automated offenses are not a requirement to participate in combat or even necessary to compete in top-tier combat. It certainly helps having something pre-built to think and act for you because an AI/script can react much faster than what's humanly possible for a person and minimizes human error to maintain a consistent accuracy within the confines of its scripting parameters but it does NOT mean you're going to win 100% of the time against someone without an AI of their own. It does not even mean or guarantee that you'll win 50% of the time against a competent opponent without an AI of their own because there is a speed ceiling in the game that's still well within reach of what's humanly possible via balance and equilibrium for example. If anything it just means you're trading off real-time adaptability to things outside your current scripting parameters for an increased reaction speed that's consistently accurate. A bonus that's not as significant as some might think (not useless either because accuracy is still important. It's just moot in my opinion because with practice you can minimize your mistakes too and be effective in combat via manual). If that were the case then those with offensive AIs would be unbeatable against those without one and for opponents that did have an AI, combat would be ultimaltely limited to whoever struck first and or had the most luck attributed to them either by latency (lag) or the random number generator elements of the game's combat system (the stupidity affliction and what have you).

What IS required of a player, however, to even participate on a meaningful level beyond temporary meat shield is a curing system and if they want to reach the middle-top end of  the combat tiers then a tracking system of sorts too (particularly since combat scrolls by far too fast to humanly comprehend without echos or something).  That's not to say that meat shields are useless or cannot be dangerous (lol, dat cleave/skewer/web tattoo/BBT/etc), just that they're not optimal/ideal/practical if they cannot live long enough to distract the enemy so that those with functional systems can take advantage of the unhindered time.

Thankfully, firstaid has been implemented in the game to help lower the entry bar because you cannot even begin to play with the combat aspect of Aetolia and hope to achieve any sort of reasonable success without a system of some kind. Unfortunately, and I don't believe I'm alone in thinking this, firstaid is not sufficient enough of a system to fairly compete with other players in its current form... Well I suppose it is sufficient for group combat provided you aren't the primary focus. Anyway, it's certainly a start but it needs to be taken a step further.

[spoiler]

If you're still not convinced then consider group combat for a moment. Simple triggers, aliases, and combat awareness (which can be achieved with something as simple as an echo or as complicated as a tracking system) are really all that is needed offensively to participate and you can still be highly effective in skirmishes considering there are A LOT of skills that mesh together on both sides of the game that can yield lethal results. The players of the Carnifex Guild are a good current example. Their elder players take advantage of their combat knowledge and use it to help introduce their newer players to combat. There isn't a highly complicated AI offense system involved for their noobs to participate successfully as far as I'm aware. What about 1v1 combat? Well, I'd say you need only look at the successes of manual players like Illidan or Benedicto or Isto (or at least I think Isto still manuals Lycan) to see that AI offenses aren't necessary. They're just as successful as players like Daskalos, Kaeus, or Tyrak (at least I think he uses an AI).

[/spoiler]

TL;DR version: I disagree. Only a curing system is a requirement to enter combat successfully in Aetolia at this point. If you want to go beyond the basics and lower tiers of combat in favor of the top tiers then a tracking system is also an additional requirement as it'll help minimize your mistakes and capitalize on opportunities that present themselves with your strategy. Those are easy enough to make though with things like discernment and what have you.

A server-side curing system benefits everyone, including the Admin, because not only does it mean more people can participate if they choose to but you'll also start to see the glaring flaws consistently crop up with combat balance if there are any since more people will be focusing their energy on learning their class and what counters what etc. This means because we can consistently identify them and more frequently, we can focus on fixing them and making a better Aetolia.

Toz

You're painfully wrong, @Haven. Combat is a matter of centimeters, and anyone competent going against anyone competent with an AI is going to lose horribly. Players typo, players miss something in the spam and hit rebounding, but an AI? An AI is almost perfect. And since an AI isn't unique to anyone, if you want to get involved AT ALL for a serious degree of combat, you need an AI. There's no if ands or buts about it - if you don't have an AI, it's like trying to pk as neutral balance. And, as things continue and the AI developed gets more advanced, it's going to become a steadier and steadier handicap.

