Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Kyzrati]

Update 160511: The new fabrication system has been implemented for the next release (Alpha 9), and I've posted a full overview on the blog. Many of the ideas were discussed in this thread--thanks to everyone who participated!

Update 160608:
Alpha 9 is ready!


Plans have been in motion to improve all the non-terminal interactive machines, as seen with the Recycling Unit in Alpha 7.

I'm not sure that Fabricators will be next, but we started a discussion on them here, and we've both passed the point where there is enough information to make informed decisions, and are nearing the point where changes to this system will be extra meaningful to the game given the new content (for example, schematics available for parts you can't even otherwise find).

First, a little background on fabrication's role in the design.

Cogmind is obviously a game about adapting to the situation. Even after you find some useful individual parts, or gather exactly what you need for a highly effective build, sooner or later it's likely that circumstances will turn against you and you'll lose something important (if not more than that ). Moreover, you may also have an idea for a build but simply haven't come across one of the better parts it needs.

Enter fabrication.

The idea is that a limited number of times during a run, you'll be able to use fabrication to supplement a missing piece or two in a build, or just get something extra special out of it. Important considerations:

• The system shouldn't be tedious (like it is now).
• Fabrication conditions should be neither too difficult nor too easy to satisfy.
• Fabrication cannot be obviously superior in every situation, a real danger with a system where you can always get what you want. The benefits cannot always outweigh the costs.

So what does the current fabrication process involve? (base list courtesy of Shobalk)

• Acquire a schematic. (Terminals, Scanalyzers.)
• Acquire a container and fill it with matter. (Found, Workers.)
• Find a fabricator of high enough rating to build the schematic. (Found, Terminals.)
• Hack to load the schematic. (May fail.)
• Hack to start the build process. (May fail.)
• Wait for the build process to complete.

It's not often worthwhile to do all these things (and even if it is, it's not much fun).

The original purpose behind the large number of restrictions was to counteract the old hacking system: Prior to the hacking overhaul, it was theoretically possible to hack forever until you got what you wanted. Now that traces can no longer be reset, unless you're an extremely good hacker, a single machine/Fabricator can only be used so many times.

Note: I don't intend for every Cogmind to be great at fabrication! That's another part of the balance equation here--it doesn't have to work equally well for everyone. In that sense requiring two separate Fabricator hacks is okay (the first has a high static chance, while the second is dependent on the part), and Terminals can be fairly reliable sources for schematics. Hackers, or at least part-time hackers with a little bit of hackware (or those who just luck out), should be able to make some use of fabrication, while others don't necessarily have that option.

To me the most annoying restriction is the level restriction on schematics, so I want to remove that entirely. It's very opaque and complicated, and there is a much better approach to differentiating Fabricators of different ratings: Time. Any Fabricator should be able to build anything if you have the schematic (and matter), it simply takes longer at lower-rated machines. So the schematic data will show XXX/YYY/ZZZ, indicating the time required at each of the three Fabricators, where higher levels are more difficult to hack, as expected, but can finish your schematic much more quickly, further rewarding anyone capable of actually hacking those. (Side note: The lifting of all level restrictions should also apply to Repair Stations and Scanalyzers as well, for the same reasons.)

Time as a restriction has the advantage of already having a cost associated with it, in the form of Programmer dispatches and increasing security level from confrontations, and is therefore tightly coupled with the rest of the game's mechanics.

I'll probably further reduce the difficulty of the schematic loading hack (right now the base chance is 70%).

To facilitate the waiting period for a build to finish, I'll be implementing a new command, Alt-w (or shift-RMB on the machine), to wait as long as necessary to finish.

To recap, here's a summary of player opinions expressed in the other thread:

• Happylisk: remove containers, reduce hacks from two to one, keep time requirement
• Decker: remove containers, remove time requirement, add some other limitation on fabricator use
• Shobalk: get rid of level restriction, reduce hacks from two to one, keep containers, keep time requirement
• Sherlockkat: remove time requirement, scale matter costs by item rarity and consume the containers (indirectly reintroduces time requirement)

I'm not yet decided on whether to remove the matter container requirement. I think it's probably best to implement these changes in stages anyway, to see how the results of the ones mentioned above play out before taking further actions, but I wouldn't be against that one if it's necessary.

Consuming containers is quite an interesting idea, though technically you can fabricate more containers as necessary

And besides, I think a time requirement is important to the whole system, and an auto-wait feature resolves any negative effects of that.

That's the extent of my notes on this topic for now. If you have anything to add (or want to change your positions ), go ahead and continue the discussion here!


"Oh yeah, and Robots!"

