Been there done that with E:WOM did not turn out pretty.  We must let the DEVs know what we would like to see in game. This particular thread is about Seigecraft in FE hence why we are talking about it.

I understand what you're getting at, but I also think features are important on their own. If there's a dev journal on magic, and they proudly present FIVE different schools of magic - one for every type of material you can create with it (gildar, materials, metal, mounts and elementium). And that's it. Wouldn't you be worried then?

But again, we know very little about combat since they haven't talked about it yet. So while not having sieges is certainly a downer, I'm eager to find out what we're getting instead, combat-wise.

I solemnly swear to deliver a PC game with siege combat, fortifications and in-depth unit customization / progression within the next 24 months.

Your both right and i understand that. I also understand the disappointment about not having siegebattles. Siegebattles are cool in AoW and HoMM. Its only that i don´t think the game couldn´t be very good without them.:)

It was not meant as an offend. Its only... i read so many theoretical discussions in so many threads here on the board and always think "How could they know all this. Its all theoretical. How could they know this would be fun ingame." Sometimes i don´t understand the gimmickry with figures not even half. Thats been said.

  Instead of trying to reinvent the wheel with all those questions, I suggest examining games from the past which do have this working in a fun way.  Play a few games of Age of Wonders:Shadow Magic and you'll not only have fun, but you'll learn what works.  The developers from Triumph Studios have always been friendly... you should try contacting them with your AI questions.

Triumph Studios are above average at making games but A.I is NOT their strength. I remember you from the AoW Heavengames forums. You're a veteran of AoW just like me and then you know how bad the AoW A.I has always been.

The AoW AI was pretty good compared to the Elemental AI.

Absolutely. And the food I made today and left in the oven will still be better in two days than the rotten tomato I saw yesterday

I absolutely demand walls, catapults, balistae, mangonels, ladders, burning pitch. Do you hear me? I demand it!!!

For WoM, the excuse was "the engine can't do it"; for FE, it becomes "the AI programmers can't do it".

If Stardock can't duplicate the tactical AI of a game 12 year old, despite cutting more and more features that it had (ships, flying units, dungeons, underground plane, differentiated factions, and now siege mechanisms), it's time to throw the towel, guys!

The only valid excuse they can come up with, is that Brad (the only A.I coder) HAD to do CEO stuff. But even then, that just doesn't cut it. So I agree with you.

But I think we're worrying unnecessarily. They've said that they WON'T release until they're ready which includes the A.I. Then again they're on a budget so....but in the end, I'll trust Brads & Dereks word.

Oh, Really?
How about going back 22 years to find a game that easily handled Line of Sight issues and *gasp* the AI handled their own Ranged and Magic-capable units fairly decently by keeping them still and plunking away at the player's forces instead of advancing them into the meat grinder.
What is even more mind-boggling was that this was all done using Microsoft QuickBasic as the main programming language!!!  *Double gasp*
Sword of Aragon by SSI.

Granted...incredibly cheesy graphics (Does anyone here even remember playing EGA graphic games? But LOS was handled seemlessly, along with customizeable, promotable, and upgradeable units.  (But don't get me started about the copy protection system!  *shudder*)  Most of that engine was used as the basis for Fantasy General (and Panzer General) later on, by the way.

Oh yeah.  Did I mention that it was programmed with QuickBasic over 22 years ago?     It can't possible be that hard to program and have the AI handle it well. Now. 
I agree that Sieges and Castle walls are an entirely different matter.  Entire games have been made (and few very well) that *just* handle tactical siege battles. These games were made with much more budget and staff time than Stardock apparantly has available in their entire games division, so I'm under no illusions why this aspect will not be encorporated into F.E.
Oh yeah... Did I mention SOA was made using QuickBasic?

Not only do I remember playing EGA games, I remember when 'Enhanced' (the 'E' in EGA) was way cooler than the games I'd played.  Heck, I remember pre-C(olor)GA.  God, I'm old.

Fantasy General was a great game.

Just because it was done before doesn't automatically make it any easier now.  I'm pretty sure that Stardock started from square 1.  Also, while it may be fun to make fun of Basic, it was... adequate enough.  C++ isn't going to magically make it easy or quick to code.  That said, they still need to get it done; the tactical AI really needs to be at least decent for FE to ever be considered a good game.

tanafres, you are a youngster, as I was attempting to WRITE text based graphical games in 1978/9/80, ie used a text display to represent graphics in BLACK and green(white), and even remember seeing the AMBER screens, and thinking how cool they were, as most of the screens WERE black and white with the expensive ones having GREEN phospher CRT and being more expensive, while the amber were even MORE expensive again.

the ancient harpo

Ok, you've got me. I bow before your antiquity experience, sir!

