Bugs, crashes, typos & bad taste

[CrazedHarpooner]

There's an issue with the new mission Forest Lodge Defense. Only the 1st unit of the entire squad spawns in. The issue lies in that the Lodge block is spawned in as a craft and there's only a single open entry tile ouside the door. There's actually 2 more entry floor tiles next to that one, but they're both occupied by a content element. For craft spawning, soldiers will only be placed on valid entry tiles, ignoring xcom nodes by what I've observed. I've experimented around and managed to get it to work. Had to redeclare the block as a terrain map and used verticalLevels to get the desired effect. When no craft has been placed on a map via addCraft, xcom units will use xcom marked nodes for spawning and will clump up randomly around them if there's not enough nodes for all units.

During my testing I discovered that one of the forest terrains: FINNIKFOREST01 had an xcom node defined so they all spawned over that side far from the lodge. I've changed that node to another type so xcom units don't spawn over there, but don't know if that is intented for another mission which could break. I haven't seen such a mission while working on my submod, but then again, there's plenty of missions I haven't adapted yet.

Also, since I no longer used an addCraft, I had to change the node at the Lodge entrance to an xcom node as at least 1 is needed to spawn xcom units this way.

I've attached all modified RMP and the small ruleset files that fixes this. You can always add or modify nodes in the lodge to let the xcom spawns spread out a bit more.

[Solarius Scorch]

There's an issue with the new mission Forest Lodge Defense. Only the 1st unit of the entire squad spawns in. The issue lies in that the Lodge block is spawned in as a craft and there's only a single open entry tile ouside the door.

Yeah... That's how it's supposed to be. The description specifically says "To avoid attention, send only one agent to investigate." Allowing more people would trivialize this mission, and I wanted a mission with desperate civilians, not yet another zombie slaughter by a group of cyborg commandos from the future.

During my testing I discovered that one of the forest terrains: FINNIKFOREST01 had an xcom node defined so they all spawned over that side far from the lodge. I've changed that node to another type so xcom units don't spawn over there, but don't know if that is intented for another mission which could break. I haven't seen such a mission while working on my submod, but then again, there's plenty of missions I haven't adapted yet.

You're right, this looks like a 100% bug. Weird...
Thank you!

[CrazedHarpooner]

Ah, my bad, I assumed more was allowed since the lodge 'craft' ruleset defined room for 18 soldiers. I'll take note of this.

[Solarius Scorch]

Ah, my bad, I assumed more was allowed since the lodge 'craft' ruleset defined room for 18 soldiers. I'll take note of this.

You're right, it should be 1. Doesn't impact the game in any way, but is quite inelegant.

[CrazedHarpooner]

I was suspecting I might be hitting the same or similar issue in Night at the Museum mission. So I tested it there and took a save when dealing with the final enemy unit, saurian ghost:

- stunning him (resulting in four squares ZZZ's) with two hits of a power mace does not give you a live capture (should it?)
- ending the mission with MC also does not give you a live capture

Note that I don't even have ghost tank researched, let alone have one built, but I don't think that should affect whether you should be able to at least see the live capture (though you can do nothing with it). In earlier games I have usually had ghost tank build on a secondary base and been able to transfer ghost units there from mission debriefing screen.

Also in debug mode for whatever reason the hotkey CTRL-J for stunning all live aliens did not work, it displays text "Deploying Celine Dion album" (neither CTRL-ALT-J), but CTRL-K for killing works.

I wonder if this is some kind of stunning mechanic issue or related to the way how ghost units are defined.

I got around to testing this hypothesis with a New Battle of Haunted House, where there are stunnable humans and ghosts both. CTRL-J works for humans but not for ghosts. In this second save I didn't manage to manually stun a ghost (and then kill the rest), but on another having one go ZZZ also doesn't give you a live capture of the ghost.

At sometime this must have changed somehow, as I have clear memory of transferring poltergeists and ghosts from mission debriefing screen.

I used the save game provided to make a copy but with a ghost tank in Venezuela. First screen shows what you get if you capture without any ghost tank anywhere (basically the unmodified save) and the second image shows what happens when there's at least a valid prison type for the capture available. And yes, it gives you the option to transfer it at that point, but you still need a valid prison for it somewhere.

