• There is NO official Otland's Discord server and NO official Otland's server list. The Otland's Staff does not manage any Discord server or server list. Moderators or administrator of any Discord server or server lists have NO connection to the Otland's Staff. Do not get scammed!

OTC 1.0 Comparison

The source code he released is actually usable and not completely useless as you claim it to be. Have you actually tried to compile it?
Yes, I've tried to compile and it's missing crucial framework files. Kondrah himself said it's not possible to compile what he shared.. It's not useless, of course, there is a lot to be studied there, but it's not that useful either, since the main things that I'd like to learn from otcv8 are in the graphic framework files that were removed! :)

See, you are just spewing it out things at this point. Only a small portion (less than 10%) of the codebase has been redacted (none of which makes the entire codebase useless). You are essentially just missing project files from what I can see, and it shouldn't be too hard getting it too work with vanilla OTC project files.
I can't believe you're serious. The 90% that he released are mainly files that already exist with some refactor and modules. However, those 90% uses a 10% of a framework that was refactored to use threads and workers in order to build graphics with better performance and this is exactly what wasn't shared and, tbh, this is the only really relevant thing in the whole otcv8 project. The rest (bot, nice refactors and modules) i couldn't care less.
 
Last edited:
Yes, I've tried to compile and it's missing crucial framework files. Kondrah himself said it's not possible to compile what he shared.. It's not useless, of course, there is a lot to be studied there, but it's not that useful either, since the main things that I'd like to learn from otcv8 are in the graphic framework files that were removed! :)
so basically nothing
 
so basically nothing
Exactly, basically he gave us a ferrari, but he removed the engine. Quite useful, but not. Ofc we can study the breaks, suspension, tires, etc. and get better on it, but it will never be close to a ferrari without engine.
 
Yes, I've tried to compile and it's missing crucial framework files. Kondrah himself said it's not possible to compile what he shared.. It's not useless, of course, there is a lot to be studied there, but it's not that useful either, since the main things that I'd like to learn from otcv8 are in the graphic framework files that were removed! :)


I can't believe you're serious. The 90% that he released are mainly files that already exist with some refactor and modules. However, those 90% uses a 10% of a framework that was refactored to use threads and workers in order to build graphics with better performance and this is exactly what wasn't shared and, tbh, this is the only really relevant thing in the whole otcv8 project. The rest (bot, nice refactors and modules) i couldn't care less.
Okay, I'll take your word for it. Must have been my imagination running amok :)
 
Okay, I'll take your word for it. Must have been my imagination running amok :)
I don't want to insult your intelligence, but did you actually opened and tested the project or you just compiled in a merge with otc?
Because the src from Kondra have even missing files you can't simply pick from otc. Also it has a lot of codes "removed" that basically makes it useless, even for cherry-picking.
It seems you're just taking his word on this one and assuming he's a "good guy" for releasing a useless, outdated and obfuscated piece of code.
 
I don't want to insult your intelligence, but did you actually opened and tested the project or you just compiled in a merge with otc?
Because the src from Kondra have even missing files you can't simply pick from otc. Also it has a lot of codes "removed" that basically makes it useless, even for cherry-picking.
It seems you're just taking his word on this one and assuming he's a "good guy" for releasing a useless, outdated and obfuscated piece of code.
I don't mind you questioning my intelligence, feel free to do so. I'm not taking his words for granted nor do I think he's a "good guy" for whatever reason. I'm one of those whom actually got access to the complete source code. I replied to your message earlier because I did a diff check on the source code that had been released publicly to see if your claim was real or not, and it turned out that things weren't as useless as one might think.
 
I don't mind you questioning my intelligence, feel free to do so. I'm not taking his words for granted nor do I think he's a "good guy" for whatever reason. I'm one of those whom actually got access to the complete source code. I replied to your message earlier because I did a diff check on the source code that had been released publicly to see if your claim was real or not, and it turned out that things weren't as useless as one might think.
Well, the invite of my initial topic also applies for moderators, feel free to elucidate us with such diffs and what actually be turned into useful PRs for OTC1.0. Any help would be more than welcome.
 
