how about these who own the source code , release it already. Many ppl got the src by reversing it, some already finished reversing it 1:1 with the binaries code...
how about these who own the source code , release it already. Many ppl got the src by reversing it, some already finished reversing it 1:1 with the binaries code...
Which of these will never happen? The release of the original source code, or the reversal of the src from binaries ? jk
I might think of a reason why the src is not yet released, yet one of jmx srcs got leaked and downloaded by many ppl, that was a kinda good base to start with reversing it from scratch, even tho that src had nothing to do with silkroad, but at least gave the base and idea how the src was built..
Yeah, right... Since the AI became so powerful, nothing can stand in front of it, I guess in a range of one year, the psro community and this forum will look different..
Yeah, right... Since the AI became so powerful, nothing can stand in front of it, I guess in a range of one year, the psro community and this forum will look different..
wouldnt count on that right now
We are seing a surge of game "Decompilations" that end up being the AI stealing code from emulators to THEN create a botched emulator
Right now, the AI would help understand the code but it wouldnt 100% reverse engineer the src
We are seing a surge of game "Decompilations" that end up being the AI stealing code from emulators to THEN create a botched emulator
Right now, the AI would help understand the code but it wouldnt 100% reverse engineer the src
Correct, we are still a long way away from anything meaningful in that aspect.
Also, do people actually like that DLSS5 **** on SRO? Because yeah, it's pretty, but ****, it looks awful. I come back for a nostalgic factor in regards to graphics, music, memories. If I wanted a game that looked entirely different, I'd play a different game
We are seing a surge of game "Decompilations" that end up being the AI stealing code from emulators to THEN create a botched emulator
Right now, the AI would help understand the code but it wouldnt 100% reverse engineer the src
Nice observation. I agree that AI, at least right now, can't reverse-engineer the original source 100% accurately - every detail, class, member, structure, etc. That's simply not realistic.
What I do wonder about, though, is the background of the people attempting these decompilation projects. Are they actual developers? Do they have a solid understanding of ASM? Are they actively involved in the reversing process, or are they basically just feeding binaries to an AI and hoping it figures everything out?
In my case, I do use AI to help reverse modules, but I've also been doing this manually for several years. I've extracted a ton of data from these modules by hand, I have a pretty solid understanding of their ASM data, and I've built up a lot of documentation along the way.
With that context, AI actually does a great job. I can guide it, correct it when necessary, and use my existing knowledge to help reconstruct the source.
That said, I know for a fact that the resulting source will never be 1:1 identical to the original. That's basically impossible.
But there's another side to it too: if you spend enough time analyzing the original code, you also start noticing mistakes, questionable design decisions, and things the original developers could have done better. In that sense, AI isn't just helping reconstruct the code - it can also help clean it up, improve parts of it, and modernize it.
For example, in my case I'm also using the process to move things from 32-bit to 64-bit and rewrite the src using modern C++.
So yeah, I definitely wouldn't say AI can magically recover the original source code, but in the hands of someone who already understands what they're reversing, it's an extremely useful tool.
I would love to see the original source code and compare it to my current progress, honestly
Nice observation. I agree that AI, at least right now, can't reverse-engineer the original source 100% accurately - every detail, class, member, structure, etc. That's simply not realistic.
What I do wonder about, though, is the background of the people attempting these decompilation projects. Are they actual developers? Do they have a solid understanding of ASM? Are they actively involved in the reversing process, or are they basically just feeding binaries to an AI and hoping it figures everything out?
In my case, I do use AI to help reverse modules, but I've also been doing this manually for several years. I've extracted a ton of data from these modules by hand, I have a pretty solid understanding of their ASM data, and I've built up a lot of documentation along the way.
With that context, AI actually does a great job. I can guide it, correct it when necessary, and use my existing knowledge to help reconstruct the source.
That said, I know for a fact that the resulting source will never be 1:1 identical to the original. That's basically impossible.
But there's another side to it too: if you spend enough time analyzing the original code, you also start noticing mistakes, questionable design decisions, and things the original developers could have done better. In that sense, AI isn't just helping reconstruct the code - it can also help clean it up, improve parts of it, and modernize it.
For example, in my case I'm also using the process to move things from 32-bit to 64-bit and rewrite the src using modern C++.
So yeah, I definitely wouldn't say AI can magically recover the original source code, but in the hands of someone who already understands what they're reversing, it's an extremely useful tool.
I would love to see the original source code and compare it to my current progress, honestly
it's not even funny
We are seeing alot of stupid decompilations that, if you look into the code, you will find it is a wrapper with a layer of emulator. NO actual decompilation is being made
We are seeing alot of stupid decompilations that, if you look into the code, you will find it is a wrapper with a layer of emulator. NO actual decompilation is being made
Have you seen on X recently, the amount of emulated browser games (GTA 5, Half-Life 1, QUAKE, CS1.6, etc etc) are being uploaded on a daily basis? I think there are also a few projects working on fully playable SRO's via browser too, which is interesting. Albeit, nothing to do with decompiling, but cool I think
We are seeing alot of stupid decompilations that, if you look into the code, you will find it is a wrapper with a layer of emulator. NO actual decompilation is being made
I can't speak for other people's projects, and I don't really want to. But have a look at this:
This is essentially a reconstruction of the original damage formula. The logic isn't contained in a single function either. It's spread across quite a few functions, all of which had to be analyzed, understood, and pieced back together. With proper guidance, AI managed to reconstruct all of that in less than 10 minutes.