Throughout this post I've been talking purely about fabrication of parts, ignoring the other important half of the system that enables you to build friends. I do eventually want building robots to be a thing, but it's secondary to part fabrication, as it's not as difficult to balance given the obvious impact on the security level looming over you in that case . Whatever changes are made to part fabrication will also be applied there, along with perhaps further tweaks if necessary because it's robots instead of parts. They're less important for this discussion.

[Happylisk]

Just a few general thoughts.  I stand by my suggestion to remove containers from the equation.  It's easy to snag a container, and it's easy to fill it with material.  Neither of these steps, however, are fun or interesting.  If carrying around material containers is not a natural part of your playstyle, making this a mandatory part of fabrication just become tedious grinding.  The requirement might be flavorful, but I'm a big fan of the DCSS ethos where flavor is always secondary to gameplay.

Keeping the time requirement allows you to introduce a tangible cost to fabbing, and helps differentiate higher level machines.  Even if you're strong enough to swat down programmer dispatches, there's still the security level cost.  If you throw in a command to pass time until something happens, it's not too tedious for the player. 

Here are two other possibilities.  One, each successful fabrication in a machine could cause the chances of success on that machine to decrease.  This would be similar to the alert purge mechanic, except localized to individual machines.  The second option is to have all acts of fabbing increase the security level; very small amounts for common/non-combat parts, larger amounts for exotic/combat parts (combat parts being defined as weapons and armor).  Either approach would reflect the same concept: there are ways these machines typically work, and a sudden change in whats being produced raises suspicion.   

The first approach also imposes a hard cap on how much any single machine can produce, which is probably a good idea.   

[Sherlockkat]

Here are some thoughts. I apologize if they are rather muddled and repeating some of your suggestions, but fabrication is hard to think about.

Cogmind is obviously a game about adapting to the situation. Even after you find some useful individual parts, or gather exactly what you need for a highly effective build, sooner or later it's likely that circumstances will turn against you and you'll lose something important (if not more than that ). Moreover, you may also have an idea for a build but simply haven't come across one of the better parts it needs.

This is my view of fabrication too. It is a support tool that allows different builds to plug holes in its inventory/setup. Speed builds can manufacture heat sinks, propulsion, sensors etc. Combat builds can manufacture armor, forcefields and combat utilities. If the fabrication system needs to fulfill this role, the primary gating mechanism for fabrication should not be hacking, but time. Hacking requires a Cogmind to be good at it, while waiting around is something all Cogminds can do. Here are some suggestions derived from this perspective:

1) Combine those 2 hacks into 1 and make it easy to pull off.

2) I don't think matter stores are very relevant. They won't deter a determined Cogmind from fabrication, especially if the system is more streamlined. Everyone would do it. They are flavorful, but getting rid of them won't break immersion. I can just as easily believe that fabricators have an inbuilt matter reserve. And you know what, you can limit fabrication by making those reserves finite.

3) It would be unfair to hackers if they can't leverage their abilities. So as we discussed in the alpha 7 thread, how about a trojan which reduces build time?

4) It would be nice if higher-level fabricators have near-instant build time. During my recent playthroughs as a hacker (I had +70 from my hackware and -20 to getting traced), I noticed that there is not a lot of things you can do with that much hacking prowess. There is access(main) and you can seal level 3 garrisons. I would like to be able to do more and being privy to instant build time from a fabricator would make hacking that much more useful.

The whole thing might difficult to balance. If the time requirement is too steep, the combat builds won't be able to use it effectively. If it is too lenient, it is going to get abused. Actually, I cannot think of a way in which combat builds can use the fabricators effectively. They already have high churn rates and are up against the clock. Whatever they are fabricating is unlikely to turn the tide of the game and the time requirement doesn't help. Compare that to a speed build. A single adv sensor array or a decent launcher can see you to the end of the game. If we are going down the time requirement route, it seems like there needs to be an alternate non-hacking way to speed up fabrication in order for it to be useful to combat builds.

Re Waiting: Isn't alt-w the command for detaching the part bound to w?
 

edits: grammar and typos

[Kyzrati]

Thanks for the input! (I've been waiting on this to see if there'd be any more conversation in the meantime, having already put down all my initial collected thoughts.)

There are strong arguments for getting rid of the matter requirement, and I agree it does lead to unnecessary tedium. Time to get rid of that.

I'm also in favor of increasing the difficulty of success for each previous hacking success at a given Fabricator. In fact, this rule should perhaps be applied across the board to all hacks, for consistency's sake? It can make sense, too, since repeatedly hacking the same target again and again is just asking to be traced. That and there aren't really a lot of hacks that you'd do more than once at a machine, anyway, so it wouldn't have much of an impact on other hacks. (Obviously we already have the map-wide restriction on a handful of the hacks that are heavily restricted.)