@Shadowsong It's hard to understand what someone is feeling by what they write, words don't quite describe what we want to say, specially considering the language barrier. I'm actually a way more nicer and interested person than I actually sound like, I'm still working to try and improve this.

I agree with a lot of what you pointed here, devs eventually will be otadmins and the opposite should happen as well, at least in some degree. But I'm trying to say is that unfortunately I can't expect someone who just want to run a project to assist me designing machine learning algorithms. They can help by testing, reporting bugs and giving tips of what could be useful to extend/improve.

Let me give you a very personal example that lately I've been strugling to decide on how to tackle. I have created a site that you simply enter, press play in some parts of it and then you get basically some sliders that allow you to generate any kind of weapon sprite in 32x32. Messing with some sliders will allow you to control intensity of details, color, shape, even the size of hilt. This is already a reality that I've been working on for about one month and a half pratically 24/7 as even when I was sleeping/working it was training in parallel.
My objective is to replace all tibia sprites, ofc this isn't a one person job because even with this tool I probably won't be able to replace a few unique models and looktypes will definetely be a challenge since you need to find versions of the same seed in different positions.

I can look each one of the infinite seeds, try to find ones that resemble tibia sprites and start replacing them. This a huge work if we consider only the ~300 weapons we have in the game. Besides this we would probably need to correct some of them manually (shadow details and some noise pixels in background). That's where community could enter in: If I share this project with others and started a repo with a list of what are the sprites missing so people could pull request the models they generated to replace them we could finalize it very quick.

While they do it, I could focus in studying more about this and try to move on to the next thing (sets and shields, probably), and after a few iterations of this we would probably have replaced 150k sprites and have tools to generate infinite possibilities of sprites. Amazing huh?

The problem is that I can't stop the feeling that if I open that, no one will be helping to replace tibia sprites and creating a single .spr that we could all use, but rather, people that have resource would probably exploit this and push even further our spriters from this community. How to deal with this problem?

I think there is no perfect solution to your problem, you will eventually have to settle for a less-than-ideal situation. Either give access only to people you trust, or release publicly knowing that it will be used in the way you described almost certainly. But that's the nature of releasing things open-source. It appears as if you are conflicted on taking the risk which comes with releasing open source code at the cost of your own time and frustration. You don't want to be that developer who spends all this time to create a tool, then someone else profits off of it or uses it in a way you don't approve of, while your original intention goes ignored. But once you let it out of your hands, it's out of your hands man.

It's a tricky dilemma, for sure, but you shouldn't be expecting anything in return if you're gonna release it open source, including that the strangers will reciprocate your aspirations for group work and collective success. The sooner you come to terms with yourself with which one of those scenarios is more acceptable to you personally, the sooner you can find a realistic solution to your dilemma.

We've already had an attempt to complete your goal, but it struck far from the finish line. Nevertheless, it was a nice effort and those sprites are still useful and out there, kudos to @Peonso and everyone who donated sprites.
But it's not such an easy task to replace all Tibia sprites, it's a lot of sprites (especially if you are aiming for newer versions... at which point, we are back to the question - which version is the baseline version you wanna go for?)

Either way, I don't think I'm the best person to talk on this topic, as I can see that my views would surely conflict with yours and you probably won't hear what you want to hear from me, which is fine, I'll have to agree to disagree with some of your views too, we can't all think the same, which leads me into the next thing:

I believe those discussions could all be avoided if we just understand that ~80% of the path is the same for everyone.

It's hard to understand that when that 80% is a gross overestimation. You can sample some people out of this thread or forum, and they'll all have different ideas what they want out of TFS/OTC/Sprites for their project. We've kinda seen that happen already.

You should try to make a successful thread where you will figure out an optimal set of features and sprites that need to be present for this custom MMO engine/client set you want to build, and then you can better focus on organizing a roadmap. It's going to be difficult, but I believe it is the crucial first step necessary to progress with your ideas.

Why bother to do it solo when we could be working on something that we all could be using and testing together?

I would enjoy giving a realistic viewpoint on this question as much as I'd enjoy running a marathon barefoot with a bunch of legos glued to the bottom of my feet, so don't ask me to explain to the kids why santa isn't real, I think you know it all too well. 😛
 
@Shadowsong
Actually the reason why I target this question to you is because I saw your signature, I would really like a honest feedback on this one.

