https://www.pcgamer.com/heroes-of-might-and-magic-2-fans-are-building-it-a-better-engine/

Very interesting. I wonder, where did they/ you get the source code form to modify it.

Are you involved in this?

As far as features go, all I ever wanted for this game was to fix resolution so that game would adapt monitor size. I guess world map in particularly. After seeing this, it is plain, that everything needs to be redrawn. Castle screen should be full screen as well not in a box.
Other features... maybe so world would scroll smoothly. I
n hot seat the fact that only the last player could see pc move was also a bitch. :)
Maybe a hd graphics in same style and manner.
I guess mainly visuals.
Functionality( or absence of it) never bothered me that much.
That fast split function it's neat but it doesn't really add that much to this game. :)

No I just seen the article thought it was cool.

And how feature complete is this engine nowadays? I've been playing it on Pocket PC, before M$ buried Windows Mobile. And functionality was missing for even for some basic castles at that times.

From what I can tell by playing with it occasionally, it's pretty much complete when playing single missions, it still needs some work to fully supports the campaigns extra features

The thing i miss the most, personally, is some form of graphical scaling, like the HoMM3 HD mod, so that one can increase resolution without everything being reduce to tiny size... even a basic 2x scaler with no filters would be nice

I also tried the PS Vita port and it works pretty nicely even using the pad, but again the tiny size is a bit of an issue

It looks like the project will bring HD rez. Anyway There is even a font resize in it.

Source would help, but it's not necessary. Since the publisher no longer releases updates, the project can pick a single version, probably the last published patch, and try to write a program that works like that version did. When the new game does not play like the retail one did, they can study the observed output of the original game and adjust the new game to do the same thing. For some games, and I think HoMM 2 is probably in this category, the community has built up a very detailed understanding of game mechanics, so the clone can match those mechanics without the source of the original game, and probably even without any EULA-prohibited reverse engineering. The project isn't trying to rebuild the same executable that the publisher released; it just needs to build a program that plays the same way in every aspect that people care about.

Since the FHeroes2 game requires original assets to run, they don't need to deal with recreating all the sound/artwork from the original game. They just need to be able to read the original game's files and play/render them correctly. That saves a lot of time, makes the game more comparable to the original (and more familiar to people who played the original), and potentially avoids disputes about whether replacement assets were created in a way that infringes on the originals.

I understand that you can recreate the game and write everything from scratch.
That's very impressive.
And insane.

I took me a while just to write a simple homm2 score calculator because of back engineering.

Original source code would be a lot of help... in Fheroes lets take world_loadmap.cpp line 620.
It looks like if a monster is ghost or elemental, it can never ever join the army. Is that true for original game? Can you test it enough?
Can you make that call? And that's just one thing out of thousands. You have to do a lot of testing...

Are you involved in this?

https://github.com/ihhub/fheroes2

From what I get out of it is it's being worked on (Fheros) and is heading for a 1.0 release it's at .93. It's a engine redo.

Devilution did it for Diablo, and even went the amazing extra step of targeting the ability to exactly rebuild the original executable, function for function. OpenXCom did a reimplementation (logic only, not executable) for X-Com: UFO Defense and X-Com: Terror From the Deep. There may be others I don't know about.In part, that's because you wanted it to be right for every case you could test. If you had settled for making it mostly right, and left a call for bug reports on edge cases people find, you could have delegated out some of that engineering work.As a tip, github lets you link to exact lines, which makes it easier for people to follow along. You appear to be referring to https://github.com/ihhub/fheroes2/blob/69d00b747f36020310e5f52374b91682e9cad9cf/src/fheroes2/world/world_loadmap.cpp#L620I seem to recall that ghosts never join. I don't recall whether elementals are likewise guaranteed never to join, but I've never had a case of an elemental joining.Yes. The nature of such projects is that you make the call that seems right, and let users report if you got it backward. I suspect if someone reported a bug, and provided a savegame where an elemental joined, the developers of fheroes2 would investigate why that worked. It's also possible that the developers did engage in targeted reverse engineering / code inspection for some of these edge cases. At scale, that would be tedious, but for targeted problems where you need to answer a specific logic question, it might be viable.No.

0.9.6 released

Another solid portion of fixes and added Nintendo Switch support. Love that project.

I just tried fheroes2, wow, the implementation is nearly complete. It's rare for remake projects to get that far, I remember checking it out what seems like a few years ago, and it was barebones.

FHeroes2 project is officially out of Beta stage and has reached version 1.0.0.
Engine currently is feature-complete for single-player and hot-set multi-player games only.