3) It would be unfair to hackers if they can't leverage their abilities. So as we discussed in the alpha 7 thread, how about a trojan which reduces build time?

I don't think we'd need this if, as you suggest, higher-rated Fabricators build things really quickly, and that'll probably work okay, since those are pretty hard to hack in the first place.

Re Waiting: Isn't alt-w the command for detaching the part bound to w?

Yeah, that was just me throwing that out there with thinking about conflicts . It should be Shift-Alt-w. The Shift-Alt modifier combo is used for fairly rare special commands.


I think it'll balance out okay in the end, but first let's implement all these changes and see what happens

[Draco18s]

Thoughts on containers:
I don't like them much, especially as sometimes finding enough matter just to keep shooting and replacing parts is hard, followed be arcane interactive to actually get the matter into the container, then the container into the machine.

The complex is supposed to be Doing is Own Thing and supplying material to fabricators for its own needs already: just merge the two. If a fab had no material, it should request some from the base AI and have it dispatched.
 * Pros: no need to move matter around yourself (but you can)
 * Cons: by not adding material yourself the fabrication might get noticed as not being legitimate

This also means that you could stumble into a fab producing items already and just have to wait for them to pop out. Or a fab already loaded with matter. Maybe be able to hack the fab to extract is working schematic. Or the fab is busy and it adds extra time to your job (back of the queue). Added possibility of another hack/trojan to bump your thing to the front and/or abort the current item.

[Kyzrati]

My original desire was to have Fabricators contain their own (unlimited) supply of matter from the complex, since that's more logical, yeah (Repair Stations already do it!), so the matter container thing was just another of those restrictions that ended up being too restrictive (and actually against what I thought would work logically--but the design tries to emphasize gameplay over anything else).

Hm, I do really like the idea of fabricators producing items on their own. A lot. It would require several extra features, though, like AI that will come collect what's produced and do something with it. Also, robots could sometimes be created there and head off to the nearest garrison (it's also a better way to replenish non-combat robots, too, rather than always sending them through exits from other floors). I'll probably do this at some point, though it'll be a non-small project...

[Draco18s]

I figured it wouldn't be a simple or small thing, but it was the impression I had from fabricators when I first encountered them.

But yeah, I was thinking having the matter delivered by the neutral bots, just to stay in keeping with the container mechanic. Introduces a way for the player to stock it to avoid being noticed (they'd have to have the material on hand first) or a player can be "lazy" about it and have a chance of a patrol being dispatched to investigate the strange order (might work it into the hacking mechanic: eg "fail by 5, get the fab going, but dispatch troops," but there wouldn't be much in the way of terminal feedback).

[Kyzrati]

I think we can do without matter specifics there, or at least that's further removed from the absolutely essential part of "what do we do with these things that are actually being fabricated"

Still, fabricator auto-production is definitely going in, because yeah we can use more machine interactivity! An easy way to continue making the world seem even more alive.

And about the lack of terminal feedback for some situations, that's best avoided since it becomes difficult for players to learn effective strategies. Even when you know how it works, the optimal way to play would be to assume that any failure might have led to a dispatch.

[Draco18s]

And about the lack of terminal feedback for some situations, that's best avoided since it becomes difficult for players to learn effective strategies. Even when you know how it works, the optimal way to play would be to assume that any failure might have led to a dispatch.

I was rambling a bit and was on my tablet.  I was more trying to say "there's a possible pitfall here that the players don't know bad stuff happened" and that either (a) there needs to be some visual feedback or (b) handle it a different way.  The idea was just referring to a middle-ground between the current failure (you get traced) and current success (the part gets fabricated): instead, the part gets fabricated and bad stuff happens.

I do feel that the player should know that this third option occurred, just that I didn't know how to handle it.

[Kyzrati]

Ah okay, I thought it was an unknown. Got it.

[Sherlockkat]

(Edit: Post moved from Alpha 8 Discussions thread, talking about drone effectiveness.)

The drone bays mechanics are fine and so are their rarity. I think the only reason they could considered a problem is because of the proposed changes to the fabrication system. I think it might be better to nerf the fabrication system than change the drone bays or any other items. I mean, I can see fabricating hacking suites being just as powerful. One thing that you can do is actually offload the fabrication difficulty to acquiring schematics than the actual fabrication process.

