SCP Beta 4 Issue Reporting

[674451da672bdvoodoo47]

yes as far as I'm concerned.

[674451da67660sarge945]

I love it, but it still seems somewhat dim

[674451da677dbZylonBane]

The hue, brightness, and saturation were selected to make it look approximately the same as it used to, just without transparency.

[674451da67932sarge945]

Another tiny one:

Hydro3.mis object 751 (champagne bottle) clips through shotgun shells (753). Should probably have physics controls enabled, but then clips through the roof. Suggest changing to a cola instead?

Something should be done about the nearby magazine (285) as well. It clips horribly with the nearby maint tool (283)

[674451da67a5cZylonBane]

For those wondering what Sarge is talking about:

Sigh.

[674451da67bbcbombum]

Wow, it's almost like he said tiny for a reason. It's not like most of SCP is correcting errors basically nobody notices to make SS2 a more well-rounded and polished experience

[674451da67cd4sarge945]

I don't need other people to fight for me. I tend to just leave ZB alone when he hasn't had his meds.

Worst case I can just stop reporting issues if it's going to be a painful process.

[674451da67f5aZylonBane]

So, some fun facts about ecologies that we've just discovered: They're supposed to work by executing at set intervals, usually every few minutes, to gradually repopulate the levels. But due to an oversight in the script that manages ecologies, every time the script starts, it starts a new timer for the next spawn interval, even if there's already a timer running. This means that every time you enter a level, or every time you save and load in a level, another duplicate ecology timer is created. So after a while it's pretty easy to get the game in a state where ecologies on some levels will fully repopulate all at once, due to the ecology script executing dozens of times more frequently than it's supposed to.

But wait, there's more! Somehow, all these duplicate timers end up scheduled for the precise same moment. This is important, because the decision to spawn, and the actual execution of a spawn, are handled by different scripts on different objects. When the ecology script decides it wants a spawn, it dispatches a message to the spawner object, which gets shoved onto the message queue. So when the ecology script gets called dozens of times in the same moment, it's looking at the current monster population every time. So it dispatches a spawn request every time it runs. When the spawn script finally receives all those spawn messages, it happily executes them, potentially pushing the monster population on the level far above the maximum defined by the ecology.

This doesn't affect all levels, because some have their ecologies set up to limit spawning to one active monster per spawn point, but it definitely affects some levels.

Anyway, it's fixed now (probably).

[674451da68073voodoo47]

translation: enter Hydroponics, save and load a couple of times, run through the level and kill all AIs, wait for the ecologies to fire, and you will get 20 spawns instead of the intended 4.

[674451da68256sarge945]

I knew something was up with ecologies. I would notice frequently they felt bugged, like they were spitting out too many enemies, and thought it was related to "bulkhead cheese" when setting off alarms. Turns out it's just always been bugged.

I always felt like the ecologies were way too forgiving, now you're telling me they're bugged and should have actually been far more forgiving?

I feel that SCP should probably increase the spawn rates and the max population of most ecologies (maybe 1 extra spawn, 20 seconds less) if it's going to fix this bug. Many maps already feel mostly empty once you clear them because respawning is very slow (and a lot of them only have 1 non-alarm ecology, like medsci1), so the map repopulates VERY slowly. Fixing this will make that worse.

I definitely need to make a harder ecologies mod.

[674451da68824ZylonBane]

I always felt like the ecologies were way too forgiving

Judging by online commentary over the last 20 years, you are in the overwhelming minority in feeling this way.

Anyway, this fix only affects levels that have been revisited and/or saved-scummed in many times, so the "default" respawning that most people experience will be unchanged.

Another thing fixed-- in the vanilla script, loading a save will restart all ecology timers. So someone could prevent respawning altogether by save/loading every minute or two. In the updated script, ecology timers will pick up exactly where they left off when loading a save.

I definitely need to make a harder ecologies mod.

Anyone can already increase ecology activity by adding "lower_spawn_min" to the config file with a negative value.

[674451da68e00sarge945]

Judging by online commentary over the last 20 years, you are in the overwhelming minority in feeling this way.

