I don't think giving horses to axefighters would add much to the game. It would be the only case in the game where a unit's type can be changed, and I think players would then want to be able to change any unit type (why not if you can give axefighters horses already?). It may also upset multiplayer balance significantly.
Of course it probably wouldn't do harm to the game and any issues arising from it could be resolved one way or another. However I don't think it really adds much so I don't see why it would be worth the cost of programming it. A feature like that would probably take 6 hours of programming then maybe another 6 hours later debugging it after it's been tested and found to have problems (almost any new feature you add can cause a bug, even if it's in a different part of the game). It would add more weight and complexity to our code potentially making it harder to maintain in the future. There's lots of unanswered questions when you start programming something like this, for example, when the first axefighter gets his horse, he can't stay linked to the remaining axefighters because that violates game rules that we rely on. So he will need to split off, but then the player will end up with 12 individual scouts. So we need some way for them to link together, but we don't want them to be linking to the wrong group if there's another group of scouts nearby. All of this just made the job bigger and more complex, requiring more coding and more potential bugs. There will be lots more unanswered questions like this when you start programming it. You also have to make sure it works properly under rare unusual cases, like if the unit is fighting when he gets his horse, does he change unit type half way through the fighting animation? The game currently cannot handle that since all units have different length fighting animations, so now we need to go and rewrite the fighting action to handle this new feature too....
In short, adding new features is expensive for us (time to implement, time to maintain, added complexity to our code, etc.). Unless the feature really seems like it would add something great to the game then I often feel like it's just not worthwhile. It's very easy to say "implement this", "implement that" but in reality implementing it is very complicated and will have consequences way into the future in terms of fixing bugs arising from it and maintaining more complex code.
If we had implemented every little feature like this one that has been suggested on this thread then the Remake would be much more buggy and harder for us to maintain, and have lots of little features that make the game seem contradictory and even more overwhelming to new players.
So please forgive us if we are hesitant when it comes to new features. If you were the one responsible for implementing and maintaining them, you might understand why
It's kind of hard to believe that when the remake's AI intelligence is uncomparable to the original game, and the SP compaigns are significantly easier than the original game. (No offense to the developers...they know where I stand in appreciation for their remake).
We don't intend for the AI to be incomparably less intelligent, but the problem with comparing to the original game is that we don't know how it works because we don't have the source code (and even if we did it would be a lot of effort to read and understand it). All we can do is set up tests then watch TPR and see what happens. This is extremely time consuming and often doesn't yield any useful results.
It's very hard to watch the AI in TPR and try to write something that behaves like it. If someone can write down step by step instructions of how the AI should work under different conditions this would be very helpful, but I think you'll find it's very difficult to convert the behaviour into step by step instructions, let alone code.
There is also the problem of features that are mostly an improvement in the Remake, like the AI pursuing the enemy (e.g. your bowmen you shot them with) without stopping and waiting every 10 tiles. That undoubtedly makes the AI look smarter, but it can have unintended problems like being able to draw out the AI's army too easily. But we don't want to change this behaviour back to like it was in TPR since it would require duplicating code (currently it's like the AI right clicked on your unit so it's the same code used for human players), and it would make the game look a lot dumber with the AI stopping all the time. How do we solve cases like this?
We want to improve the campaigns, hence we have got pawel to test and fix the glaring bugs for us like AIs attacking at the wrong time/place, colours being wrong, etc. But many of the issues cannot be fixed by editing the campaigns. Our AI behaves very differently to TPR and fixing this is not easy. But that doesn't mean we don't want to fix it. So PLEASE don't imply that we made the AI less intelligent and the campaigns easier on purpose.
There's only two of us developing and just making the game engine work is a massive task in itself. Maintaining singleplayer, multiplayer, testing campaigns, testing/choosing multiplayer maps, maintaining/improving multiplayer balance, fixing bugs, organising nightly builds/testing, improving the map editor, it's just too much work for us to do all of it all 100% perfectly. We need help from the community, like with pawel fixing the campaigns. Remember this is a hobby project that we don't get paid for and we both have very busy lives to live (in fact I'm writing this before I go to work when I should be packing my lunch).
Seriously, if you can make a list of how the AI's behaviour is different in the Remake and what features are missing then it may help us in solving these issues and making the campaigns as intelligent and challenging as the original game. Or just describe what features the AI needs in order to be smarter and more challenging. It will probably be a lot of work to implement but if we at least have an idea of what needs to be done it will help.