Another idea: how about having a global limit on fabrication? Make it progressively harder as we fabricate more stuff and higher rating stuff and let the difficult persist across levels. Let us fabricate point singularity launchers by -7. But after we do that, we can't fabricate anything for several levels. Make it similar to alert level. Fabricating a point singularity launcher should be the fabrication equivalent of blowing up 3 garrisons. We have to hack our fabrication alert level down if we want to fabricate more stuff. Make the fabrication alert level decay down slower than our regular alert level.

[Kyzrati]

Making schematics more difficult to acquire would work if Scanalyzers were removed, since they'd either otherwise be frustratingly difficult to succeed with, or a good enough hacker could still use them to get anything. Unless we say it's impossible to get anything but common schematics from a Scanalyzer, and the rest have to come via other means. That approach seems to take away a lot of the freedom of the system.

While I wouldn't want to introduce a whole separate alert system for fabrication, some kind of global/larger restriction will need to be put in place on the fabrication side of things for it to balance correctly once the tedium and other limits are removed.

It would be nice if the system scaled so that it's still possible to produce larger quantities of common parts, if you really want to, but the really good stuff will be more limited. I'll think about what other kind of overarching system could be behind that. Certainly now that the matter will be sourced from a central/alternative location, that could play a role as well.

[Decker]

It would be nice if the system scaled so that it's still possible to produce larger quantities of common parts, if you really want to, but the really good stuff will be more limited. I'll think about what other kind of overarching system could be behind that. Certainly now that the matter will be sourced from a central/alternative location, that could play a role as well.

I fully agree with this, and with Sherlockkat about imposing a global limit on the use of a fabricator.

Here are some more thoughts I had about the system.

1) How about a low-difficulty trojan(botnet) on a fab? Then you earn "fab" tokens each time you hack a new fab. Or you just earn tokens every time you find a fab.

2) Yay for the matter requirement being gone!

3) Building robots over time sound like fun.

4) Finding a fab should be a "weeee!" event: fun, positive, no tedium. I still strongly disagree with waiting for fabrication, because I fear it will lead to all four of the following problems.
- False positive: waking up when it's not necessary, because a patrol was detected over sensors.
- False negative: not waking up when it would be necessary.
- Suboptimal: while the player would resort to waiting to avoid hassles, the optimal play would be to continue mapping with a terrain scanner, explore, kill workers for matter, etc.
- Hide-and-seek games: if there is a patrol heading the fab way, then the player has to actively hide until the patrol strolls through. This is easy but not particularly fun.

[Kyzrati]

I thought about it more today and the likely new system will be something on the order of global matter stores allocated for the fabrication network (probably on a per map--or map type--level), and you can draw from those stores to build things. (The AI also draws from those stores and occasionally builds things, too.) Better parts will draw significantly more matter (unlike the current relatively small spread), and as the stores drop below a threshold, there is a greater chance of a floor-wide fabrication block and an immediate jump in security level, possibly with more patrols dispatched (we'll see--whatever it takes to balance in the end).

The stores will also increase and decrease on their own over time.

Auto-waiting does pose those first two problems, and honestly I think the best option is still to just show some kind of countdown timer to completion so you know when it's coming, it's not very hard to hold a button for a few seconds to wait out a timer. This timer could be shown directly on the map over the Fabricator itself as it runs to zero. (That would also solve the problem of new players not realizing you have to wait around for it. In fact, this same timer should be shown for repair stations. You can already get the info by right-clicking on the active machine, but why require a click )

The optimal play of continuing to do other things while building can still be dangerous. You know you've had those situations where you're like, oh, I'll just check around this corner, or go over to that room, and then be on my way, and suddenly you are in very serious trouble! I don't imagine it being a long wait, anyway, really similar to what it is now, with better hackers who can access high-tier Fabricators getting them even faster. (Like, much faster.) Waiting will be less of a heavy requirement with a global system in place, but I don't want it to be instant.

- Hide-and-seek games: if there is a patrol heading the fab way, then the player has to actively hide until the patrol strolls through. This is easy but not particularly fun.

Same thing: "Easy" doesn't always work in your favor when unexpected things happen... Undetected trap blows a hole in a wall; Hunters masked against sensors were patrolling on the other side; time to party. Just one of many possibilities.

[Draco18s]

Sounding awesome!

[Sherlockkat]

@K: Just saw the new fab changes on twitter. The on-map timer looks cool. The only thing about Fabrication that I still don't like is the time reqs. I think you can eliminate if it the global matter store is tight/limited enough that we have to pick and choose between the high-tier prototypes, which is all I intend to do, that we have to build. That alone would limit fab farming.

It's not a deal breaker by any means though. I still look forward to using/abusing the fab system in alpha 9. Maybe we can convince you to get rid of the time reqs in alpha 10 .