Additionally, New Battle won't help testing for this scenario as I've observed in my experiments. It simply churns all captures (MCed or stunned) in the debriefing blender and they all become corpses. I used vanilla UFO Defense to generate a medium scout with sectoids, warped my way around in debug mode with the stun rod and bonk them all to sleep. Debriefing reports none made it back alive anywhere.

Edit: self correction.
The issue with New Battle is not the option itself, but rather it recreates a dummy base using the starting facilities. Since most starting bases usually lack the required facility, all captures of any kind will just be killed in the afterbattle report. If you mod it so this starting base comes with a compatible prison type, or mod any of the starting facilities to act as said prison type, the debriefing will reflect that upon completing the New Battle.

[psavola]

Thanks for testing and verifying the engine behaviour. A bit odd that you get more points out of live captures when you have the corresponding containment facility built versus when not, but there we go..

[SBBurzmali]

One aspect of the game that has never felt great is the engineer lottery. Basically, the transition from the mid game to end game is largely gated by capturing dozens of engineers and leaders and hoping you pull one of the few interrogations that are valuable. This can be rough as there is plenty of overlap with navigators and even some with medics leading to a pretty complicated "optimal" way to interrogate aliens. I'm not sure if there is a better answer, long research times gating progress is a staple in Xcom-like games, but it still doesn't make spinning the interrogation wheel any more fun.

[Solarius Scorch]

One aspect of the game that has never felt great is the engineer lottery. Basically, the transition from the mid game to end game is largely gated by capturing dozens of engineers and leaders and hoping you pull one of the few interrogations that are valuable. This can be rough as there is plenty of overlap with navigators and even some with medics leading to a pretty complicated "optimal" way to interrogate aliens. I'm not sure if there is a better answer, long research times gating progress is a staple in Xcom-like games, but it still doesn't make spinning the interrogation wheel any more fun.

I'm not a huge fan of this either (worked better in FMP), but at this stage I do not have a good alternative. I'd probably have to redo the whole tree of alien culture/tech (Alien Vocabulary etc.) to make it dependant on other stuff, not just captured specialists. And that's a major writing problem.
To be honest, I haven't received much criticism for this, so I haven't been motivated to do something about it yet.

[Stone Lake]

The optimal way to interrogate aliens is capture 10-20 leaders (there's 1, or 2 in some medium UFOs), and interrogate them. Almost every topic they give is useful, and they can be quickly exhausted. The inverse (saving leaders for last) leads to said lottery. Interrogating engis/navis/medics is mostly wasting brainpower, and should be done after you get good tech.

[psavola]

Leaders are more effective, yes, but in the midgame (before and briefly after the start of 1999) the most chances you have at alien captures are small UFOs, from which you can only get soldiers, navigators and a bit more rarely engineers (depending on craft). So quite often you can and should get started with navigators and engineers. But it is of course possible to restrict your initial research to only a select few samples.

[SBBurzmali]

Often I find myself with 1 leader, 2 or 3 engineers, 2 or 3 navigators and a fistful of assorted others after a mission. At the beginning, leaders have, for example, 25 techs of which 5 are useful. One pull at 20% isn't great and early on leaders aren't easy to come by, so if you interrogate a soldier or two first, then the navigators, then engineers, you might get those odds up to 25 or 30%. If you are swimming in Leaders, sure, but an early hit on one of the important techs can improve survivability in the mid game substantially, so I do my best to maximize the chance on early pulls of the wheel.

As to a fix, well, it's hard to say. The "substantial breadth" is a big draw for XCom Files, the issue is just that it is a mile wide but inch deep at some points, often when you've stalled and are waiting for that one mission type you need to advance to spawn. You could add options to make it easier to get those techs, or alien techs in general, but that'd just exacerbate the wait for specific spawns near the endgame. Long term, do you have a goal for how research should work in the end game? Is it expected that the player will fully research every option? One path some games take is to provide research chains that while not mutually exclusive, require more resources than the player has to fully explore by the end game, sort of like a softer version of the codex from Xpiratez. So instead plasma being the default endgame weapon, maybe have terawatt laser, warp cannon, and monofilament flechette, but each requires a data core that is a one off from a very slow to reoccur mission and the data core has to be "cracked" providing only one of the technologies.

[Mathel]

I'm not a huge fan of this either (worked better in FMP), but at this stage I do not have a good alternative. I'd probably have to redo the whole tree of alien culture/tech (Alien Vocabulary etc.) to make it dependant on other stuff, not just captured specialists. And that's a major writing problem.
To be honest, I haven't received much criticism for this, so I haven't been motivated to do something about it yet.