Probably.

Throughout various playthroughs, I have gone from "the respawns suck, I want to disable them" to "yeah okay they are fine" to "wow these don't actually respawn all that much".

I guess it depends how many times people play the game and how familiar they are with it. For the average person who will only play it once or twice, the respawns are probably fine.

That said, I'm sure almost everyone in various LP's I have watched has experienced increased spawns in eng1, which is a map you do return to several times. In almost all of these cases, it seems like people comment on them ("wow, there's a lot of spawns", etc) but also kind of enjoy the experience.

This could be another typical case of gamers not knowing what they are talking about when it comes to their own fun, which happens pretty often.

Anyone can already increase ecology activity by adding "lower_spawn_min" to the config file with a negative value.

This I did not know. I'm going to try it. However this may not actually solve the issue completely because most ecologies are still capped to 1-2 total spawns, which IMO is not enough to properly repopulate the level, so a mod might still be needed.

For instance, the only non-alarm ecology in medsci1 (1187) has a max population of 2, which means that regardless of respawn times, at most you will see 2 extra pipe hybrids repopulating the level, which isn't much. It does have a rand of 3, which from what I understand means it can sometimes spawn a third hybrid, which still barely helps.

[674451da68fbeZylonBane]

And "raise_spawn_rand" can also be given negative values, thought if it pushes the base value into the negative it'll probably break things. I should clamp it to zero in the updated script...

Irrational did a really half-assed job of explaining what the spawn config variables do. Basically, the logic is this:

Code: [Select]

if ((count < max) && ((count < min) || (Data.RandInt(0, rand) == 0))) {
   ...then issue a spawn requestWhere "count" is the current ecology population. Which basically means:
- Never spawn beyond max.
- Always spawn when below min.
- When between min and max, spawn randomly. The closer "rand" is to zero, the more likely a spawn.

[674451da690d0sarge945]

hypothetically, if one wanted to read this script code, how would they go about doing it?

[674451da691b4ZylonBane]

Wait.

[674451da69389voodoo47]

heh. anyway, quick tweak to the fusion cannon. like/dislike/whatever?

https://www.youtube.com/watch?v=yl-fABe56FM

[674451da694d7sarge945]

Anything that visually shows a weapons state is good in my books. If it was up to me, every weapon in the game would show it's mode visually.

I would probably change the colour of the green core at the front (and the actual projectile) for DEATH mode, though.

[674451da695fcvoodoo47]

yeah, more extensive changes definitely are possible, but at this point, I don't think we want to get into custom resources - the current tweak can be accomplished by one simple model edit and one line of code, using all vanilla resources, so we can just throw it in and call it good.

better than nothing™

[674451da696e6ZylonBane]

I was thinking it would be interesting to, on the Rec deck basketball court, add some additional worm piles to the upper area on the hard and impossible difficulties. Thoughts?

[674451da697c5voodoo47]

no objections if it's just a few.

[674451da698ebZylonBane]

Also, we've been discussing ways to make Tinker less useless. We're somewhat limited by what we can do with OS upgrades since we don't have engine access, but something we can do, that's relevant to the Tinker skill, is modify the properties of weapons. For example, Tinker could decrease the minimum skill requirements to repair, modify, and maintain a weapon. Or maybe that would be overpowered?

[674451da69a07voodoo47]

it would make the trait almost too good, but I think this is still acceptable, as you will be losing an OS upgrade slot.

it would basically be naturally able in disguise - save a bunch of cms by not having to upgrade tech stats that much.

[674451da69b11ZylonBane]

Well how about just decreasing the requirement for two stats instead of three then? Modify and Maintain? Leaving out Repair would actually simplify things since that can be used on non-weapon items.

Also remember that players will want to upgrade Maintain anyway if they use energy weapons.

[674451da69c26voodoo47]

yeah, if just Modify and Maintenance is doable, then that's definitely less op.

so less nanites when modifying weapons, and lowering modify and maintain requirements for all weapons by one? sounds good to me (ideal for energy specialists as they will be able to maintain their emp rifles at maintenance 5 instead of 6).