The 'gross' estimation is actually a pareto distribution that seems to work when we compare diffs of different projects, so it's not something I just made up. They can have different ideas, but in general they don't manage to actually put those ideas in practice, and that explains the similarities.

I'm bet by now you are judging me naive by my beliefs towards people working together, this is actually happening already. I'm not asking for something I'm not willing to commit myself to, I'm inviting others to participate in something greater.
Each of us have different beliefs about perfect server and there's even people who don't bother to make a server at all. Vast majority of the things are common to everyone.
 
@Shadowsong
Actually the reason why I target this question to you is because I saw your signature, I would really like a honest feedback on this one.

I told you in my reply how I would approach this in your place.
I will share with you how I feel about this topic:

I tried already to work on such a project (replace all tibia sprites), and it's an astonishing amount of work, especially if you want them to look anywhere decent. If you try to include multiple artists to accomplish this, you will, first of all, have trouble even finding them (assuming they're not complete beginners), and then on top of that you will face wild style inconsistencies which will make the project look like a clown fiesta in summation.

I know people currently actively looking for decent Tibia spriters (not just other people, I myself was looking for a spriter until a few days ago too), who are willing to pay good money, and it's next to impossible to find. Thinking you will find someone to do this workload, do it right, and do it for free is lunacy in my experience.

Furthermore, the current .spr format and the tools we have currently for editing .dat/.spr are not optimal and result in huge wastes of time when doing sprite addition or replacement. You need to take a weekend off to just implement a few (finished) smooth-walking idle-animated outfits, not to mention the amount of effort it goes into creating them. That's about 4608 32x32 px frames for a single outfit (male+female). Try doing it, and check how you feel realizing you have to do this about 30-40 more times, and that's just human outfits - what about literally everything else?

You need to put into scope what you're aiming at.
Cipsoft has been building their sprite repository with numerous artists and (now) millions in their budget, for 2 decades.
Your best bet, and smart move, is to try and employ AI to match that kind of workload, or reduce the workload very very drastically by cutting features and inclusions left and right, with a machete.

I personally wouldn't do that kind of massive amount of work and then release it (which isn't even OT-exclusive and can be abused and resold literally anywhere where there's a demand for sprites), especially not after going through the kind of stuff I've gone through with various community members while trying to be fair to people. ¯\(ツ)

The 'gross' estimation is actually a paretto distribution that seems to work when we compare diffs of different projects, so it's not something I just made up. They can have different ideas, but in general they don't manage to actually put those ideas in practice, and that explains the similarities.

I'm bet by now you are judging me naive by my beliefs towards people working together, this is actually happening already. I'm not asking for something I'm not willing to commit myself to, I'm inviting others to participate in something greater.
Each of us have different beliefs about perfect server and there's even people who don't bother to make a server at all. Vast majority of the things are common to everyone.

EDIT: Sorry, this part got cut out.

What's the point of comparing diff when there is no real choice or alternative? Most servers here use some version of TFS and that's an undeniable fact. Not because they really want those features, or that server, but because there's not much other choice, and they don't have the knowledge nor time to put together a new engine which fills only their needs. If there is no real choice there, of course the majority is using the same codebase and basing everything on the same code.

Yes, I judged you by what you said, what else is there to judge by, it really came off as naive. If you are able to prove this wrong, then please do, but before you invite people to do projects of such scope, show some plan, show some roadmap, show a strict set of goals and that you're serious about it. I haven't yet seen an agreement among the community on how this should be done.
 
Last edited:
I don't mind you questioning my intelligence, feel free to do so. I'm not taking his words for granted nor do I think he's a "good guy" for whatever reason. I'm one of those whom actually got access to the complete source code. I replied to your message earlier because I did a diff check on the source code that had been released publicly to see if your claim was real or not, and it turned out that things weren't as useless as one might think.
That explains a lot! So enjoy the 5k you wasted and please stop saying what you don't know. As I showed before, kondrah himself said what he released has nothing of the core functionalities. :)
 
