EXE patch ideas

[bvanevery]

It's a good idea.  I think assigning a weight for any possible outcome, would be the right way to implement it.  Add up all the weights, divide by the number of possible results, roll the RNG.  No idea what it takes to code or patch that.

General problem with new .txt capabilities is interfering with alphax.txt.  Best policy would be to create some completely new modpatch.txt or some such name for completely new capabilities.  That way, when a binary patch is installed, it doesn't step all over alphax.txt and people don't have to make "Yitzi versions" of their mods, for instance.

[Vidsek]

  Fine idea Dewbacca!  That was exactly what I suggested to Yitzi in his last (active) patch proposal thread.  Well, I was thinking of being able to assign percentages, but three bins of none, some, lots, would be just fine.

  And Bvanevery, an excellent suggestion to have all the additions to alphax.txt shifted over to a separate text file so we have the mods/new features collected in one obvious place.

  Plus, all of them should have either an easy (probably default) disable setting, or a clearly labeled setting that is identical to the original.  The player could pick and choose among them and customize single player games and scenarios conveniently.

  There might occur issues with certain new features being incompatible with certain others, but a little feedback and documentation could give warning.

  Oh, say, I've forgotten where Yitzi's explanations of his work wound up.  Is it over in the WIKI perhaps?  I'll go snoop around.

[bvanevery]

He had some stuff in some files with his patch.  I don't think there's a consistent clearinghouse of such stuff.  I think that's been cited as one of the ongoing problems of maintenance.  Awhile back, I declined to become the "Yitzi maintainer / disciplinarian" because I didn't see how my own concerns and priorities would benefit from the effort.  I admire the amount of work he did, but he mostly didn't work on things that I thought were critical.

[PvtHudson]

Today, when my precious independent Unity Chopper once again was lost trying to intercept enemy worthless trivial missile bomber, I once again remembered discussion in one of Yitzi's "Vote for feature" threads about change to interception behavior. Wouldn't it be reasonable that interceptors do interception runs only when put on alert (Shift+L)?

[bvanevery]

Not really because then it becomes a permanent bureaucratic task for the player.  Instead of building an Interceptor that performs the job of well, you know, interception, you'd have to build it and activate it.  Dumb, tedious, bad design in a game that already has too much to do.

This is basically an "opt in" vs. "opt out" debate.  I think it would be perfectly reasonable to have an opt out available.  Can't really use "H" or "L" because those already have legitimate uses for expected gameplay.  Perhaps "M" for Mothball?  Maybe that's too obscure, but that's the concept.  A unit you don't want to have do anything, not even fight unless it's the very last thing in the base.

[PvtHudson]

What's a rationale behind feature "open diplomatic conversation (not first contact) on units' touch"? It not only results in clicky and gamey behavior for a player not wanting to talk to AI, but can put AI at a disadvantage. If such contact occurs near the end of AI's turn and results in war, the other player gets a first strike opportunity, which obviously should be an aggressor's luxury.

[bvanevery]

What's a rationale behind feature "open diplomatic conversation (not first contact) on units' touch"?

I reread this thread and didn't see such a proposal anywhere.  Assuming someone proposed it somewhere sometime, I assume it means that upon physical contact, the other faction is forced to talk.  Politely declining to speak to you would not be possible.

This is a bad idea in 2 ways.  First is that when any substantial amount of mutual movement in an area occurs, you'll be talking all the freaking time!  That's gonna get old real quick.

Second is that factions have legitimate reasons to avoid talking to you.  Such as not wanting to pay you, not wanting to be sucked into wars by you, not wanting you to beg them for technologies, not wanting to be threatened by you, etc.  If you do various things without observing the diplomatic niceties, your Reputation suffers.  So other factions endeavor to have your Reputation suffer.  They deny you the avenue by which to perform a diplomatically legitimate action.  They want you to be seen as an aggressor that cannot be trusted.

Incidentally, I have found that there's no clear pattern to when an AI will start talking to you.  Bringing a threatening military unit into contact, or a probe team, does not automatically trigger a conversation on the AI's part.  I used to think that probe teams had a higher chance of this, when say you're in a Truce with a faction.  That the AI is trying to talk its way out of being probed, just before you do the probe.  But I've seen plenty of instances where my probe just did whatever it was going to do, no AI conversation triggered.

[PvtHudson]

What's a rationale behind feature "open diplomatic conversation (not first contact) on units' touch"?

I reread this thread and didn't see such a proposal anywhere.  Assuming someone proposed it somewhere sometime, I assume it means that upon physical contact, the other faction is forced to talk.  Politely declining to speak to you would not be possible.

