@Winni - thanks for the link on MOM. I bought it from GOG a month or so ago, had no idea there was a cool update like that. Will check it out.

It does. It sounds like the classical IT problem with so many prjects: Too many hats, too little time. I keep hitting my boss with that ( its part of my job as a lead engineer to do that)...

I really really really hope you make it with FE. Nothing would please me more than to see a true classic come forth. Good luck!

Based on what I have read from Frogboy's previous posts, that's what they tried to do. However, being an engine in the progress of being built, it's capabilities ended up being changing... In fact, I think Frogboy ended up taking some sort of emergency leadership position because of this.

Point being, it's hard to judge the capabilites of an engine until you've tried to use it to create a game already.

Sorry to say this guys but I don't think you understand what i was getting at here.  If I design a board game that calls for 12 pieces per side but it turns out I only can have 8 pieces per side due to manufacturing costs of the board game, I have to alter my design.

Every game, regardless of how it manifests itself (board game, video game, card game) makes use of the resources at hand.  Try playing Texas Holdem if some of the cards in the deck have been lost. You could either not play the game or try to alter the rules of the game. 

The WOM original design was dramatically different from the final design (the tactical battles being the most obvious example) because of what the underlying game engine turned out able to do.

Even today, we struggle with limitations (how long have we been battling memory limitations for example) that require us to make changes.  I originally wanted to see battles involving tens of thousands of units.  But we also wanted to have units that players designed.  Only one of those things was possible to do so we had to dramatically lower the number of units on screen and the size of armies and such.  The list is nearly endless on things like that.

With Galactic Civilizations, we knew what we could do because the engine was 7 years old by the time we made it (Pear was the engine and we used it for Entrepreneur, The Corporate Machine, LightWeight Ninja, The Political Machine and other games).  

It's easy to make facile analogies like square pegs and round holes but that presumes that it is possible to simply make a carve what you want. 

I am going to be honest with you Brad but I would have rather have had the option of thousands of units on the battlefield than the option of customizing units. In my opinion, having the ability to customize units obviously killed that idea of having thousands of units but it also killed the individuality of factions. If you had several unit types for each faction rather than customizing units, each faction could have been more unique and you could have kept in those thousands of units running around. I think people would have been much happier that way.

I think this is an important point. If the units for each faction were pre-designed and the variable squad sizes had been dropped then the game would have been much easier to balance and each faction would have felt unique.

I also believe that AI is the least of the issues that need addressing in Tactical Combat. If the system isn't redesigned to allow units to be balanced then it is always going to be an up-hill battle. Here is an earlier post about this, and while there was disagreement on the specifics of how combat should be approached there was consensus that the system needs work to allow balance.

You put on a good point- one of the areas I feel SD games could use improvement in as they sometimes they feel mechanical.  I think Derek will cure that, given his history with FFH though, so I haven't said much on thaat.

I'm just hoping the AI gets fixed for FE at some point, as that has been a crippling weakness to me when it came to WOM, and what made me give up on the game.  I'm still of the belief a strategy game can only be as good as its AI.

In software, you don't control all the variables.  You're constantly making choices that have both good and bad consequences.

Let me give you a couple of examples:

Early on, we wanted the game to be able to run on Netbooks. That meant picking Pixel Shader 2 as the means we would interact with DirectX. But it turned out that Pixel Shader 2 had a lot of limitations that we had not counted on (we had never made a strategy game with "organics"). 

This, in turn, had another consequence that was totally unforeseen -- newer video cards don't handle the older Pixel Shader 2 very efficiently in terms of their drivers. This is the biggest source of "memory leaks" that really came and was pretty crushing.

Now, all of these things could have been addressed with sufficient time. And that is where the key error in judgment on my part came into play. I believed I could re-design the game at the 11th hour based on what the engine could do (this is back in May 2010) and still have it come out really good rather than stick with the original design and add 18 months (the estimate on how long it would take including the abandonment of pixel shader 2).

With Fallen Enchantress, one of the first things we did is drop support for Pixel Shader 2. So no old netbook support. A lot of "bugginess" disappeared overnight just with that.  It's almost tempting to retrofit that into WOM but there'd be riots (because N% of users could simply no longer play the game) so instead we continue to find workarounds (if you want to see a pixel shader 2 driver bug in WOM, look for a case where the portrait becomes all black, if you have an older video card, that won't happen but if you have a "Good" video card it'll happen occasionally).

Anyway, to summarize, when you make an engine, you're really creating middle where. There is a whole world underneath that engine (DirectX, video drivers, the OS itself, and of course the hardware) and the interactions of those aren't always -- or even usually -- what you'd anticipate.

Any chance for a variation of these tactical battle Sovereign AI enhancements in the next WOM update?

• AI sovereign will cast wind shield if the opposing army has archer units and ranged attacks are projected to cause notable damage.
• AI sovereign will cast magic shield to protect himself from enemy magic if enemy army contains sovereign or magic casting champion.
• AI sovereign will cast Spell Blast or Flaming Arrow
• AI Sovereign will cast Guardian Aura to improve dodge rating of targeted friendly unit

Question- how hard would it be to drop support for 32-bit OS's and make a future Elemental game a 64-bit application?  I think that would drop most of the memory issues, at least until the very long term.

I know this isn't practical for FE, and am not asking for that, but would it be practical for the 2nd expansion, both technically and in a business sense?  I think the amount of "free copies" that are being given out in a way, gives you guys some freedom to take some experimental approaches if FE does well, and I can't see GalCiv3 not being a 64-bit application (Assuming that's next) when you get around to doing that in 3-4 years.

Not if he just puts in tags and weight measurements.

It's good to see you back, Brad

When Winnihym said "I know you don't mean what you said in your post, or at least I hope you don't" my feeling is he was referring to this sentence in particular.

I'm neck deep in the software industry so I entrely get what you say about the engine and its limitations. But I think a strong case can be made that the fundamental problem with WoM is that it is just a very very thin layer over the engine. And I don't just mean when it first came out, I mean now too.

The key point being that the game part needs to have a great deal of depth to it, both conceptual and implementation. That means an overarching design which includes limited competing resources which are applied to try to achieve multiple competing goals with many meaningful and hopefully interesting (read fun!) choices along the way.WoM has a very disjointed resource model, there are many resources but most of the time only one (or none) of them is limiting and there are few meaningful choices in managing resources. Conversely on the goal/choice side there are few meaningful choices because certain strategies are so much more powerful than the alternatives that you would only choice the suboptimal choice if you are trying to gimp yourself. And of course the lack of depth and fun factor in the available fluff is a topic covered in many other threads.

FE clearly has a more overarching design so I am cautiously hopeful that it will be decent. But I fear they are not going nearly deep enough at balancing these core mechanics. I am confident it is a difficult thing to get right BUT it is also the sort of thing that distinguishes mediocre games from brilliant gems.

BTW for what it is worth I feel that this is where they went wrong with Civ 5 compared with Civ 4. I find myself having to make much more difficult decisions about resource allocation (primarily trade and production, but also settling and improvements) in Civ 4 than in Civ 5. This is because in Civ 5 your gold income/happiness/research/culture rates are largely preset based on choices you have already made whereas Civ 4 you have that but also the big lever of how much of your trade you allocate to gold/research/culture. Similarly in Civ 4 almost any building or unit is semi-viable to build at any time (depending on your overall and regional goals) while in Civ 5 due to the much higher maintenance costs of both buildings and units there is almost always just one or two obvious choices. These things don't seem like much but they make all the difference (of course there is also the core problem that the Civ 5 unit AI just can't run a decent war on the hex grid but that is a separate discussion!)