summery

1: no-bell villages are not recording villagers or not being recorded properly.

2: villagers are not being properly simulated in an "unobserved" but loaded chunk within chunk simulation range.(see "villager tick radius" below)

3: villagers are not being properly saved to chunk for the aforementioned reasons; so technically they don't despawn. they just dont get saved before being unloaded from memory/simulation.

 

the villager tick radius

this is the radius centered on the player that allows villagers to function. villagers do not use the chunk simulation. for a villager to persist, they would need to be saved to chunk when they leave this radius.

 

its distance is about 70, but most probably 64...

 

i based its range on some testing where in i monitored my iron farms production while idling at certain distances under the farm for 20 minutes each test.

at 70, 4 ingots

at 55, 10 ingots

at 45, 20 ingots

 

the VTR generally explains why my village wasnt repopulating(rather than the idea of parasitic errant villagers refusing to give up their old beds...)

in reality, the village wouldnt repopulate because the villagers where not functioning/being simulated.

 

 

reassessment of the issues

 

villagers seem to be vulnerable to despawn if 

a: outside the players VTR

b: have no ownership to a bed under the influence of a bell

c: ownership to said bell bed was not accomplished while within the VTR

 

this is currently the only theory i have that explains the depop of bell bed villages and the depop of non bell bed villages.

this might also explain why children seem vulnerable. when they grow up while the player is "out of range" their child forms ownership of their bed dosnt transfer to thr adult and the adults new bed ownership isnt properly recorded to the village.

 

 

 

technical details

dorm-A(golem farm)

aprox center location with bell at center

250, 94, -32

beds & composters

31

 

dorm-B(breeder site)

aprox center location no bell

273, 14, 24

farm area dimensions

20x3x20

one bed, one composter

has path way to breeder stalls

four breeder stalls

dimensions 2x3x1

each with 2 built in cartographer tables

16 adjacent but not accessible beds.

 

my usual hangout(where i usualy am)

aproximate center

280, 10, -24

 

further testing

 

i opted to do a "hard reset" of dorm-A(the iron farm). this meant simultaneous destruction of dorm-A's bell, beds and worksites. then i left the villagers to mill about for a few sunsets while i went exploring around the area with the intent of keeping thier chunk loaded.

prehard reset the population was at 28. post reset the pop had dropped to 21.

i set the village back up and fed them and went back to mess around as soon as i was sure their pop was near limit of 31. pop was 29.

after two sunsets i returned and found the population had dropped back to 25. the kids had despawned apparently. so i resolved to hang out in the dorm and watch everything.

after 3 sunsets of watching and waiting, population was at max and fully adult. i have since let the place alone, periodicaly returning in the evening for head count. population has been stable now for about 5 sunsets.

update post hard reset.

after about 10 sunsets since post reset reestablishment, i noticed my iron farm production halted. upon bed check, i found pop had dropped to 30... 1 villager is now missing since reestab.

because of how golem production works, this leads me to beleive that there is still at least one errant "parasitic" villager trapped somewere around my base. 

 

 

old defunct stuff... aka errata that no long pertains to the issue at hand but rather than deleting, i moved down here.

 

parasitic villiager theory. mostly disproven by discovering the VTR.
 
villagers transplanted from a bell bed village to outide that villages boundries will persist in maintaining ownership over thier former bed despite range, inacesability if
a: they can not reach a new bed
b: are still in a the same simulated chunks as their home village
 
this means that a villager who falls into a cavern and is trapped will continue to occupy their previous bed despite being out of range and inaccessible... this appears to be a issue but unlike the despawn issue, i cant back this up 100%. however it would seem to be related over all or at least contribute to the issue of a village having trouble repopulating.
as this issue seems persistent for my iron farm, one idea is that, because all my villagers are lvl 0, that the game/village views them as ubiquitous. so these parasitic villagers cause the game to delete redundant villagers. "hes not here, hes in the cavern"
 
 
old assessment
if a villager dosnt have ownership over a bed within "range" of a bells influence and they are not within an unknown range of a player, they run the risk of being despawned.
more over, and this is fun... if a villager has some kind of seniority on the village roster, they can actually claim ownership of a bed within their home bells influence at ranges greater than 34 blocks... even if they cant reach it and even if another villager is closer and could actually use that bed... in that case, this unlucky "younger" villager runs the risk of the aforementioned despawning......

 

in summation, beds dont make villages... bell influenced beds make villages.

https://www.minecraftforum.net/forums/minecraft/discussion/3029315-this-is-why-villagers-disappear-and-despawn

and heres how to reproduce it

https://www.minecraftforum.net/forums/minecraft/discussion/3029304-curse-of-the-zombie-baby-farm

 

this has become such a pain i decided to go ahead and let you know. the irony that you dismissed my "villagers disappear" issue is not lost on me.