Trilarion: The 100+ bashes you need to get a door open is surely a design flaw [...] although I am often surprised that other people may see it completely different
For the walking part, I've heard that was a very intentional design decision on the part of the developer. I would not be surprised if this "fully interactive bashing" was as well (maybe to "reward" players who get high lockpick or lockmelt skills by punishing everyone else). My philosophy with RPGs is that the player should be aware of the time and effort the PC is spending, but the game should not require the player to directly experience it. I believe a big part of the problem is that other consequences which should be present are not modeled in the game, and their presence would be adequate motivation without making the player artificially suffer. For example, if time actually mattered in the game (e.g., events play out over game-time whether the PC is active or twiddling their thumbs), those who learn the portal spell would have a real in-game advantage (at least when it came to traveling) over those who walk everywhere. If bashing doors attracted unwanted attention (baddies coming at you all at once instead of in more manageable numbers), or the game-world (not real-world) time spent bashing meant sentries would find you and sound the alarm, or the weapon repair skill wasn't so over-the-top effective (able to repair your bashing weapon to 100% every time), those who learn lockpick/lockmelt to high levels would then have real in-game advantages over those who bash all/many of the locked doors.
Trilarion: making your own design better without introducing other flaws... is not so easy
I think identifying general philosophies and rules (such as the one indicated in my above paragraph) and sticking to them may help with this.
Trilarion: I helps already thinking about possible implementations problems (path finding, AI, ...). Not all game elements one can think of are easy to implement.
I have identified AI as one my top two "hard problems" to deal with. (Efficiency is the other simply because I am going for a full "living world" system, and want it to be very large -- I don't know that I'll be able to match the
62,000+ square miles Daggerfall provides, but that's the kind of scale I'm shooting for as I want to support "real", directed adventuring rather than having yet another "explore every nook and cranny of the whole tiny world" game.) For AI I plan on using a generic approach -- I will not be hard-coding AI behaviors into the NPCs (at least for the most part). Instead there will be a machine-readable description of the game-rules to which the AIs have access (partial access for some things as they will need to be discovered), and the AIs will have certain built-in goals (varying from one AI to another -- things like survive, get rich, have kids, be famous, have some adventure, etc.). (There may also be some other personal traits like moral disposition mixed in.) Between the rules, goals, and the AI's access to its internal "user interface" will be a generic AI system based to some degree on the work of Douglas R. Hofstadter and his grad students. (If you're interested, check out the book "Fluid Concepts and Creative Analogies". What they are doing makes a lot of sense to me as I see similar processes in my own brain. They've implemented their ideas in various contexts, and the results are pretty impressive even though they've only been tried in limited domains. An RPG with combat, trading and actual communication (actual information transfer) occurring both between NPCs and PC(s) and amongst NPCs, production and shipping of goods, theft, guards, government, etc. will make for a very interesting large-scale test case for their ideas.)
Trilarion: Trying to put a turn based strategy game together currently.
Not that you necessarily care (especially if you're focused on making the game you want to play rather than what others want), but... the two "key features" for me when it comes to TBS games:
1. have random maps
2. have a decent AI that can handle random maps (with limited or no cheating -- if it cheats, let the player control specifically in what ways it's allowed to cheat and how much)
(I also have more fun building things up than tearing them down, but that probably varies greatly from one player to another. And I generally don't care for campaigns, except Warlords Battlecry 2, but that's because its "campaign" wasn't some excuse-for-bad-storytelling-and-railroading/linearity but rather was an additional layer of strategy added on top -- I really wish more games would do that. If you haven't played it yet, I'd humbly suggest it's time for you to do some more "research"!)
(And just in case anyone still thinks I might be sane:) One of my goals for my RPG is for it to be kind of a "Swiss army game". Not only will someone be able to play it as an RPG, they'll be able to play it to some extent as other genres as well. Become a general in an army, and you may end up commanding troops like a strategy game (though direct "mind linking" to units may or may not be available). Become a (somewhat tyrannical) mayor of a city and start directing construction like a city-builder game. Buy some wagons and ships and hire some drivers and captains and start building a trading empire like a trading game. Or maybe just buy one ship and hire some scoundrels and start pirating your way to riches. (And of course the NPCs will also be engaging in these and other activities.)
