So, after re-reading This thread yet again, I feel compelled to post, yet again, a supply suggestion.
First off, I want to make it known that for those of you who don't know, it was said by a Paradox representative (Kallocain I believe, but due to 1/2 of these forums, including Search, not working for me, I can't find the thread to link it) that given the choice, Paradox would have loved to give every single division it's own "supply trace", which would solve most of the "oddball supply" issues this game seems to have, as well as just about ALL of the misunderstanding issues of the supply system.
However, said post said that such a system was simply unfeasable, as it took the already complicated set of calculations and exponentially expounded them.
Any suggestion to the supply system would have to take into account that the existing system was chosen because, among other considerations, it allowed the computer not to kill itself trying to pull it off.
Often I see the "make the HQs the supply hub", without further elaboration; the "suggestor" then seems to imply that this wonderful idea would solve everything. The only universe where making broad, top-level assertions with no follow up is successful is the Dilbert comic.
Not that EVERYONE who has said this has ever not expounded, but... the full-hearted attempts are few and far between.
And as such, I post this idea. Let's see how it goes...
Main Step 1: Leave the "single supply depot per connected landmass" basic concept.
--- This depot would be the "supply hub" of any division not attached (in some way) to a 6* HQ, or any division who is not land-connected to it's 6* HQ. Also, it would remain the hub for all Naval and Airborne supply requests (see later on for simple suggestion regarding those).
Also, and most importantly, this hub would also be the "servicer" of 6* HQs.
As this system is "already in place", the reprogramming for it, and expected necessary computer cycles, should be minimal.
Main Step 2: Convert 6* HQs to be localized supply depots for all land-units attached to them (including themselves).
--- Of course, I have no access to the source code, but I would assume that changing the portion of the supply-draw code from "get_supply_route($main_depot)" to "get_supply_route($HQ)",
having a "$supply_from" flag per operational division that defaults to "$main_depot", gets set to "$HQ" when attached or moved, and re-set to "$main_depot" when de-attached, would causes minimal computing time, and minimal code changing.
---The supply draw of each entity on the board is unchanged, save for the 6*HQs, which will have a value of sum($attached_land_divisions + $self). The draw pirority of the 6*HQ, however, is LOWEST of all types of draw; I.E. for each province that is determining what is "drawing" from it, it figures out all requests BEFORE the 6*HQs request (to prevent HQ stockpiles with out-of-supply divisions, and also to prevent supplies ping-ponging between HQ and divisions).
--- rebased air and sea units often end up out of supply upon arrival while the system tries to catch up, which is poor practice, both historically AND programatically.
--- All ports and airports have a draw of their own, equal to 5*maxlevel (value subject to debate, but calculation is based on how large the port is when fully repaired; strat-bombings should not affect draw requests). However, while they have a draw, said ports/airports continue to NOT use supplies themselves.
---As supplies do not "decay", the overall effect of having this would mean that the port maintains a basic supply while unusued and out of the way; and if an unused port/airport is in the middle of a supply highway, once it gets it's desired value, it will then pass through 100% of future supplies that hit it's province.
When a 6*HQ moves a province, the supplies it takes with it are based on the following calculation:
A) figure out the supply cost of the 6*HQ (and included brigades, if you like to put AAs with your HQ's ;p) by itself.
B) figure out the draw value of all other units (divisions, fleets, wings) on the province, using the current system's values.
C) Figure out the value of sum($attached_land_divisions + $self)
Call the total supplies/fuel on the 6*HQ's province as X.
if X <= A, then the moving HQ takes all the supplies in the province with it.
if X > A and X < (A + B); OR if B > C (for example, if you just created a 6*HQ and haven't attached anything to it yet), then the moving HQ takes A supplies with it, and leaves the rest fot the divisions that are still in the province.
if X > (A + B) and X < C, then the HQ leaves behind B supplies, and takes the rest.
if X > C, then the moving HQ takes C resources with it, and leaves the rest, which then follow the game's existing logic of "trying to get back to the $main_depot".
This amount is INFRASTRUCTURE INDEPENDANT; regardless if the province is a 1 or a 10, the 6*HQ + supply depot concept means that the 6*HQ would have sufficient trucks/horses/etc. to ship an amount of supplies = C (which is already beyond the game's abstraction anyway), and as such would "move with" that amount of resources anyway, just like a regular division brings it's 30-day supply when it moves.
If a 6*HQ is disbanded or destroyed, all atatched units immediately revert their "$supply_from" flag to $main_depot. The supplys that were with the 6*HQ remain in that province, and will try and get back to the "$main_depot" as well (if possible)
As each "unit" already has a self-contained calculation for it's supply draw, and each province already follows "supply rules" based on what "draw requests" it receives, these changes, I would assume, would result in a minimal increase of processor power, as the number of "new" equations is only a linear factor of the number of 6*HQs on the board, with a slight overhead of each time a unit is attached or removed from a 6*HQ (to reset that unit's "$supply_from" flag).
What this would result in, with the game's abstraction level:
---When a 6*HQ moves, all subordinate units will start "drawing" from it's new location. This may (depending on current supply lines) result in them ending up in short-term supply deficit. I personally have no trouble with this, as moving "supply depots" in the middle of heavily supply-intensive times would have the same effect IRL, but your opinion may vary.
---Moving a 6*HQ into a specific area in preparation for a future offensive would have the passive effect of causing a "supply buildup" to automatically begin.
---The 6*HQ (or more specifically, the 6*HQ's province of residence) would become a major Logistical Bombing target. Much care must be taken to prevent air-raids on it, just like Log-bombing the current enemy's main depot (or capital if land-connected). However, if the 6*HQ's province is decimated, it can move to a new one, so single-province megabombing would not be a total-army crippling factor.
---It would now be VERY BAD to lose a 6*HQ to a concerted blitz/paradrop effort. Guarding them with some actual combat units would be a good idea, and targetting them would (maybe) be a viable strategy
--- Poor infrastructure would still stall a large offensive, but the impact of unit movements that aren't directly in line with a previous supply path would have their impact on the supply route cut from (from position to $main_depot) changed to (from position to 6*HQ); assuming the 6*HQ is even remotely near the units, this turnover time would be substantial (if a 6*HQ unit is 1/2 the distance to a division as the $main_depot is, then each division movement not along the current supply path would take 1/2 as long to reconcile).
---Small operations will likely not end up in "out of supply" territory given a small buildup time (logistical bombing notwithstanding), but large operations (I'm looking at YOU, Barbarossa) run the risk of having the 6*HQ run out of even it's dramatically-inflated 30-day supply value, causing a stall, or potential necessary retreat. Which is exactly how "the logistics of extended campaigns" ARE supposed to be challenging, IMO, rather than issues like "I moved 5 divisions 1 province north, now they need 40 days to get fresh supplies from Berlin"
---Bold individuals might push their 6*HQ REALLY close to a front line to minimize supply time. This would be dangerous yet effective; also, whichever divisions are closest to the "pushed forward" 6*HQ would be getting the equivalent of "preferred supplies", which is another suggestion I see often.
---Trying to mix 6*HQs in the same area may result in a "draw war" between them. This should be left up to the player; you are a leader of a WW2 entity, doing something as nuts as giving multiple 4-star generals command over the exact same region is something YOU should avoid ;p