H7 vs taking things into our own hands

[Steven Aus]

Are you going to make it available for Linux? I know a lot of people are getting sick of the privacy intrusions by Microsoft and Google and all the rest, and it would be great if a homm-inspired game made by the fans for the fans also included Linux, since it is the closest thing we have to an unpolluted OS environment. I myself am considering switching to Linux. I know for example that two of my favourite games that I bought on Steam are available on Linux - "Thea: The Awakening", and "Conquest of Elysium 4". What makes it interesting is that they both can be played without Steam too. Apart from the updates, these two games can be played entirely without Steam. And Thea has made it to GOG.com recently too, although AFAIK they are not into Linux as much as Steam is.

I'm looking forward to more news about your project - I'd love to hear when you're ready to talk about it.

[Galaad]

Are you going to make it available for Linux? I know a lot of people are getting sick of the privacy intrusions by Microsoft and Google and all the rest, and it would be great if a homm-inspired game made by the fans for the fans also included Linux, since it is the closest thing we have to an unpolluted OS environment.

+1
Linux may be a bit hard at the beginning but is totally worth it. Stability, performance, no trackwares no BS. And it's free. I would totally love it if this goes to Linux.

I myself am considering switching to Linux.

You should, start with Ubuntu then when you understand how it works and are comfortable in the Terminal go Debian: it's basically the same without the pre-installed stuff for noobs thus even more stable, freaking solid.
Just a note however, avoid hybrid cards, and if you plan on buying a new machine check that pieces are compatible (will save you some trouble).

And yes, more and more games (and some of the great ones) get Linux support nowadays, I don't know if GOG is less into Linux support than Steam, what makes you say that? Just curious as I use both.

I'm looking forward to more news about your project - I'd love to hear when you're ready to talk about it.

+1 again, to me is the only hope left for a good modern Heroes game (still playing h3 and h5 heh).

[Steven Aus]

Oh, do GOG games have linux support on GOG itself? I was under the impression that GOG was Windows-only, but I may well be wrong.

[Galaad]

Nah you can get the Linux package directly from gog as well if it exists.

There is even a little program called lgogdownloader allowing you to get your stuff directly from command line lol

[Steven Aus]

Thanks for that. Really good to know.

Anyway, getting back on topic: I love to see Heroes of the Ancient Order get a Linux release (it would be perfect if it would be at the same time as the Windows release).

[Galaad]

I wouldn't count on it too much but I obviously agree, also on philosophical level. Though if Windoze version gets released first and Linux is TBA would also be very good.
Anyway, can't wait to see how Heroes of Ancient Order will turn out to be.

[Steven Aus]

Quantomas, do you have some more news?

[Quantomas]

I wished I had a bit of funding. So many small obstacles and detours that all add up and eat my time. Not much work left though.

[Steven Aus]

Quantomas, how are you going?

[Quantomas]

Mostly all parts of the AI have now been written. The next step is to review and adapt all AI routines to the changes in this development cycle.

Essentially the AI makes an initial guess of the strategic and tactical situation to arrive at a suitable behaviour. Then it can drill down to analyse any aspect it sees fit in more detail. Technically there is no strategic situation the AI cannot handle. But it will take a while to make sure that everything is implemented proper.

Will be interesting to play against the AI. I suspect that the AI might prefer to play defensively because even successful battles can loose you a game. I will shape proper AI personalities to provide different levels of aggressivity to make the game fun. If done right, it will not make it more easy for the human player.

And there is a fair bit of additional work to get everything running.

[Karmakeld]

I've been following this tread for some time, and think I should state my support.
What about listing what you need of help from others? Like artists, programmers etc. and for what tasks? It might be a bit too soon, if you're not entirely done with the AI, but I know that people on HC have been asking about support to create a HoMM like game from scratch, and some also started making ideas for new factions. I don't know if it would be of interest to combine some of those projects? Might speed up some things.

[Panda Tar]

Probably most things will have to be developed in unison with AI, especially due behavior of abilities and units due combination of designed factors created for factions and gameplay.

