My dream is a Fallout/Arcanum style game with a sci-fi space opera setting. Basically combine Mass Effect with Arcanum and you'd be about there. It's a shame most of the recent "old school" RPG revival games have had typical fantasy settings. The ones off the top of my head that did not, Underrail and Atom, copies the Fallout setting.

I think the difficulty of Fallout 1 was pretty solid. If you made a character archetype... stealthy, charmer, brute, whatever... you should be forced to stick to it. You should also feel the game getting easier as you level, so it should be kinda hard early on, but nothing super frustrating. I definitely demand save anywhere systems, but steps should be made to reduce save scumming like hiding speech check numbers.

Where do you store that unique id?

(The game I make is not going to have perma-death of any form. Even if I do decide to have characters die permanently, the game will not auto-save and will let you reload from before that happened, but even that is not likely.)
Sounds like some influence from Rogue Legacy and Romancing SaGa 2.

To some extent, I feel this way about SaGa games in some ways. There are some aspects of the games I dislike, mostly introduced in Romancing SaGa, like having quests close if you fight too many battles, and having enemies move in real-time outside of combat (as I don't like my RPGs turning into action games when I'm just trying to avoid combat; unfortunately this is the one SaGa element that's spread to other JRPGs).

Encrypted in the save files, auto or user generated, and in a temp file. Each session with perma-death a check is made to determine player state. When the query returns dead the seed is invalidated and save states will be erased. This is how I would do it if I were to introduce it to a game with that mechanic.

My favorite use of time in a game is actually in Zelda: Majora's Mask. Events happen as time passes, and the world will end in 3 days, but you can restart time (and the world) by playing a song. When you restart time, you keep everything major that you acquired (but lose things like rupees (except those deposited in the bank, for whatever reason) and ammunition), but events are reset, which allows you to do things differently.

There's one other Square game that uses time travel to a significant degree, and that's SaGa 3 (Final Fantasy Legend 3), However, it only uses it for the first half of the game, and doesn't do that much with it. It also doesn't really play like a SaGa game, mainly because Kawazu wasn't involved (but he was involved in the DS remake, and it shows, as that game redid the mechanics, replaced XP/levels with a SaGa-like growth system, though the ability to change race was kept).

Incidentally, there's a zero cost game on itch.io called Rxcovery, which combines a time limit with New Game + that lets you keep progression items, as well as SaGa-like growth mechanics. It's not a big game (otherwise it would cost money), and it's not without flaws, but I still found it to be enjoyable.
https://lovegames.itch.io/rxcovery
And then I figure out where the temp file is stored, and copy that as well.

Or, if I *really* feel like it, I reverse engineer the encryption and figure out how to modify save files. However, that would take a lot of work and is something I would only do if I really like the game. (Sometimes, cheating can take more time/effort than playing the game normally!)

I actually disagree with this. Unless the game is super short (like SaGa 1 length or shorter), I would prefer to be able to try out other character archetypes without having to start the entire game over.

I don't really agree with doing this directly. (Also, I dislike the term "save scumming".)

I think the best approach is, rather than limiting saves or reloads in some way, it would be better to remove the incentives for doing so. Avoid having situations where bad things happen that can't be easily recovered from them, and on party wipe have the party respawn in a safe spot, without losing anything since the death. Also, avoid non-respawning random treasure. (Some games have non-respawning random treasure, which really does irritate me; Wizardry 8 and the SaGa 2 remake are two examples of this.)

The game could be designed (for hardcore/permadeath) that the player can't explicitly save. Saves would only be made when the player quits the game -- in which case, the save would be of the player's immediate state. Yeah, you could still cheese the system a bit with that, but it would at least minimize "I'll save, then drink the unknown potion, then reload if I don't like the results" situations.

And I say "minimize" because if someone is really intent on cheesing the system, they could simply copy the save/config files before starting the game, then quit and restore if they didn't like the outcome of that session (and they could potentially do so while the game is running if I used good programming practices and only left files open when I needed to directly read/write to them). If someone is really intent on cheating in their local single-player game, I see no reason as a developer to go out of my way to prevent that.