Based on the results, the accuracy is incredibly close to the original behavior, I'd say around 99.99%, including some of the smallest details. And I'm saying this from my own experience, having spent years manually reversing these modules long before AI came into the picture.
Of course, to confirm whether every single detail matches the original implementation, we'd need someone with access to the actual source code. But the results so far speak for themselves.
I'm not sharing this to prove that AI can reconstruct an entire source code perfectly, nor am I making any claims about other decompilation projects. It's just one example, albeit a fairly complex one, of what AI can accomplish when it's in capable hands and given the right guidance.
As for the bigger picture, I'd rather not jump to conclusions just yet, my initial claim was in a period of 1 year..
In majority I agree with many of your points, but...
-- edit:
Obviously the GameServer itself, isn't the source-code... The whole source code consists in server modules, client interface and 3d modelling tools.
Have you seen on X recently, the amount of emulated browser games (GTA 5, Half-Life 1, QUAKE, CS1.6, etc etc) are being uploaded on a daily basis? I think there are also a few projects working on fully playable SRO's via browser too, which is interesting. Albeit, nothing to do with decompiling, but cool I think
it feels like a coordinated strike against gaming. Kinda like there was someone trying sabotage the entire decomplilation projects
Quote:
Originally Posted by leo2111
I can't speak for other people's projects, and I don't really want to. But have a look at this:
This is essentially a reconstruction of the original damage formula. The logic isn't contained in a single function either. It's spread across quite a few functions, all of which had to be analyzed, understood, and pieced back together. With proper guidance, AI managed to reconstruct all of that in less than 10 minutes.
Based on the results, the accuracy is incredibly close to the original behavior, I'd say around 99.99%, including some of the smallest details. And I'm saying this from my own experience, having spent years manually reversing these modules long before AI came into the picture.
Of course, to confirm whether every single detail matches the original implementation, we'd need someone with access to the actual source code. But the results so far speak for themselves.
I'm not sharing this to prove that AI can reconstruct an entire source code perfectly, nor am I making any claims about other decompilation projects. It's just one example, albeit a fairly complex one, of what AI can accomplish when it's in capable hands and given the right guidance.
As for the bigger picture, I'd rather not jump to conclusions just yet, my initial claim was in a period of 1 year..
In majority I agree with many of your points, but...
-- edit:
Obviously the GameServer itself, isn't the source-code... The whole source code consists in server modules, client interface and 3d modelling tools.
that's a different thing. We are talking people that are doiung absolutely jack and are delivering botched projects made of pieces of actual projects that people worked hard on
Art Chit-Chat 09/21/2026 - General Art - 8188 Replies Ich habe mir mal so gedacht : "Wieso denn nicht einfach mal über unsere Kunstwerke labern?
Die Hintergrundidee ist eigl. um über Kunstwerke oder allgemein über Photoshop & Co. zu reden ohne Warnings zu kassieren.
Vllt braucht manns vielleicht auch nicht.
Soll sowas wie ein Längster für uns artists sein xD
Naja viel spaß.
_____
Hardware Chit-Chat 02/02/2025 - Hardware Discussions / Questions - 19418 Replies Hallo,
ich werde in naher Zukunft noch die Admin um Erlaubnis fragen, doch erst einmal werde ich euch erzählen, was ich hiermit meine.
Dies hier soll eine Art Diskussions- und Unterhaltungsthread (kein Spam!) sein. Das bedeutet, dass Leute die sich sehr für Hardware interessieren hier mit anderen über gerade verschienene Hardware diskutieren können oder sich auch einfach aus Spaß an der Freude über verschiedene Zusammenstellungen auslassen können. Man könnte das Ganze hier auch als ein...
Diablo 3 Chit-Chat 09/29/2021 - Diablo 3 - 12834 Replies Vom OT- und LoL-Chit-Chat inspiriert, einfach mal erstellt.
Nicht dazu da um Probleme und ausführliche Diskussionen zu behandeln, dazu sollte es immer noch eigenständige Threads geben. Keine Flames oder Zwei-Wort-posts. Have fun. :)
[Vindictus] Chit-Chat 01/05/2018 - Vindictus - 1028 Replies http://www.elitepvpers.com/images/us.png
English Version:
Hello Users,
please use this thread to discuss Vindictus related problems or to have a little chit-chat about the game. All the usual forum rules apply to this thread.
You might post here if you have a small question or if you want to share some videos or screenshoots with us.
http://www.elitepvpers.com/images/de.png
co partner chit chat 01/19/2006 - Conquer Online 2 - 3 Replies people plz post the latest progress in the new co partner here for the latest progress plz
jk
