HQ supply chains not working in 1.02?

[Barry Eynon]

In version 1.02, it appears that HQ are not being rechecked for supply increases from other HQ as described in the HQ supply chain rules (and as always used to work). For instance, consider the Axis HQs stretching east across the southern front at the start of the 1942 scenario. (tried to put a picture in here but couldn't get one small enough to upload, sorry). The unit information and the supply display both show the Rumanian HQ at 10 supply, and the 2 German HQs further east at 8 supply, all of which they could get from local resource supply sources. But by the chaining rule the Rumanian HQ should offer 8 supply to Manstein which is 2 squares away, and raise it to 10 supply according to the chaining table. Then Manstein should be offering 6 supply to Bock, which should raise Bock to a 10 as well. Right?

This problem was particularly limiting in doing some of the user scenarios like 1938 Calm Before the Storm, where resource supply sources are much more spread out.

So either there's a bug or something I don't understand, appreciate clarification either way.

Thanks,

Barry Eynon

[Wushuki]

I don't think that such a chaining rule has ever existed. Headquarters can relay supply to each other, but this supply is only used by the HQ itself, it is not distributed further among units. So only the supply from cities, towns and fortresses is used to calculate the distributable supply of HQ.

If I misunderstood you, please use imageshack to post a screenshot and clarify what you meant.

[pzgndr]

I don't think that such a chaining rule has ever existed.

Correct. The manual fully explains the supply rules. HQ supply is based on their range from valid suppply sources, not other HQs. No chaining.

[Barry Eynon]

In the manual, on page 34, please read the two paragraphs beginning "Friendly HQs can also be linked in a supply chain over extended distances ... as well as any nearby units ..."

This may never have worked properly in SCGC, but you could certainly pass level 10 supply along a chain of HQs 4 spaces apart and have it distributed to the units at the front in earlier iterations of SC, in particular SC: W&W.

Here is the picture I was trying to upload, thanks for the suggestion to use imageshack:

scgc102supplyexample.png

[Hubert Cater]

IIRC in WaW if an HQ was in range of a 10 supply resource then it would maintain maximum supply, i.e. 10, but I believe this was changed in later releases to the current system found in the Global Manual as it was a bit too inconsistent. For example, along the border areas with the USSR, German HQs could maintain full supply and then it would suddenly drop as soon as it was more than 10 tiles away from a 10 strength city.

For the linking, the way it works, and this was the same as far as I can tell in WaW, is that if an HQ is calculated to have a low supply giving value, as determined by the table on page 33 then it can be boosted by a nearby HQ that has better supply.

For example (looking at 1939 World At War), let's say Germany controls much of the USSR (but not Kuybyshev) and is about to make a push onto Sverdlovsk from Gorki which has a supply of 5. If a German HQ is located 117,10 it is more than 5 supply tiles from Gorki and therefore considered out of normal supply. In this case the HQ would have a supply value of 5.

Now let's say we place a second HQ at tile 116,10 this HQ would have a supply value of 8 because it is within 5 tiles of Gorki. The HQ at 117,10 would still originally be calculated to be out of supply at it is more than 5 supply tiles of Gorki, i.e. it would have its maximum supply calculated to be 5, BUT since the neighbouring HQ at 116,10 would give any adjacent unit at tile 117,10 a supply value of 7, the HQ at tile 117,10 now will also have a supply value of 7.

Hopefully this makes sense,

Hubert

[Barry Eynon]

Thanks Hubert, your last paragraph matches exactly my understanding of how supply chaining is supposed to be working, for HQ that are far enough apart to not to have supply available from resourses. I do not believe this to be working correctly (or at all) in 1.02, here are two screenshots showing this:

scgc102example2.png Here the 3 German HQs are all able to get at least 1 point of supply from the Finnish HQ in Leningrad, hence should be promoted to 8 supply as per the table.

scgc102example3.png Here supply comes from Smolensk which is at 6, and this reaches the southern of the 2 German HQs and puts it at 8 supply. The northern HQ is out of range of Smolensk, but according to the table it should end up with supply 8, not 7, by chaining from the southern HQ.

I will mail you the save files from these two situations for you to examine.

But isn't this supply improvement also supposed to happen for HQ that are closer together, e.g. the HQs in the first screenshot I posted above?

Since you think the rule may have changed since WaW I'll leave it out of the discussion, though I know there was HQ chaining in WaW.

