Page 1 of 2
Posted: Tue Feb 26, 2008 10:00 pm
by arsan
(...)
On the rest of the fronts, USA have been a little active on east of Kentucky but no more: nothing around the Mississippi, little on Missouri and nothing on Texas.
That last is particularly strange: with the last patches the vanilla AI is very active on Texas, trying to overrun the place with all the forces of the far west boxes (i fear this happen since Gray changed the links

maybe the travel time between boxes is too low??)
(...)
[color="DarkRed"]This is a thread split off from Leader Mod for 1.09 (
http://www.ageod-forum.com/showthread.php?t=7539) to let the jumplink-discussion have a life of its own.
Regards, Rafiki[/color]
Posted: Tue Feb 26, 2008 11:34 pm
by Gray_Lensman
deleted
Posted: Tue Feb 26, 2008 11:37 pm
by lodilefty
Gray_Lensman wrote:Concerning the travel times to/from the Far West boxes:
There wasn't much choice between types of Transition JumpLinks to work with
1) takes 12 days
2) takes 48 days
3) takes 96 days
I wish there was a choice between 1 and 2. However, since you are pointing this out about the AI using the shorter time periods in its activities, I may revamp the Far West Transition JumpLink(s) to inhibit this AI behavior, since historically, the Far West really didn't have that much of a role in the Civil War. The problem is that I will have to change the 12 day links to 48 day links, which in a lot of cases is way more than the time necessary to make the transit.
Gray,
Have you tried differrent terrain in thos OMB's? Maybe we can slow things down by making them mountains or something.....
Posted: Tue Feb 26, 2008 11:44 pm
by Gray_Lensman
deleted
Posted: Tue Feb 26, 2008 11:58 pm
by arsan
Hi
Gray,
I didn't know there was only those choices
Still, it should be good to try the 48 days options on some links... the USA ai seems to use the off map boxes like a highway to Texas, "backdoor of the confederacy"
Another example of the problem: to go from Houston to Dallas, Arkansas, or Missouri the shorter way for a stack (the one taken by default as you drag the unit) its via... Tucson, Arizona!
A mixed stack needs 101 days to go from Houston to Fayeteville (AR) by the "Direct" route over the map.... but only 70 taking the roundabout Tucson off map box "highway"
And sorry for the off topic...
Regards!
Posted: Wed Feb 27, 2008 12:56 am
by Jabberwock
Gray_Lensman wrote:I really wouldn't like to make the terrain something different from what it actually is, and in the case of Transition JumpLinks, it wouldn't matter, they don't take terrain into consideration anyway. They're basically a "teleportion" JumpLink between 2 areas (whether adjacent or not). So that's not a workable solution to the problem. Has anyone else observed this AI behavior that Arsan is pointing out in either the vanilla scenarios or using the "Leader MOD"?
I have. Both vanilla and LM. Athena goes nuts over Tuscon as either side. As the Confederates, she starts sending ranger raiders through all those jumplinks.
When I was testing your canal work, I noticed that the movement times were often cut in half from the base twelve days. I wonder if something like that may happen here as well?
Rafiki / Korrigan - can we branch this off into a new thread?
Posted: Wed Feb 27, 2008 1:36 am
by Gray_Lensman
deleted
Posted: Wed Feb 27, 2008 1:43 am
by lodilefty
Jabberwock wrote:I have. Both vanilla and LM. Athena goes nuts over Tuscon as either side. As the Confederates, she starts sending ranger raiders through all those jumplinks.
When I was testing your canal work, I noticed that the movement times were often cut in half from the base twelve days. I wonder if something like that may happen here as well?
Rafiki / Korrigan - can we branch this off into a new thread?
I also have a mod I'm working on, and Tuscon is a magnet for Athena....
re: canal movement: was the shorter time for ships moving? The ship movement through these jumplinks is significantly faster....
Posted: Wed Feb 27, 2008 1:47 am
by Jabberwock
lodilefty wrote:I also have a mod I'm working on, and Tuscon is a magnet for Athena....
re: canal movement: was the shorter time for ships moving? The ship movement through these jumplinks is significantly faster....
Yes. That explains that.
Posted: Wed Feb 27, 2008 2:02 am
by Gray_Lensman
deleted
Posted: Wed Feb 27, 2008 2:56 am
by Jabberwock
Gray_Lensman wrote:I don't recall any reworked canal JumpLink(s) in the vicinity of Tucson.
However, naval movement thru Transition Links is half the time of land units.
Gray_Lensman wrote:Also, I need region numbers in order to check out these areas. I just don't have the extra time to waste looking them up individually by name.
1095 South West
1094 California
1092 Rockies
1034 Decatur KS
1036 Platte KS
1038 Prairie KS
1039 Plymouth KS
1041 Colby KS
1044 Dodge KS
1047 Smith KS
1050 Grand Bend KS
1053 Hunter KS
1057 Fenimore KS
Maybe it could be the number of links that makes it a magnet? Half a dozen of the KS links never get used.