I personally wouldn't do that kind of massive amount of work and then release it (which isn't even OT-exclusive and can be abused and resold literally anywhere where there's a demand for sprites), especially not after going through the kind of stuff I've gone through with various community members while trying to be fair to people. ¯\(ツ)
your opinion sounds a lot like a few brazilian spriters I talked to such as Leeshrot, mind adding me in discord so we can discuss this topic in particular? I don't want to go offtopic Night Wolf#8602
 
I personally wouldn't do that kind of massive amount of work and then release it (which isn't even OT-exclusive and can be abused and resold literally anywhere where there's a demand for sprites), especially not after going through the kind of stuff I've gone through with various community members while trying to be fair to people.
1) Much of the world is already starting to receive competition from AI, it's inevitable and only a matter of time and impact and for us a matter of timing when, not if, to strike with this kind of technology. In other words only crucial details are missing such as human timing, organizing things on a human level etc, it's not a matter of "should we do this" - no, the cat's already out of the bag, now it's just a matter of not having the cat pee all over the place by getting it a sandbox and potty-training it.
2) From what I've seen so far, I'd say any half-decent artist should be able to compete with GAN2 as applied by NW currently in terms of quality (NOT quantity!) so far. This is of course good, it means spriters just has to up their game and provide "better" quality content, such is the state of the world anyways, technology improves and competition increases with it. At the same time the quality is pretty decent, make sure to note that this is not an insult, it seems very powerful in terms of quantity imo and I saved a couple of sprites on my computer in amazement of the quality of some of the weapon sprites generated, but in terms of quality spriters still has a place -- always keep in mind that what AI has in terms of production value will likely lagg behind in human creativity, at least for a while, so stay creative or imaginative, create what you want instead of mass-produce things, and you'll hopefully stay ahead of AI, at least for a while.
 
Last edited:
1) Much of the world is already starting to receive competition from AI, it's inevitable and only a matter of time and impact and for us a matter of timing when, not if, to strike with this kind of technology. In other words only crucial details are missing such as human timing, organizing things on a human level etc, it's not a matter of "should we do this" - no, the cat's already out of the bag, now it's just a matter of not having the cat pee all over the place by getting it a sandbox and potty-training it.
2) From what I've seen so far, I'd say any half-decent artist should be able to compete with GAN2 in terms of quality (NOT quantity!) so far. This is of course good, it means spriters just has to up their game and provide "better" quality content, such is the state of the world anyways, technology improves and competition increases with it. At the same time the quality is pretty decent, make sure to note that this is not an insult, it seems very powerful in terms of quantity imo, but in terms of quality spriters still has a place.

You're absolutely right, it's scary what can already be done in terms of graphics via AI, and in the coming years, I can only imagine the kind of possibilities and advances that will be done in this field, it's possible that it will make digital artists, as we know them today, completely obsolete, alongside many other professions. Half of me is dreading the day when such things become an easily usable reality, but the other half is excited for how much other things you could focus on if unique graphics were no longer a big deal to generate.

I can't speak for other spriters, just for myself, but I already suggested to @Night Wolf that he should employ that AI to do such tasks if possible, because employing human labor on such scale would be insanely costly. Out of all the tools and things that have been contributed to this forum since its inception, a full rework of the (not even current, let's say - 9.60) Tibia sprite pack would probably take most of the time.

If @Iryont is willing, he could share for comparison some interesting data of how many sprites he had to replace to get Medivia where it is - and how much time that took. I think the only ones who managed that feat next to Medivia are Zezenia, but I don't think anyone from their dev team is here. And keep in mind that any newer datapack has way more sprites than the old version Medivia is based on.

Either way, it is a fun tangent to explore but I think I've already gone way off-topic in my last few posts, so I'll stop there.
 
You're absolutely right, it's scary what can already be done in terms of graphics via AI, and in the coming years, I can only imagine the kind of possibilities and advances that will be done in this field, it's possible that it will make digital artists, as we know them today, completely obsolete, alongside many other professions. Half of me is dreading the day when such things become an easily usable reality, but the other half is excited for how much other things you could focus on if unique graphics were no longer a big deal to generate.
I'm gonna stop insisting on driving this point through after just this one more try: keep in mind that human creativity fortunately is ahead of AI's scope.
AI doesn't know what it is to eat food, to sleep, to feel fear, or get laid, it only knows how to generate things (or respond) based on the input it receives - at least basically so, maybe I'm off on some technical details, but the general point should stand true.
In other words it has massive power in terms of quantitative production, but it can't really tell, on its own, what the human experience is like, this is (originally) the field of the artist, so there's now the time for a return.
 