If someone wants to play "perma-death"... but then writes a utility that constantly backs up their save state to prevent perma-death... well, that kinda defeats the whole purpose of perma-death and is exactly why I would make perma-death an optional toggle. But if someone REALLY wants to go that way... umm... well, "you do you", I guess.

Personally, I'd be more interested in 2D overhead view (like those with tile graphics) RPGs; I don't really like the isometric perspective.

The original Dragon Quest is a good example of such an open world RPG, as are the Ultima games up through 5 (and even 6 at least still lets you follow paths in a straight line, unlike most isometric games where you need to move diagonally, which is a pain).

What about games where the world isn't contiguous, but where the player is given freedom? You see this in the Romancing SaGa series, where you just choose a location on the map to go when you leave an area, and in SaGa Frontier, where the regions feel like separate planets and the game provides no indication of where the regions are in relation to one another.

From the few roguelikes I've played, I've decided that I really don't like having consumables not be identified.

My answer to this varies every time someone asks. My current mood:

Culdcept Saga has a baby with [Slay the Spire/Monster Train].
Deck building "Monopoly area control" game.

Like STS/MT, I'd be hypothetically a solvable game where victory can come easy with the right decisions.

Yep, I agree -- unless (1) you can save-scum (which I **REALLY, REALLY** don't like, so I actually don't consider this a viable "solution") or (2) Identify scrolls/spells/doodads are very plentiful and/or shopkeepers can identify items for you and shopkeepers are reasonably accessible (i.e. Diablo style: travel back to town, talk to the "Stay a while and listen!" dude, identify your items).

Option 2 intrigues me, in that it presents another strategic element in that you have to consider which items you will identify immediately versus which you will save to identify later. So, if you only have one "Identify Scroll" and already have a magic suit of armor but are using a mundane weapon, then you might put off identifying another suit of armor and instead identify a weapon. Again, assuming that you had a *reasonable* number of Identify opportunities through consumables or vendors... yeah, I could run with that kind of system.

*builder
*simcity, Majesty 2/1
* Leading a town in a fantasy world.... dnd also somewhat of a inspiration, others told me there was a very early try at it
* It looked pretty nice in a dream, trying to keep your village/city going through a period of increasing unrest. Different guilds of course such as the mage guild and the thieving guild, i guess the best description would be ankh morpork. The whole idea of such a lively background just coming alive in front of your eyes. How would the thieving guild come into being, the merchant ones first of course ... or the sowers one ;)
* difficulty, the same as any builder with lots of untold stats etc

Some people want to do everything in one playthrough, yeah. I just don't think that's very RPG-ish. Your character's choices and role should open some opportunities and close others, that's what makes these choices interesting. Even if you never replay games, it still tailors the game to your individual playthrough.

The problem with consumables is that you can't identify something through experimentation and still be able to use them later.

(Gameplay spoilers for Shiren the Wanderer)
One good example of identifying items by experimentation that's done well happens with Shiren the Wanderer's pots. For those unfamiliar with the game, a pot is an item that you can put other items (but not other pots) into, and which will (presumably) store the item; to get the items out, however, you need to throw the pot at the wall, which will destroy the pot. In any case, identifying a pot by usage works something like this (put an item into the pot to start, then check the contents):
* Get the option to rub the pot rather than open it? It's a healing or monster pot. Use it when it's safe to do so to distinguish the two. Or it could be a pot that sucks in an item from a distance, which does have some uses.
* Game lets you select an item in the pot and take it out or use it? I forgot the name of this pot, but it's the one that lets you take items out of it without breaking it. (In a game with limited inventory space, it's quite useful.)
* Item disappeared? It's either a storage pot (items in it get sent to storage, where they can be later retrieved, even after a death) or a bottomless pot (item is gone forever; throwing it at a wall will create a pit, which can be useful if you're a thief).
* See a different item in the pot? It's either a Change pot (items in it are transformed) or a Liar pot (item lies about its contents, and may be different if you check the pot again). In this case, put some useless items into the pot, throw it, and see if anything interesting dropped out of it.
* See two of the item? Congratulations: You found the rare duplication pot! Yes, there's a legitimate way to duplicate items. (Unfortunately removed in the DS version.)
* Nothing happens? There are some possibilities still left. Some pots identify their contents. There are also unbreakable pots; throw one at a wall and it won't break.