Is it possible to make a research topic consume multiple items from the stores/prisons?

If yes, vital researches might be able to be brute-forced by consuming an alien and a relevant object. I am thinking something along the lines of using one UFO Propulsion and any alien Engineer to research Alien Gravity Generator.

Or perhaps make all brute-forcing depend on one expensive resource, such as Alien Data Slates. One could either decrypt and research it to play the roulette, then research a Leader/Engineer to play it again, or they could consume both to guarantee a certain topic that the Leader/Engineer in question has a chance of telling.

[psavola]

For what it's worth, I have never had a major problem of doing those dozen or two dozen alien research projects. You can get all of them done reasonably smoothlessly in 1999-2000 timeframe.

They also follow exactly the same pattern as XCF in general in many other aspects, except that alien tech research is mandatory for progress while some other areas are just fillers (for example, almost all things Alien Hybrids will tell you are just additional information or the same thing aliens will tell you with a greater research cost; the same with syndicate scientists and MIBs, especially MIB scientists). There are multiple layers of information you can glean from prisoners, and it might make sense to start working from the ground up (from the prisoner who is knows least and/or cheapest to interrogate). Or you could just try to get what you need and skip the rest. Personally, I've always researched everything, but I'm a research junkie (I have usually built 200-300 research space by the end of 1999 so you advance the research quickly).

A bigger problem - and where I have usually stopped playing the campaign and started from scratch after a while - is playing the MIB commander and/or ethereal commander capture roulette, waiting for your chance to grab them. And I think based on reports SS already ackowledges this is not optimal, but how to get it better is the question.

[Solarius Scorch]

Long term, do you have a goal for how research should work in the end game? Is it expected that the player will fully research every option?

Insert the "Do I look like a man with a plan?" meme here...

In seriousness, I'm hoping for a certain amount of randomness in what you get in a given campaign and what you don't (or you do, but too late to count).

One path some games take is to provide research chains that while not mutually exclusive, require more resources than the player has to fully explore by the end game, sort of like a softer version of the codex from Xpiratez. So instead plasma being the default endgame weapon, maybe have terawatt laser, warp cannon, and monofilament flechette, but each requires a data core that is a one off from a very slow to reoccur mission and the data core has to be "cracked" providing only one of the technologies.

That's a commendable direction, but XCF's premise gives little room for such branching. Piratez takes place in the far future, with many alien civs and such available for examination, whereas XCF focuses on retroengineering just Cydonian tech, with some aspects of Dimension X and T'leth solutions.
I'd like to add some more techs from secret groups like The Grid or Discordians, but their capabilities would also be limited to what you can do on present Earth.

Is it possible to make a research topic consume multiple items from the stores/prisons?

Nope.

Or perhaps make all brute-forcing depend on one expensive resource, such as Alien Data Slates. One could either decrypt and research it to play the roulette, then research a Leader/Engineer to play it again, or they could consume both to guarantee a certain topic that the Leader/Engineer in question has a chance of telling.

I guess, but I'd rather choose something more interesting than data slates. Even live engineers are more interesting.

A bigger problem - and where I have usually stopped playing the campaign and started from scratch after a while - is playing the MIB commander and/or ethereal commander capture roulette, waiting for your chance to grab them. And I think based on reports SS already ackowledges this is not optimal, but how to get it better is the question.

One day I'll actually get to the end game... and then I'll start making adjustments.

[Stone Lake]

One pull at 20% isn't great and early on leaders aren't easy to come by, so if you interrogate a soldier or two first, then the navigators, then engineers, you might get those odds up to 25 or 30%. If you are swimming in Leaders, sure, but an early hit on one of the important techs can improve survivability in the mid game substantially, so I do my best to maximize the chance on early pulls of the wheel.

This is exactly the erroneous thinking that won't get you anything. 20-> 25% is like nothing. No, leaders need to be prioritised if you need techs - e.g. a lucky laser pull, and progress to adv lab. Granted, after researching at least one engi / nav, etc.

Survivability is not really the function of the lucky techs either. Most core survival techs (PSI, tanks, advanced soldiers, extinguishers, medkits) are alien independent. Some good late armor can be looted from captures.
And to produce those armors, you need advanced lab = ALL the techs, not just a single lucky one. And that means exhausting leaders.