Let's Talk About 11.0

[Y2Quakepc2]

Illuminati?

[Raphael]

Yea.. The first person to create a decent bot for the new client will be one rich motherfu*ker.

The first one doesn't really matter. Take the scenario after the decay of Neobot, for instance: right after it went out, both RedBot and XenoBot were already arround, but it was iBot, released quite a while later, that dominated the market until WindBot, which came over an year later, took its position. Of course, you do have some lead by being the first one but it's not that much.

[Lilithwins]

Today, Cipsoft finally released a BETA of their new client. We've known this day was coming for almost three years, and that gave me a sufficient amount of time to plan for the impending doom of XenoBot.

Let me repeat myself: the impending doom of XenoBot.

However, it's not just XenoBot. This will be the doom of WindBot, MageBot, iBot, and every other bot. This isn't an update in the way most people think of them; it's completely new architecture. Every memory value, function signature, data structure, and little chunk of assembly code is different. This means that every bot will have to undergo a major overhaul, which can amount to thousands of hours of work, in order to function with the new client. In some cases, such as with XenoBot and likely others, it's safe to say that writing a completely new bot will actually prove easier than adapting the existing bot to the new client.

However, this isn't the end. There's a few ways the release of the client can play out:

1. The classic client stays around for a year or more while testing is done and bugs are ironed out of the new client.
2. Cipsoft pulls the trigger and kills the classic client within a few months and makes the move to the new client mandatory,

Here's how each scenario will play out from the perspective of the bot community:

1. Everything stays how it is, and bots keep updating for the classic client. In parallel, bot developers work on new bots for the new client, and plan to release them when the old client is dead.
2. We're left without bots for a while while bot developers work on new bots for the new client.

Clearly, the first scenario is what we want to happen. Either way, though, here's what you can expect from XenoBot:

• I will continue to update, improve, and support the bot until the classic client breathes it's dying breath.
• Once the classic client is dead, XenoBot will live on and be supported in the world of OT servers for years to come.
• When the day comes that the demand for XenoBot is too low to for me to keep up maintenance, I'll:
  
  • Create a stand-alone version of the bot that can be used by anyone, anywhere, for free.
  • Release the source code to the bot so that the public can learn from the code and, in some cases, ridicule the terrible coding decisions I've made.
  • Take the forums offline and leave up a minimalist website with download, usage, and compilation instructions.
  
  
• I don't see the above happening for a few more years, though, so rest assured we'll be around for a while.

Now, depending on how everything plays out and how insurmountable of a bitch the new client is, I may or may not be making a new bot for Tibia. At this point in time, I honestly am not sure. But, if I did, it wouldn't be XenoBot. It would be a fresh new bot that takes into account all of the learning and growing I've done over the 8 years that I've spent on XenoBot. If and when that day comes, I'll make sure everyone knows about it. But, right now, with no clue how everything will play out, what I'll be doing in life when the classic client dies, or how much demand there will be for a new bot, I really cannot give a definitive answer.

No matter how the next few months or years unfold, there's one thing I can say with certainty: it's been a pleasure. I started this bot when I was 15, and I made it for fun. It was my first project in C++. Over the years, you guys, the community, turned it into the monster it is today. You've pushed me to better myself and my software, to try new things, and to write code I never dreamed I could in the past. Turning this hobby into a business took me from a homeless high-school dropout living in a car to a successful C++ engineer with a bright career, a nearly published book, and a whole hell of a lot more experience than I'd have gotten otherwise. You're my customers, but you're also my friends and my fellow Tibians, and I wouldn't have gotten here without you. My life was made forever different by your support, and I thank you. Stick around and let's see how we can weather this storm together.

//DarkstaR

Youre not allowed to stop.

[Stusse]

The first one doesn't really matter. Take the scenario after the decay of Neobot, for instance: right after it went out, both RedBot and XenoBot were already arround, but it was iBot, released quite a while later, that dominated the market until WindBot, which came over an year later, took its position. Of course, you do have some lead by being the first one but it's not that much.

Got any information about WindBot's continued development in regards to this news?

/Stusse

[Mexen]

Got any information about WindBot's continued development in regards to this news?

/Stusse

According to them Windbot will run perfectly fine but may need some minor tweaks.

[DarkstaR]

The first one doesn't really matter. Take the scenario after the decay of Neobot, for instance: right after it went out, both RedBot and XenoBot were already arround, but it was iBot, released quite a while later, that dominated the market until WindBot, which came over an year later, took its position. Of course, you do have some lead by being the first one but it's not that much.

While that's true, you need to look at why. The reason iBot won is because RedBot was terrible and people were scared of injection because of the precedent set by NeoBot. It had nothing to do with the merit or funtionality of iBot, or a failure on the part of XenoBot, but rather the fear made by Ekx.

I'm sure when the new client comes to stay, this stigma will be somewhat depleted due to both the new environment and the example that XenoBot has set. Because of that, whether or not the first bot is injected, its timing will propel it to the top *unless* it happens to be terrible.

[DarkstaR]

According to them Windbot will run perfectly fine but may need some minor tweaks.

Minor is an understatement. In order to send accurate mouse clicks to the UI, the bot absolutely must have some idea of the structure of the client's UI. That's one of the most complicated things to work with currently, and the only reason they've managed (imho) is because the structure and information was freely released on tpforums.

They'll need to start from square one and reverse engineer this structure, and likely the creatures and containers structures, before they're able to do things like attack monsters, move items, and use the shop. Even with a perfect proxy, there's simply no way to do all of this without some way to identify the rough shape of the UI.

In fact, while XenoBot also relies on this data, it would likely be much easier for me to replace the existing XenoBot UI with an external UI, thus breaking my reliance on this data, than it would be for WB to reverse engineer the new UI structure.

[Rebellos]

Just when I planned better TA integration into the client... ;P

Anyway, the new client might be actually easier to RE because Qt unifies lots of stuff.
Just whole events ping-pong is annoying, but I believe it requires some IDA scripts matched to the Qt5 meta compiler used, and it should be easy enough to move around.

I'll drop a post on TPForums if I figure something out.

[DarkstaR]

Just when I planned better TA integration into the client... ;P

Anyway, the new client might be actually easier to RE because Qt unifies lots of stuff.
Just whole events ping-pong is annoying, but I believe it requires some IDA scripts matched to the Qt5 meta compiler used, and it should be easy enough to move around.

I'll drop a post on TPForums if I figure something out.

While it's true that QT leaves behind a sufficient amount of metadata, it's actually the structures that will be difficult. When everything is coded with C++ best practices, it's the data structures that bite your ass. Proper OOP means pointer chains for days. It's slightly more complex to read an std::vector than a flat array, and extremely more complex to read an std::map or std::set. std::list is somewhere in the middle, but if they're using things like QString and QList, it's much harder again. It really depends on how it was designed.

[spook]

No matter how the next few months or years unfold, there's one thing I can say with certainty: it's been a pleasure. I started this bot when I was 15, and I made it for fun. It was my first project in C++. Over the years, you guys, the community, turned it into the monster it is today. You've pushed me to better myself and my software, to try new things, and to write code I never dreamed I could in the past. Turning this hobby into a business took me from a homeless high-school dropout living in a car to a successful C++ engineer with a bright career, a nearly published book, and a whole hell of a lot more experience than I'd have gotten otherwise. You're my customers, but you're also my friends and my fellow Tibians, and I wouldn't have gotten here without you. My life was made forever different by your support, and I thank you. Stick around and let's see how we can weather this storm together.

//DarkstaR

You're inspiring! Keep up the good work!