I'm gonna stop insisting on driving this point through after just this one more try: keep in mind that human creativity fortunately is ahead of AI's scope.
AI doesn't know what it is to eat food, to sleep, to feel fear, or get laid, it only knows how to generate things based on the input it receives.
In other words it has massive power in terms of quantitative production, but it can't tell, on its own, what the human experience is like, this is (originally) the field of the artist, so there's now the time for a return.

This is what machine learning is for, what are the odds such A.I is able to gather and save all information and data based on images online about wars and actual weapons and use them as source for creativity? In modern's world, an A.I intellect would come entirely from the internet, and that's human made and everything we are is in there.
 
This is what machine learning is for, what are the odds such A.I is able to gather and save all information and data based on images online about wars and actual weapons and use them as source for creativity? In modern's world, an A.I intellect would come entirely from the internet, and that's human made and everything we are is in there.
I'm aware of the concept of a singularity. It's a hypothesis, yet to be proven. Even the most advanced AI technology I've looked at, not to say that I've looked at much of this at all, is lagging far behind in for example social understanding.
Algorithms are developed by humans after all.
In case this is going off-topic we should/could continue this discussion elsewhere at some point in time, I just wanted to drive this one point through.
Also, just for the memes.
(Dry version without the optional ba dum tss)
 
Last edited:
You're absolutely right, it's scary what can already be done in terms of graphics via AI, and in the coming years, I can only imagine the kind of possibilities and advances that will be done in this field, it's possible that it will make digital artists, as we know them today, completely obsolete, alongside many other professions. Half of me is dreading the day when such things become an easily usable reality, but the other half is excited for how much other things you could focus on if unique graphics were no longer a big deal to generate.

I can't speak for other spriters, just for myself, but I already suggested to @Night Wolf that he should employ that AI to do such tasks if possible, because employing human labor on such scale would be insanely costly. Out of all the tools and things that have been contributed to this forum since its inception, a full rework of the (not even current, let's say - 9.60) Tibia sprite pack would probably take most of the time.

If @Iryont is willing, he could share for comparison some interesting data of how many sprites he had to replace to get Medivia where it is - and how much time that took. I think the only ones who managed that feat next to Medivia are Zezenia, but I don't think anyone from their dev team is here. And keep in mind that any newer datapack has way more sprites than the old version Medivia is based on.

Either way, it is a fun tangent to explore but I think I've already gone way off-topic in my last few posts, so I'll stop there.

Machine learning is here to stay, and even if we don't want to adapt to this new reality, sooner or later we'll be forced to. I've been discussing the issue with several friends, and we all concluded that many jobs will disappear / become redundant in the near future. Yes, near future... at the rate these technologies are advancing, in 5 years we'll be seeing machine learning do things that we thought impossible today, just as at this very moment we're seeing computers do stuff that if you told someone about 5 years ago, you would have been laughed at.

Digital artists will be amongst the first ones to go, and I don't mean any offense to anyone who does it professionally or as their hobby. If you think about it, its the same issue as with Uber vs. Taxis, or McDonald's employees vs "Automated PoS machines". At the end of the day its more efficient, faster, cheaper...and even if some people in the affected sector try to fight it off, it'll take over eventually.

For those that still think this is all a fantasy, I highly, highly recommend watching some of the videos in this youtube channel Two Minute Papers (https://www.youtube.com/user/keeroyz)

As it's been mentioned here, could use machine learning to help us complete the OT sprite pack. In addition to using the AI to directly generate sprites after training it (I have no idea on how difficult it'd be to achieve the desired results), there's other machine learning programs out there that can generate sprites or sprite-esque graphics out of images, which could then be lightly edited to match with the desired style. We could start by creating a replacement pack from the lower versions and work our way upwards 7.1 has like 2000~ different sprites, 7.4 4000~ 7.7 9000~ 8.0 20,000~ ... and so on and so on, I'm already thinking about starting such a project.
 
Back
Top