Posted: Wed Feb 27, 2008 3:47 am
by Gray_Lensman
deleted
Posted: Wed Feb 27, 2008 4:32 am
by Jabberwock
Gray_Lensman wrote:Okay
Here's a .ZIP that actually contains 2 (fast-install) style .zips
One of them is the Experimental version and the other is the Regular or as it is now.
They both contain the 40 regions along the Western edge of the map from Wisconsin all the way down to Texas. All JumpLink(s) to/from these border regions have been changed to the 48 day JumpLink type. When/if you want to revert back to the 12 day JumpLinks, just run the other (Regular) fast-install .zip
Please let me know if you actually see any real difference in the AI behavior using the 48 day Experimental set.
Jabberwock:
All the regions you listed are included in these files. I think you are misnaming Transition Links by calling them Canal Links. They're not Canal links per se, but they were used to make the canals work correctly. In the ExMap program, they are called
Transition Link
Long Transition Link
Very Long Transition Link
They function like a "teleportation" connection, and terrain (movement wise) actually does not matter to them unless it is a completely illegal move for a unit to make. Example: deep draft naval vessels cannot move into shallow water, no matter what the movement cost.
Regards
.
Didn't call 'em that (recently ... in this thread). I'll try to remember for future posts. Thanks for posting this info, it is interesting.
I'll test these, and mb run a few other tests of my own. It may take a few days, working on something else.
Posted: Wed Feb 27, 2008 4:37 am
by Gray_Lensman
deleted
Posted: Wed Feb 27, 2008 7:59 am
by arsan
Thanks a lot Gray!
I will try them this afternoon... after work
Regards!
Posted: Wed Feb 27, 2008 10:13 am
by Rafiki
Jabberwock wrote:Rafiki / Korrigan - can we branch this off into a new thread?
Yup

Posted: Wed Feb 27, 2008 8:08 pm
by arsan
Hi Gray!
I´m probing the new links you have posted but they seem to not work all right.
I got the same movement time with the vanilla ones that with the experimental long ones.
A cav unit can go from a texas border region to Tuscon or Laredo in just 9 days, like before...
I have even tried starting a new campaign but no to avail...
I have already double checked that i have made the installation ok (looking at the files's time stamp).
Any idea?
Regards!
Posted: Wed Feb 27, 2008 9:45 pm
by Clovis
trying them with my mod.
Posted: Wed Feb 27, 2008 11:08 pm
by Jabberwock
arsan wrote:Hi Gray!
I´m probing the new links you have posted but they seem to not work all right.
I got the same movement time with the vanilla ones that with the experimental long ones.
A cav unit can go from a texas border region to Tuscon or Laredo in just 9 days, like before...
I have even tried starting a new campaign but no to avail...
I have already double checked that i have made the installation ok (looking at the files's time stamp).
Any idea?
Regards!
I didn't give him the region #s for Texas, those links pre-existed the issue. We are testing from Tuscon (which should be spelled 'Santa Fe' :niark

to points north, those were the new links.
Posted: Wed Feb 27, 2008 11:32 pm
by arsan
Jabberwock wrote:I didn't give him the region #s for Texas, those links pre-existed the issue. We are testing from Tuscon (which should be spelled 'Santa Fe' :niark

to points north, those were the new links.
Ups!
I think the southern regions were the most problematic.
The region link from Tucson to Dallas ans south to the gulf are the real backdoor entrance where most of the USA forces enter and go rampaging all around Texas

leure:
Hope they can be included also on the fix
I will try to check the others...
Regards
EDIT: Jabber, it seems Gray included also the Texas links and a lot more: check the attached rar. Inside there are 40 rgn files!
Cheers
Posted: Thu Feb 28, 2008 1:07 am
by Jabberwock
The link from South West to Kaufman has existed as long as I have been playing. Do you think it should be longer? It is too bad we don't have El Paso and San Antonio (as buffer regions) to work with, but we don't (and it would be more effort than it's worth to add them). I think it is important for balance to keep that one as is, with regards to the Sam Houston event.
The new links I was aware of are the ones from South West to points north. I wasn't aware of any new additions in Texas. The USA didn't used to be able to send all its west coast and Kansas troops easily to South West. The CSA could reach South West, but it didn't used to be able to send all its Texas troops easily from South West to Kansas and the other places.
South West should have links to California (the historical premise for the CSA campaign there), Rockies (used historically by USA), and Texas (used historically by CSA). Those are all in Foote. I'm not so sure about Kansas.
I think Athena may be having a problem with the sheer number of links leading from South West, causing her to overvalue it, but it may be the convenience of the links. We're testing the second hypothesis right now. I will test my hypothesis as well, if we don't see better results from the first test.
Posted: Thu Feb 28, 2008 3:35 am
by Gray_Lensman
deleted
Posted: Thu Feb 28, 2008 4:38 am
by Gray_Lensman
deleted
Posted: Thu Feb 28, 2008 10:02 am
by arsan
Thanks to you, Gray for taking care of the problem!
Hope its easy to fix for Pocus.
Regards
Posted: Thu Feb 28, 2008 1:57 pm
by simovitch
Playing the Confederates in the April '61 campaign I was noticing this behavior as well, large formations headed into the Rockies and Southwest and raids into Texas and Sherman is sitting in Tuscon with a rather large Corps. - glad it is being fixed.
Also - 2 areas in Texas have 'railroad' in the tooltip but no RR graphic: Cotton and Sumpter. Playing 1.09 with latest RR mod.
And just wondering if The Rockies and the southwest should be "Civilized"? that doesn't sound right to me.
Posted: Thu Feb 28, 2008 3:30 pm
by Pocus
Can you post your testbed Michael, telling me the test region and also the regions ID (and nature of link) of 3 connected regions?
Posted: Thu Feb 28, 2008 9:57 pm
by Gray_Lensman
deleted
Posted: Fri Feb 29, 2008 12:09 am
by Gray_Lensman
deleted
Posted: Fri Feb 29, 2008 5:59 pm
by Pocus
Thanks Michael, I'll check that.
Posted: Mon May 05, 2008 8:14 pm
by Turbo823
I've also seen this in the AAR game I am posting in the AAR forum. In that game Union Athena put a full division and Halleck in the Tucson box. A strange move as she left all of Arkansas wide open and it allowed me to move price into Missouri. I posted a screen shot of this there.
Maybe a solution here is to remove the strategic VP for Tuscon?