Thanks,

Barry

[SeaMonkey]

My apologies Barry, didn't mean to take this "off topic", but your posting underlines what I have been thinking about for a new SC. Sorry Hubert this SC gig is up, SC needs to evolve, not necessarily radically, but a new foundation needs to be laid and its got to be flexible for the initial layers to display the needed dynamics for gameplay.

There's a lot of good things that have come out of this run of version two of SC and I'm not advocating dismissing any of those features, but its time for an SC3 and it needs to be built from the ground up. I've got lots of ideas as I'm sure everyone else has also, but a little revision of the current engine is not going to carry this legacy any further.

Its time to start anew!

[Barry Eynon]

Returning to the topic of the thread, I think it's becoming clear what you're saying, Hubert - the table on page 33 does not get applied to the supply level a HQ gets based on another HQ, only to the best traced supply from a resouce. If that's how it's supposed to work now, then everything seems to be working correctly. Thanks for your patience!

-Barry

[Hubert Cater]

No problem Barry and glad to hear it is all cleared up now

[Hubert Cater]

SeaMonkey, I'm intrigued as I am always considering improvements to the game

Care to expand a bit as I am not sure what specifically you are referring to here in terms of the shortcomings.

Thanks,

Hubert

[SeaMonkey]

Of course Hubert, it would be my pleasure and I was just poking around with generalized statements to see if the World's best developer was receptive to a new SC vision. It seems you are, as I'd hoped. So let me gather my thoughts, some of this stuff you've heard before, perhaps a little guidance would help.

I intend on initially examining the foundation of a good wargame, my thoughts are that will encompass the units and the map(pieces and playground), your thoughts?

Am I on the right track?

Oh.....again....sorry for taking this thread off...I'll start another..I mean....

are we serious here Hubert? Are you sure you want to embark upon this new path?:confused: I mean, I'd rather play than post, with my limited gaming time. If this is it for SC...I'm OK with just playing, love the game!:cool:

[Hubert Cater]

I'm always serious

To clarify though, I unfortunately don't make enough on these games to promise anything concretely, but if all goes well in the next little while then I would likely embark on a project that will hopefully make a lot of people happy

Essentially the desire is there but time and money are always an issue. If in the meantime you've got something to give me to think about now is as good a time as any. Reason being that since development cycles can take quite a while, now is really the time to be pointed in the right direction.

Definitely feel free to start a new thread or send me an email in private and we can discuss your ideas/suggestions further.

Thanks,

Hubert

[SeaMonkey]

Understandable, no promises necessary, just the inclination of pursuit, it'll be fun and thought provoking. If you don't mind, I'd like to keep it in the open, that thing about more heads are better than two and I know we'll get substantial contributions from the forum.

Now I have miniscule knowledge of Eiffel, or programming for that matter, but I remember you saying something about the different layers of SC and how the processors slow down by computing the complicated algorithms associated with the supply, communication, terrain, weather, etc layers that are incorporated into SC.

Is that a correct assumption? If so, then we'll want to emphasize minimization of those layers and make more editable features inherent into each layer. For instance the initial building platform is the map, defined by the tiles, hexes, coordinates, whatever, representing the 3 dimensional playing surface(air, surface=sea & land, and the submarine realm, interspace). What else should be considered for the foundation, the initial layer?

[Hubert Cater]

Just trying to catch up here and no problem at all SeaMonkey, out in the open is good as I agree others can share their ideas as well.

Regarding the different layers, there are two real issues, one is the amount of data layers and the other is the amount of graphics that need to be drawn per layer.

Unfortunately putting the data all into one layer doesn't help that much in terms of speed as there is still a lot to process... really, and if we are talking about the AI, it is just the amount of data and the number of processes such as calculting supply, best attack poisition, running through multiple attack possibilities looking for the best one etc., that slows the AI down, as otherwise the search times for the data is pretty quick.

The biggest issues is the increase in map sizes that slow the AI down because even let's say doubling a map size from 256x64 to 512x128 doesn't double the data, rather it quadruples it!

[SeaMonkey]

Crap Hubert! Through all these postings for SC3 ideas I kind of completely blew off the AI considerations, my bad. I know how important an AI is to SC, it has to be competent, to a degree, so now I'm going to have to rethink things and keep that in the back of my head.

Possibly you could comment when we start discussing things that you'd naturally recognize as an unfriendly AI feature, guidance helps moves things in a more pragmatic way.:cool:

[Hubert Cater]

No worries SeaMonkey and I could comment on possibly unfriendly AI features but the problem is that sometimes I may inadvertently shut down a good idea that might come along as a result of the initial discussion... even if at first it is not necessarily AI friendly.