Notice how these methods typically only use one charge of the item (few exceptions, like using a monster pot that you don't want to keep anyway), and you still have more charges to use? That's why the mechanic works fine here, whereas it doesn't work so well for herbs (which are identified in the game's 99 floor dungeon; said 99 floor dungeon has basically become a standard feature of Japanese roguelikes (at least the ones decended from Torneko/Shiren).)

Having opportunities opened and closed like that is IMO good if the game is short, so one can experience a character fully in a short amount of time and then replay with a different character to see other paths. I don't think it works in 20+ hour games, however.

Well, depends on the exact question here.

If I was to make a game, I'd have another go at using RPG Maker or something similar and not digging deep into scripting or custom assets, even if I'm not keen on what that will result in. Sort of trying to get some story in despite disliking the gameplay I'd need to make do with.

If I was to magically learn some programming and also have the assets, as something I could probably handle myself to get a feel of what it's all about might be some Castle of the Winds style dungeon crawler, icons for everything, that system where everything moves once the player does, but with handcrafted maps, no procedural generation, and a lot more complexity, lots of kinds of enemies and skills and gear and complex calculations for effects.

What I'd want made for me, by a team of skilled developers, on the other hand, is... Didn't I already say this in another topic of yours? Anyway, it's the game based on the story I had for the books I meant to write, and gave up on after several years of struggling with the first. Story's still there in my mind with no outlet, and it'd suit a game quite nicely... Well, more like a series of games, because it'd be quite impossibly huge for one game and also require entirely different genres, since the gameplay is dictated by the story.
So it'd start with a sort of tactical/stealth RPG as the characters need to run away and escape the island, since there's no way they'd survive on it with all the might of those out to get them, so difficulty will be quite high and you'll need to pick your battles, fight too much and you'll get worn down, but evade too much without killing some of those after you and too many will gather and become impossible to avoid or defeat when they surround you, and there's also the fact that a couple of characters are basically useless in combat and need to be protected. Then the sea voyage would be more of an adventure game with some survival elements, there are some battles on one other island they'll land on for a while but other than that it'd mainly be exploration and making the right choices to reach the mainland alive and before the situation gets too bad, and difficulty will be rather low, the best outcome may be harder to obtain but a real loss state would require quite a string of bad decisions. Then open world RPG, a heck of a lot of exploration, little handholding or even direction really, it should feel overwhelming and confusing, but the difficulty will depend on how well the player makes use of the characters, because by then two would feel like demigods in terms of power and a few others would be outstanding on their niche as well, but the challenges faced would also be at that level, think basically the whole world having evolved to prevent people from returning to it, and you're doing just that, and trying to survive enough to prove you'll be different, and figure out just how one would go about doing that. Then there's the return to the island, and a mix of adventure and stealth/tactical RPG, gathering allies.
The above may fit into one game, but then things switch to battles between armies, for which I'd go TBS since I can't deal with RTS, and I don't really think these two parts would fit in a single game. There'd still be RPG portions, so it'd be like one of those hybrids where you have a strategy part with armies and a RPG part with a party of heroes. But with this being so far off in the story too, there are a lot of details I never worked out even for the books, so can't say too much more about it.

Personally, my game design interests tend more to gameplay than to story, so I would prefer to code the combat mechanics and such by myself so I can do interesting things there.

Playing Rxcovery made me aware of how the RPG maker engine is somewhat opinionated and has to be coaxed into doing things it wasn't meant to. For example, on the save/load screen, you can see the levels of your characters, which are always 1 due to the game not using level/XP based growth. Another is seeing current HP on the non-combat status screen, even though the game fully heals you after each battle.

Those are all good games, although I personally prefer the isometric style WRPGs with higher-resolution artwork.

I don't dislike those games at all. Baldur's Gate 2 is like that and is my favorite RPG. It seems a little strange to me though that Ultima VII is considered to be so influential, but so few isometric WRPGs that followed tried to replicate its open world model. For some reason, it seems to have had more influence on 3D open world games, like Elder Scrolls, etc.

I haven't tried any of the SaGa games, but they sound interesting. It looks like they're not available on GOG though?