[Humakt]

Instead of spending too much time to play against the AI, you could make the AI play against another AI. This can help you to test the AI, find bugs and even to test the balance to some extent.

Might make sense to do different testing environments for both adventure map and battles.

Similar to this http://ukkosnetti.fi/tribetron/#/rumble

[bitmaid]

It looks like there is a month or so before they release the expansion along with the big patch including alleged improvement on AI. Currently it's broken for everyone I believe, every creature uses wait and initiative doesn't mean shit, enemy heroes don't skill properly and don't use magic effectively. Somehow I just don't have faith in them anymore. It's like they don't even see that AI is essential to the game. But anyways, enough with the negativity.

It's great to read about your progress here. I look forward to this much more than what the dodoes over there are baking.

[mr.hackcrag]

I don't know why they are too cheap to just pay Q to work on the game as A.I. developer. After all Q has done, he's earned trust to do an excellent job.

[Steven Aus]

Quantomas, how is it going?

[bitmaid]

I was right, they delayed the freaking patch. Now the game will have shitty AI almost one year since release! I agree Ubisoft should totally pay him.

[Quantomas]

Since my last post I hardly got to work on the AI because too many other things demand my time. The situation is far from ideal. It's a pity though because there is not much missing.

But you can bet that I won't give up on the opportunity to have a Heroes game demonstrate what a proper AI can do for the game. It will be an eyeopener.

Thanks for the encouraging words. I also do not understand why Ubisoft and Limbic did not inquire about the AI work I do. Actually, I spoke to the CEO of Limbic in depth about H6, way back before H6's first expansion. So, they definitely have my contact details. As an explanation I can only guess that the management does not recognize the central value a proper AI has for a strategy game and its playing experience and the effort it requires to develop one. It seems a bit that they may have thought along the line that a modder developed an AI, so we can hire any talented one who is qualified. Where I hail from, our hiring process is much more demanding and we would favour applicants who can prove that they have the right skillset and ideally have a track record of a successful project in their line of work. I have all of this.

From what I have heard of H7 (didn't play it myself) it's not only the AI that is lacking, but the state of optimizations and how the different parts of the game are tied together tell me that they were also lacking a good lead programmer. An experienced lead runs the show, so to speak, and organizes the entire development process in a way that all elements fall into their proper places naturally. If you look at a game like Torchlight II it is virtually bug-free. Games like this are complex. The only way to get them free of bugs is a good lead organizing the development in such a way, that the code is robust and and its function verified along the entire track.

Regarding the development of the AI itself, Ubisoft has the source of H3 which contains a working AI. Looking at H7 at its current state, they did not only ignore all the lessons that made H3 and H5 a success, but were also ignoring this valuable source. While it will still take you years to develop an AI, starting from the H3 source code will teach you the right lessons how to develop an AI and get you on the right track. Gus Smestad's AI for H3 is still unsurpassed by any other AI released for a strategy game until now.

That's the cue for us of course. What is shortly before completion is the next generation of AI, which is entirely non-cheating, versatile and reactive to whatever is and happens on the adventure map. It is also easily portable to other games. If Ubisoft is not interested, we will explore licensing it for other games. We are TESLA MINDS LTD now and will set up our studio in Dublin.

[bitmaid]

Some exciting news Quantomas!! What a surprise. I'm so glad that you decided to incorporate and thought beyond. This is the right way, others will definitely benefit from even looking at the library of some proper code. Yes save other games from having shitty AI as well! I'm a bit curious how you were able to fix the AI in H7 without having played it. Granted I don't know much and perhaps AI can be examined independently to a degree especially when you have your product in mind. Either way I look forward to both the fix and your new venture. Good luck!

[Yurian Stonebow]

Dear Quantomas,


Please contact me ASAP. ( I hope you read the PM already?) I may have some interesting news to share with you.

With best regards,

Yurian Stonebow
H3, H4 and H5 map maker and enthusiast