Haven

I think we have different definitions of manual combat and automated combat.

When I say manual, I don't mean players literally typing out "envenom left with curare;envenom right with aconite;dsl Toz" every time. I mean they've aliased those commands and popped them into a macro. For instance, I punch F1 on my keyboard and it'll do my standard luminary combo (adjusting the affs for what they don't have) but I must continuously press it whenever I have balance. If want to do a different tactic or combo I use the other F keys. As opposed to what I consider automated combat being something like: type kill Toz and it does everything until either I die or you die. 

Angwe

Missari manuals. 'Nough said. Lol Edit: stupid auto correct

Moirean

@Haven, adjusting affs for what they have is what (at least most of us, I think) are considering automated. 

I finally got on board with that about 2-3 months ago when I picked up Syssin. Before that, I was literally typing raze if they had rebounding and picking from a bunch of individual aliases (2x for xentio and vernalius, 2g for slike and gecko, 2a for aconite and slike, etc) and still have stuff like that for things like Indorani which I haven't rewritten for efficiency. An offense like that is simply not as good as one that picks the afflictions based on what a target has and automatically does things like raze rebounding instead of attack.

Toz

^

What @Moirean said. You have to do beyond 'basic' aff-tracking in order to be able to pk. Missari can manual all she wants, but with systems improving she's going to fall further and further behind until she's entirely irrelevant because of the way systems work. Nobody does accounting by hand any more. Nobody codes with punch-cards. And if you want to be worth anything 1v1 OR in a group, lately? You need to start building an AI.

Azton

Angwe said:

Missari manuals. 'Nough said. Lol Edit: stupid auto correct

I do both, depending if im on my phone or not... and yes I do lessers on my phone too...

 

It's all a matter of preference as I know ALOT of people freeze up so the auto side of things help them, but they is only so much you can do auto. Manually you can change things up ALOT more

Irruel

Each class is different. Lycan and Shaman are very much suitable for much more manual offenses, though good affliction tracking and the display of that is still essential. Luckily though, Oleis is committed to making that as simple as possible.

A few years ago, when dhuriv was first given to the sentinels and I was still maintaining my own system, I built an aff tracker. These days I use tripwire and the built in aff tracker, but I have built my own so I know what is involved.

I have spent far more time coding GUI displays of the information Kaeus's system hands to me on a platter than I ever spent writing my original aff tracker, all so I can semi-auto and (hopefully) still be competitive.

Angwe

I'm mostly auto, still. Initially, this was due to a lack of coding knowledge. Given my class, 'chasing' balance as I had with my Blademaster (Achaea) and Rogue (MKO, pre-venom) characters would have resulted in me going through a LOT of wasted venoms, spamming my opponent if they didn't gag wipe messages and possibly running myself out of venom mid-fight. No bueno.

Now, I think I could code around that using functions like cancan or w/e in TW to prevent me applying/wiping off balance. By the same token, I get near the same level of versatility I'd have if I manualed it (with an auto-envenomer) by just by using the queue function that comes with TW. I just queue the manual alias (I made manual aliases for all my attacks, along with combos and some pre-venomed combo aliases i.e. 'pastu' will wipe dhurive, apply curare and aconite and then slash/stab while doing pa1 will always apply curare first when using one of my manual aliases).

There are... a few reasons I let the AI push the actual button for me. The 'stupidity' affliction is a huge hinder for me, and I'd be missing balance a LOT even with a limiter. The AI doesn't waste time. Latency is another issue that the AI seems to overcome better than I would. Afflictions like amnesia and balance knocks go in the same category, natch.

Edit: PS: Having said all that, I'll be the first to admit that there's a good possibility that I'm not using my class to the full extent of it's capability yet, and that switching to a more manual (I will always use an envenomer for an aff class like Sentinel) style may give me more control to use the class a bit better. I've just not been convinced yet, and I've not made the time to find out for myself (shame on me, I know).