Max request distance for chunks

What is the farthest that a CF, refinery, or fabber can pull a chunk down a conveyor? I've seen people mentioning that there is a limit, but not what it is.

 

The problem is that it depends upon how many other crafting blocks (Factories, Refineries, Silos) and also connected to that line, upstream. Each one shortens the distance. And looping the conveyor massively shortens the distance and can create more dynamic problems if the crafting AI decides to send stuff around and around the loop. It might even be unpredictable currently, with the problems we've been seeing with long conveyor crafters...

You can tell if it's gone too long to reach a Silo because the little white light won't flash on those Silos when it's trying to craft. But well before that you'll get the 'stutter', 2 second hangs, when trying to edit a tech that's crafting: https://forum.terratechgame.com/ind...ting-with-lots-of-silos-causes-stutter.10309/

 

So nobody has a number because it all depends on the conveyor setup?

I get the crafting/editing freeze all the time but I haven't seen the flash stutter.

 

I suspect the number to be 96 / 127; however I have *NOT* counted it out.

References: https://forum.terratechgame.com/ind...ting-ais-read-distance-along-conveyors.13404/ Where I determine the problem to be max conveyor read distance.

https://forum.terratechgame.com/index.php?threads/zeds-tech-zoo.9126/page-4#post-79405 Contains a Grabbit of the tech where the number of conveyors could be counted. (IIRC when requesting the fabrication of Hawkeye blocks materials from the first three silos (directly across from the Hawkeye Fabricator) will not furnish resources.)

 

I Think 'Stutter' is just a short freeze, now, in the borderline designs between no freeze and full 2 second hangs. Or else the game was changed little since that initial report and there aren't regular stutters any more.

@Twigjack - you could test this out for us: see what the theoretical maximum conveyor length is. i.e. for a single Fab calling down a long conveyor with only 1 silo on the end.

You could also then see how/when that length is shortened by other crafting blocks on it.

 

I tried this in creative and the call distance cap doesn't seem to exist there. I couldn't recreate the issue with inactive silos.

1. Built the first line out from the anchor- all the silos active.
2. Added the hairpin and came all the way back- all the silos active.
3. Added loop and more Fabbers- all silos active.
4. Added second hairpin and 4-stack silos with filters and exit conveyors- all silos active.
5. Ran multiple fabbers on repeat- all silos active.

EDIT: line 5, changed 'silos' to 'fabbers'

 

Heh, good to know.

But I wouldn't expect Fabs to interfere with each other. I think if you start adding Component Factories and/or Refineries, instead, it should shorten the active range...

 

I'll try that when I have some time.

 

Ok, so I went back and marked out distances with GSO line blocks, then marked every 25 from the HE fabber with construction lights. The track is about 220 conveyors long and I feel that's enough that even the most spaghetti of factories should stay within it.

1. Added CF 1 to 6- all silos active
2. Added CF 7- call distance dropped to 150
   1. 220 to 150 is a drop of 70 conveyor spaces.
   2. I am more inclined to believe that CF number 7 is where the limit begins, not a drop.
3. Added CF 8- call distance dropped to 117, a loss of 33 conveyor spaces. Strange number imo.
4. Added CF 9 (I know nobody uses 9 but this is for science)- call distance dropped to 103
   1. 117 to 103 is a loss of 14 conveyor spaces.
   2. The numbers remain strange and inconsistent.
5. Added CF 10 (SCIENCE!)- call distance dropped to 89
   1. 103 to 89 is a loss of 14 conveyor spaces.
   2. Eureka, a pattern!
6. Added CF 11 to confirm pattern- call distance dropped to 81
   1. 89 to 81 is a loss of 8 conveyor spaces.
   2. Boohiss, not a pattern after all.
7. Added CF 12 because it would fit- call distance dropped to 77
   1. 81 to 77 is a loss of 4 conveyor spaces.
   2. 33, 14, 14, 8, 4???
8. Returned to 8 CFs (because that is the most a normal factory would need) and added one lvl 1 dongle- call distance dropped to 102
   1. 117 to 102 is a loss of 15 conveyor spaces.
9. Added second lvl 1 dongle- call distance remained 102
10. Added lvl 2 dongle- call distance remained 102
11. Added second lvl 2 dongle- call distance remained 102
    
12. Added lvl 3 dongle- call distance remained 102
    
13. Added second lvl 3 dongle- call distance remained 102

I think we can conclude that 102 is a good number to keep in mind in most instances. The lack of a clear pattern makes me a little bonkers, though.

The forum keeps erroring out when I try to upload the snapshot, so here's a Google Drive link: https://drive.google.com/file/d/1BbRJTBh0s0k57aSgE8fdn-16OGS3KxIK/view?usp=sharing

 

Heh, you managed to make a snapshot image that's *just* over the forums (hidden) 1MB upload limit, congrats.

Good work with your careful measurements and observations. But I'm not surprised that there's no simple numerical relationship between length reduction and CFs added. It's not going to be a hard coded limit, more the result of the expansion of the total number of combinations of possible crafting routes as the complexity of the system increases...

So from the look of your snapshot image, a major things you've not spelled out above is that all the Component Factories are on a loop together. Then the conveyor length is attached to that (starting from count 1?).

I think you'll find that if you break the loop open you'll be able to add many more Component Factories before it starts limiting the call distance. I'd be interested to know what the limits are on an open conveyor length, by comparison.