This is already existing behavior. Not forced talk, just occasional. Contrariwise, I propose to disable this feature. If AI has something to tell you, let he say at the start of his turn, not when his last unit runs into yours. About avoiding conversation when player dials vs just denying player's demands in a conversation (if AI doesn't get provoked into vendetta you still are aggressor) - it's controversial feature, I'd prefer it be optional. After all, you might just be trying to improve your standing with him.

[bvanevery]

How is the AI going to know that you have units in range that can affect it?  Waiting until you actually make physical contact is a way to ensure that.  Otherwise the AI has to sit around trying to predict what you could, would, or might do.

Denying you an audience, is actually a historical method of diplomatic control.  Stalin did it to Mao a lot.  Drove Mao nuts.

In a perfect world of diplomatic discipline, you may be right that a diplomatic AI should be expected to just tirelessly rebuff the player.  However, I see a problem with "getting that train started".  What if there are exploits available to the human player?  Can't exploit what you can't even talk to, it's a more secure implementation.

The game is also conditioning you that you're not supposed to / allowed to spend infinite amounts of time on this.  You don't get to talk to other faction leaders forever, they get tired of you.  This has some benefit of protecting the player from himself, his desire to drive himself nuts with diplomatic micromanagement.

[PvtHudson]

Before ThinkerMod AIs rarely made it to the satellites stage, but now AI's sats are a common sight. And this raises a couple of questions.
First, one of the best SMAC features is that there are several ways to do most things. But sats distinct in that there is only one way to deal with them - ODP. Wouldn't it be logical to add a probe action to destroy or even hijack enemy sat? Probably restricted to bases with Aerospace Complex.
Second, how does the game store sat (sans GSP) data? Are they tied to bases that built them or it is a common pool? In latter, far more probable case, there is a possibility for an option what to do with sat pool of defeated/surrendered faction with variants: a) let them orbit idly as a memento; b) destroy; c) transfer to conqueror.

[bvanevery]

Before ThinkerMod AIs rarely made it to the satellites stage, but now AI's sats are a common sight.

I see them in SMACX AI Growth mod as well, if I diddle around long enough, maybe on an Enormous map.  I have fought major, major, spawn-from-hell satellite wars.

But sats distinct in that there is only one way to deal with them - ODP.

If you can deal with them endlessly that way, you're more stalwart than I.  The computer does not get bored with the mouseclicks, but I certainly do.  From a mouseclick sanity standpoint, just letting the AIs build and build and build satellites, as I do the same, is a pretty reasonable way to go.  Satellite combat favors the defender for the most part.  ODP pod attrition rate when they attack is pretty horrific.

If you're determined to take out enemy satellites, it is far more efficient in terms of mouseclicks to launch missiles at ground targets.  The AI will burn up its ODPs to intervene.

Wouldn't it be logical to add a probe action to destroy or even hijack enemy sat?

You are asking to participate even more deeply in greater and greater quantities of spam.  It is not worth it.  You can build your own satellites with fewer mouseclicks than stealing them.

I also think intervention on the ground is the True Way to deal with this spamming problem.  No matter how big the map, there's always a rail.

[PvtHudson]

AI (usually) refuses to trade SP-enabling techs, presumably because of large impact the SPs can have. From this standpoint, it is logical to have an option to restrict trade of other impactful techs, namely reactor, resource caps lifting, crawling-enabling, PB-enabling. Maybe even generalize approach: add to Alphax.txt #TECH lines TradeValue column, then map somehow these values to AI behavior through an option.

[bvanevery]

A decent idea, although I wonder how one prevents everything in the tech tree being deemed "important", for "some reason".

Couldn't we hack this as is, by creating bogus Secret Projects?  What do Empty Secret Projects 38 through 64 do when you associate something with them?

I'd definitely like to make the AI completely unwilling to trade Probe Team technology to anyone.  I'm not willing to spend a real Secret Project to do it, as there are not enough of them, and they need to come at appropriate times and places in the tech tree.  I don't allow any Secret Projects before Tier 3, and I've got Planetary Networks at Tier 2.

[PvtHudson]

Given how powerful TECHSTEAL and Spoils of War are, a 3-level option to govern them is desirable: 1) as it is now; 2) only give tech if captured/destroyed base had a Network Node; 3) only give tech if captured base still has a Network Node.

[bvanevery]

Why do you think TECHSTEAL is powerful?  Are you biased towards primarily playing multiplayer games?

In single player, I don't even see it as consequential.  The cost of stealing a tech, is making a completely disposable probe team.  You have to either sneak to the base, or bring a shielding unit to get into range safely.  You could even use the armor exploit and send an armored probe team, although it would be more expensive than just sending 2 units.  It's still easier than conquering a base.