:karma:

I disagree.  QuickBasic had absolutely zero graphical manipulation commands. The game had to use far too much inefficient code to accomplish what it did, and actually ran fairly slowly on the hardware of the day (8088?). On the other hand, C++ has commands and vast classes for graphic manipulation already developed and available that can be optimized for speed and visual effects.

But I do agree that Stardock probably has to start from square one if they are irrevocably tied to their game engine, and that engine simply does (as I suspect) not lend itself easily to a tactical battlefield with elevations and LOS issues. (for some reason?).  I'm simply saying that the algorithms to handle LOS issues for both the game engine and the AI behaviour with it should be vastly easier to code than having to incorporate all the intricacies of siege warfare into a game.  The Paxton quote above seems to reverse this claim, and I don't see how that is possible.

Which doesn't matter a whole lot when it comes to AI. It's a turn based game, the AI has time to figure some stuff out. The quality of the algorithm is more important then the speed, and basic does have some advantages in terms of getting out of the way compared to C++.

C++ is certainly better for something like the graphics because speed is key there. But there's a reason a lot of corporate apps use languages like Java or C# instead: they take stuff that C++ makes you deal with and removes it so you can focus on business logic (or AI in this case). I mean if you wrote the AI in VB.net and the AI was good, nobody would notice or care what it was written in.

As Tridus mentioned, we were talking about AI, not the graphics.  My point was that while similar logic has been written before, since the AI is very engine-specific in what it has to deal with, my impression is that you still start more or less from scratch each time with a new engine (or your first game on a licensed engine).  So to duplicate the AI of that game, would take about as much time from a standing start, since they probably only had 1 AI coder, and FE seems to have 1 also.  The developments that have advanced other areas seem not to apply well to the AI.

As for why he brought up line of sight and such for siege warfare, it seems like for walls and sieges to be meaningful you'd likely want those sorts of features. Personally I was OK without (MoM), but I can see why adding it would create a deeper experience.  It does worry me though, as it implies that the actual battlefield in FE will have little or no effect on tactics, only strategy (neat-o unit producing buildings, for example).

Can we expect other areas of focus to receive similar treatment?

Speed does matter for the AI, as making the player wait is considered a bad thing (even though that would allow for better AI) these days and is therefore avoided.

"Parts" of the AI can be reused if written properly; however that requires more work (the big diffrence in writing code in C++ and Basic as I know it is C++ focuses more on making it possible to reuse old code rather than make it easy to write new code*). However since AI has to wait till the game is pretty much finished I would suppose that making the code reusable is of low priority compared to finishing it on time and having it work.

* Or so it is intended. Oppinions may vary on which is easier to write in.

Unless the object model for the code is completely screwed, coding the AI is a completely different task (mainly maths) to showing the result on screen (the job of the engine).

And we are dealing with a really simple grid with extremely limited parameters, and units with similarly really simple parameters. The math to work with these objects should not be as difficult now as it was before object oriented programming came in.

It sounds more like the engine doesn't have a usable API yet - which would be what you would expect from something that is being used for the first time.

Honestly, it doesn't. Speed matters in the graphics engine because a 2FPS game is unplayable. If the AI is good, people will tolerate waiting for it until a patch later has some optimization. It's far more important that the AI is intelligent then that it's fast. A great, slow AI is great. A bad, fast AI is bad. Speed is a requirement in an RTS because the AI has to react instantly to whatever the player does. It's a "nice to have" in a TBS.

Secondly, the task is straightforward enough on computers today that the bare metal access C++ gives you doesn't matter. You could do it in Python and get perfectly acceptable speed out of it.

At the end of the day the problem with the WoM AI is that it's not very good, not that it's slow. If a higher level language trades some speed in exchange for more developer productivity it's a no-brainer to go with that because that's what we really need.

Gameplay and good AI>Speed. Within reason of course.

Why even ask? I mean, I think it's nice that you can stand up and say that you want it.. but Elemental has been pretty cautious towards throwback features pretty much since it was announced. The series (seems to be) designed for a more casual audience, it's not really intense, not real deep, not very difficult.. comes with a construction set, friendly graphics, and a reasonably tame story. Not much of that looks to be changing with FE, and I doubt stapling on walled siege battles would make the experience more engaging.

I dont mean that to sound overly critical, pretty much everyone at Stardock seems decent, good at what they do.. I like the company, I've got a few of their products. Elemental is probably the only bone I've ever really had to pick with them and that makes for a pretty good track record in my opinion. It's just not designed for people like me, or.. really, fans of any older TBS. I think you're bashing your head against a wall with a thread like this.