Are you kidding? Lol. There's plenty here who'd call that modest to small. I like to go minimal and that's 4 to 8. Many new (crafting enthusiast) players will be oblivious to the problems that 'more' can cause and just stick more on to get it done, enjoy building big, or whatever.

Also, I don't think you've accounted for the conveyor length limiting factor of the Silos themselves (currently arrayed along the measurement line). Like quantum physics, the act of making a measurement with these affect the result.

I'd also be interested to know for sure if 3 node Silos are more limiting that 1 Lotsa Silo, or if it works per block length adjacent to conveyor.

 

My count started at and included the conveyor adjacent to the HE fabber input, which I alluded to but did not spell out. I'm not really sure how one would not have the CFs on a loop, since in my experience, there's no way to make the ones with higher dongles not try to build the basic pieces, then have nothing building the highest thing you need. (Maybe I was just unlucky)

Looks like I've found my next SCIENCE!

Jeez, how do they cram all their materials into the call limit?

I used node silos so I could get a more precise measurement. As far as I can tell, like dongles past the first, there is no interaction with the number of silos and call distance.

 

1. Exchanged 24 node silos for 8 Lotsa silos- call distance remained 102
2. Broke loop and replaced 8 Lotsa silos with 8 CFs (8 total CFs on entire call line)- call distance returns to indeterminate
   1. CFs call fully twice as many resources as they need and ignore half. (bug report)(bug report)
      1. Set Pacemaker from Turbo to Normal, no change in behavior.
   2. Silos faff about with resources. (like this bug but constant and on normal)(here we go)
   3. At least one CF calls resources, then cancels the recipe, destroying them.
   4. CFs flicker through recipes, calling and destroying materials. (bug report)
   5. After 4 tier 1 components and 1 tier 2, all CFs seem to give up and cancel all recipes.
   6. Fabber recipe still active.
3. Noticed Refinery is not downstream of CFs. Move to upstream, though there are no unrefined materials in the system. (bug report)
4. Manually canceled fabber recipe and restarted it.
   1. CFs continue to call excess resources.
   2. CFs without dongles call and refuse all Plumbia Ingots to build Plubonic Greebles.
   3. Same CFs have the Luxian Crystals they need but continue to call more.
   4. No further progress is made.
5. Manually cancelled fabber recipe.
6. Removed all silos downstream of CFs and inserted dropper/reciever setup to prevent clogs from extra materials.
7. Restarted recipe.
   1. Process hangs because of CFs producing Blast Caps and Plutonic Greebles refusing some materials.

I think I'd rather deal with just the call distance bug.

 

Extra Sciency Bonus Round!

1. Removed a lot of node silos- call distance > 220
2. Added 1 CF (total 9)- call distance > 220
3. Added 1 CF (total 10)- call distance > 220
   
4. Added 1 CF (total 11)- call distance dropped to 217
5. Added 1 CF (total 12)- call distance dropped to 191
   1. 217 to 191 is a loss of 26 conveyor spaces.
6. Added 1 CF (total 13)- call distance dropped to 174
   
   1. 191 to 174 is a loss of 17 conveyor spaces.
7. Added 1 CF (total 14)- call distance dropped to 151
   1. 174 to 151 is a loss of 23 conveyor spaces.
8. Added 1 CF (total 15)- call distance dropped to 137
   1. 151 to 137 is a loss of 14 conveyor spaces.
9. Added 1 CF (total 16)- call distance dropped to 112
   1. 137 to 112 is a loss of 25 conveyor spaces.

• Now at twice what I consider a normal amount of CFs, call distance is approximate to half that number on a loop.
• Useless information at the moment, because it won't craft top tier items.

 

It would also be fixed by patching the material refusal bug and making CFs smarter.

 

Yes, @Zed, you've previously suggested that this conveyor 'look-back' lenght be increased. But what I'm hoping for in this thread is some more numerical data on how and when this length changes. Because it's clearly not a fixed limit that could be trivially extended, ever, right?

 

I'd very much like to know if your setup was experiencing the old bug or the new one (particularly with regards to all the resource deletion that you report after that). Did you (not) make a point of removing the Fabricator after each modification to the Component Factory (and Refinery) layout?

 

I somehow missed that detail in both reports. I'm gonna be a little put out if this bug is that easy to work around.

1. Returned to 8 CFs, 2 at each level.
2. Removed and replaced fabber.
3. Fabrication succeeded.
   1. Recipes in CFs flickered around once, I didn't see if resources were destroyed.

Yup. Same bug as reported before. I'm annoyed at my lack of reading comprehension.

 

I would disagree with this assessment given that there is clearly a governing initial variable somewhere from which these proven permutations are derived to reduce length along the way. Unless @Greg / @zanzistar share information to the contrary most likely this variable could be modified with an [ Variable (x) ] * 1.50

This type of modification would be fairly simplistic and permit the creation of larger bases without expending significant coding time on Crafting which is most likely "Done and Dusted" until post 1.0 release Given a scant 11 full weeks remain now and the end of "Summer 2018" Within which MP / Console / Control Schema *Must* take priority.

 

It would be interested to know if the initial variable was arbitrary or intentional. If arbitrary, yes a quick change would be helpful. If not, the reason needs to be considered. The variable may have been set to reduce constant lag.