Add additional "buffer" slot in labs

[nadnad]

TL;DR

Add additional "buffer" slot to labs so that when using input science packs of variable quality labs will not stutter.

What?

Add an additional slot to labs to labs for a small buffer of mixed quality science packs.

It could work like the heating tower does where there are two slots for fuel, but instead labs would need 2 slots per possible science pack.

Example of lab with stuttering due to mixed quality of input science packs.
Imgur Link:

mp4 file:

FactorioLabExampleSmall.mp4

(737.63 KiB) Downloaded 10 times

Example of buffer slot

FactorioBufferSlotExample.png (64.24 KiB) Viewed 283 times

Why?

When using variable quality science pack inputs, labs will briefly stutter off and then on when a science pack of one quality runs out and a new science pack of a different quality level is inserted. It is a bit unsatisfying to see your science production line fluctuate for this reason. You could mitigate this issue currently by running labs dedicated to specific qualities, but it would be cool if you didn't have to do that.

[macdjord]

Note that adding only one buffer slot will only help if you have a single mixed-quality science.

Otherwise, let's say you have red and green, each in normal and uncommon. Whichever kind of red science arrives first, normal or uncommon, will get put in the red slot, while whichever kind of green shows up first goes in the green slot. Whichever of the two remaining types and quality of science shows up first will get put in the buffer slot. Now the lab starts researching, using the science in the non-buffer slots. At some point, one of those sciences, red or green, will run out. Now, if that happens to be the same colour that's in the buffer slot, great; the lab can just instantly move the buffered stuff into the active slot and continue on. But if its the other colour, the lab will have to stop researching until an inserter adds a science pack of the other quality into the now-empty slot.

I think if would work if you had one buffer slot per science type. Even if you have more than two quality levels on the belt, as long as their are two of them in the lab, when one runs out it can continue researching using the second, while an inserter add new packs of some other quality into the now-empty buffer slot. The buffer slots would have to be filtered by science type, though.

[atarihomestar1]

I think that, unfortunately, the easiest solution is to not mix quality sciences. If you're throwing quality modules into science assemblers to get that bonus % from quality,viewtopic.php?t=126882 < This discussion and others show how in nearly every case, it's better to use productivity modules.

However, if you have dedicated "quality science" assemblers that use leftover quality materials from your base, I guess another solution could be just having separate labs for each quality. I know that's kind of a pain, I just have my doubts that the devs would implement something like that buffer in labs.

[nadnad]

Note that adding only one buffer slot will only help if you have a single mixed-quality science.

Otherwise, let's say you have red and green, each in normal and uncommon. Whichever kind of red science arrives first, normal or uncommon, will get put in the red slot, while whichever kind of green shows up first goes in the green slot. Whichever of the two remaining types and quality of science shows up first will get put in the buffer slot. Now the lab starts researching, using the science in the non-buffer slots. At some point, one of those sciences, red or green, will run out. Now, if that happens to be the same colour that's in the buffer slot, great; the lab can just instantly move the buffered stuff into the active slot and continue on. But if its the other colour, the lab will have to stop researching until an inserter adds a science pack of the other quality into the now-empty slot.

I think if would work if you had one buffer slot per science type. Even if you have more than two quality levels on the belt, as long as their are two of them in the lab, when one runs out it can continue researching using the second, while an inserter add new packs of some other quality into the now-empty buffer slot. The buffer slots would have to be filtered by science type, though.

This is true and not something I had considered. You would need ideally need a buffer slot per science type.

I think that, unfortunately, the easiest solution is to not mix quality sciences. If you're throwing quality modules into science assemblers to get that bonus % from quality,viewtopic.php?t=126882 < This discussion and others show how in nearly every case, it's better to use productivity modules.

However, if you have dedicated "quality science" assemblers that use leftover quality materials from your base, I guess another solution could be just having separate labs for each quality. I know that's kind of a pain, I just have my doubts that the devs would implement something like that buffer in labs.

I am aware that using productivity modules in science production is more efficient than using quality modules. There are plenty of other drawbacks to using quality in science production as well (e.g. interplanetary logisitics). Labs flickering off and on can be considered a minor inconvenience in comparison. The one part that of your post that I disagree with is that the solution to mixed quality science having drawbacks is to not use mixed quality science at all.

[computeraddict]

This discussion and others show how in nearly every case, it's better to use productivity modules.

However, if you have dedicated "quality science" assemblers that use leftover quality materials from your base, I guess another solution could be just having separate labs for each quality.

For sciences whose ingredients can't use productivity such as green and purple this is a real consideration. Idk if there are others off the top of my head.

[boskid]

Given that the scenario provided is highly artificial (high technology cost multiplier on a technology that has low unit consumption time and is usually already researched by the time quality is involved) combined with the proposed solution not working if more than 1 science pack gets involved, i am going to throw this to wont implement. Also there are some changes for 2.1 that may affect this.