[Kyzrati]

I don't know if I'd go quite that far... but yeah we'll see how Alpha 9 turns out. I played some with the new system and it's nice, though I'm sure it will be abused and I would like to see how first

For everyone: I just posted to the blog an overview of the entire new fabrication system, which has already been fully implemented for Alpha 9.

[Sherlockkat]

I can think of some creative uses for it in speed builds. Can't wait to fabricate those distraction..um I mean..drone bays in alpha 9. I recently discovered this cute tactic with mini drone bays when you are being pursued by swarmers. Turn a corner, release a single drone and let him run back around the corner using the goto command. The swarmers would waste a turn or two shooting at it giving you plenty of time to escape. But, seriously there is a whole list of items which would make speed runs not necessarily easier..but more interesting.

Question : Do you have any plans to add more interesting sensors to the game? I remember there being talk about a sensor which tells you how many enemies are pursuing you. It would be cool to have some new kind of sensor models apart from terrain scanners, optical arrays and sensor arrays to play with. Might make speed builds feel less samey..

edit: Typos and sentences in general

[Kyzrati]

Yeah, I agree that once players can somewhat more easily pick and choose a few parts they really want (come Alpha 9) we'll start seeing some changes in strategy, especially for the less combat-oriented builds.

With regard to this, I've been thinking about changes to drones in particular. What if you were forced to keep the bay attached in order to maintain a remote connection? I'd like to take the restrictions even further than that, but doing so would be illogical... Opinions?

Do you have any plans to add more interesting sensors to the game?

Yes, there are at least three sensor ideas still sitting on the list for serious consideration, including that one. I've been waiting until later to get to that stuff, but any time a current player brings something up it gets moved higher on the list . I like to add features that help change up the whole game, alongside the usual world/map expansions, and those are good candidates so I'll look them over in the near term. (Though for Alpha 9, machine updates are already the big one.)

[Sherlockkat]

Yeah, I agree that once players can somewhat more easily pick and choose a few parts they really want (come Alpha 9) we'll start seeing some changes in strategy, especially for the less combat-oriented builds.

With regard to this, I've been thinking about changes to drones in particular. What if you were forced to keep the bay attached in order to maintain a remote connection? I'd like to take the restrictions even further than that, but doing so would be illogical... Opinions?

The change that you propose could be an interesting test of how powerful they are. I don't think any more nerfs is necessary. Leaving the drone bays equipped when you are strapped for mass support early in the game is difficult. Late in the game, mass support is not a problem. But, I feel that speed runs do need a little something that deep into the game. Mini-whine (feel free to ignore) : Access has a lot of shiny weapons that speed builds cannot use. It is a pretty big floor and you are just running around praying for a terminal which are surprisingly rare. Then, there is the problem of getting those sealed doors to open. Looking through my scores folder, of the last 10 serious runs, I made it to access in about 6 of them and died. It's usually one thing or another : lots of hackware but no terminal, found an energy cycler but not enough hacking gear, Found the exit but no way to open the doors. Now that I look back on my previous speed wins, I can appreciate the amount of luck involved in me getting out. (edit : To be clear, I don't think that access needs to be any easier. I just think speed builds need more toys to play with)

General thoughts on drone bays : I think drone bays are very powerful, but not abusable. The drones are a finite resource and you need to be very careful with their use. That takes a certain amount of skill. I rarely manage to recover the drone that I deploy as I usually use it as a forward scout to figure out enemy patrol routes. The only abusable one I think is maybe the mini drone bay because of the ample number of drones it provides. Nerfing it by reducing the drone count would make it strictly weaker than the mapping drone bay. So, maybe mapping drone bay = 2 drones and mini drone bay = 3 drones?

I also think their use a distraction tool makes them very versatile and covers a niche that no other utilities address.

Edit : Also dude, broken clavicles are serious business. Take care..

[Kyzrati]

Remember that the late-game branches aren't in yet, and those will change that part of the game a lot, including also the addition of new toys. Only a small portion of the alien artifacts have been added so far, and there are other non-alien prototypes to be discovered, too--the best in the game that cannot be found randomly. So the balance in that part of the game will remain somewhat off for now.

Maybe a drone count reduction for those is in order, yeah. ... I just now changed the Mni. variant from 4 to 3 (also changed its art), and I'll be leaving the Mapping variant as is at 3 because I see it was supposed to be more restricted than the current settings--that drone is actually based on Derelict parts, but it was allowed to spawn anywhere. Now as a Derelict-only part the bay itself will also be more difficult to get.

Regarding distraction tactics, I think that now that fabrication is easier, buildable allies will also likely be used as distractions as well.

Take care..

Yeah, woulda been nice if that just didn't happen, eh?