Leader bug and WitPDecoder

[Cpt Sherwood]

Now that Woos has given us the excellent WitPDecoder we can see things that we couldn't before. I noticed that when I wanted to change leaders and used WitPDecoder to help me, I was given leaders to select that were all ready in command of another unit. Here is one picture. LTC T. Kelly is commanding the AO Brazos. I asked to change the leader and as you can see, the first three on the list are commanding subs somewhere in the Pacific. I know it also happens with war ships. I have not really looked at gound commanders yet, but I will be. If you do not use WitPDecoder to anayize your leaders and game, you are taking a risk of problems.

[jwilkerson]

Yup, there are a fair number of "duplicates" (leaders used in more than one unit) in the database and until these are corrected they will cause problems. The system was designed to handle only one use of a leader at a time, but the editor has no check to enforce this.

[Cpt Sherwood]

But I think you are missing the point. I am trying to select a new commander and some of the choices given are commanders that are commanding another ship. Why is the program letting me chose a leader that is all ready commanding a ship? Most interestingly, I have made most of my leader changes in the subs. Now I find that the first three commanders I am presented with are currently commanding subs.

[LarryP]

Where is this WitPDecoder located?

[Cpt Sherwood]

You will need to read this thread. WitPDecoder thread

[michaelm75au]

Out of curosity, what version of WITP are you using.

[Cpt Sherwood]

This would be version 1.804. The senario is BigBs 192.

[qgaliana]

One problem I've noticed seems to be when I create a TF from an existing TF and move the flagship out. I (sometimes, not sure if always, may depend on being docked or not) wind up with the same commander in charge of both TFs.
 
Once the leader winds up in 2 places, it seems to be difficult for the program to track him correctly. In particular, it looks like when you disband the TF or assign a new commander, the old leader gets flagged as available and the capability isn't there to check to see if they are still assigned elsewhere.
 
After some brief experiments, the only way to tidy this up seemed to be to remove the leader in question from every assignment. Then when you assign them somewhere they seem to finally get correctly tagged as 'busy' and stop showing up in the list of available leaders. (NB - I talk about tags and flags, but I don't know what's going on under the hood, could just as easily be a complicated pointer tree linking everything. I'm just metaphorically describing the observed behaviour).
 
And yes, couldn't track this until witpdecoder, after which I found some of my more choice commanders in charge of 3-4 subs.

[cverbrug]

Hi all,


Now we are talking of leader problem. I am playing a pbem game and i lost my naval HQ leaders from P-Z. I have no idea why and how (gues it is related to installing a patch on an already existing pbem game). None of the leaders with a name starting with P-Z is there anymore. Even the units having one of those leaders assigned showed 'staff officer'. It only occured with the naval HQ leaders. The others are fine.
Even Yamamoto is gone.

Has anyone ever observed the same?

chris

[rtrapasso]

ORIGINAL: Cpt Sherwood

Now that Woos has given us the excellent WitPDecoder we can see things that we couldn't before. I noticed that when I wanted to change leaders and used WitPDecoder to help me, I was given leaders to select that were all ready in command of another unit. Here is one picture. LTC T. Kelly is commanding the AO Brazos. I asked to change the leader and as you can see, the first three on the list are commanding subs somewhere in the Pacific. I know it also happens with war ships. I have not really looked at gound commanders yet, but I will be. If you do not use WitPDecoder to anayize your leaders and game, you are taking a risk of problems.

i thought WITP was DESIGNED with producing "clones" - and i am not sure if this is a bug. i will consult with higher-ups...

[rtrapasso]

OK - i am told we need to look into it - but you need to provide a SAVE file (nonfinalized) that happens immediately before the bug. If you have one, let me know and then we can proceed.

[Cpt Sherwood]

I have all the save files you will need, it is always present now. I need your email address and I'll send one to you with a write up.

[Cpt Sherwood]

Harsher message than I should have sent, sorry.

[rtrapasso]

Capt Sherwood -

Sorry - i didn't see your post sooner - i check the support forum once or twice a day... somehow i missed your prior post...

send save to rtrapasso @gmail.com (no space in address - blank put in to foil spambots)

Give details of where the leader(s) vanish - and as mentioned in previous pm, i will need passwords (if PBEM).

Cheers,
Bob T.

[rtrapasso]

bump

[Cpt Sherwood]

You will have to wait until I gete home from work. At least 6 more hours

[rtrapasso]

OK - just wanted to make sure it didn't get buried... [8D]

[Yamato hugger]

From the sounds of it (my un-educated impression) its going to take a data-base change to fix this. Keep in mind, I dont see the code or the files, so Im as in the dark on this as the rest of you, but sounds like a flag needs to be put in the leader file and when he is used / dead, the flag needs to be tripped and every place in the code that it assigns a new leader it needs to check this flag. At least thats what it sounds like to me.

[Cpt Sherwood]

I believe you are on to something Yamato hugger. One thing to note is that most of the commanders I have changed by a large margin have been sub commanders. I think it is interesting that most of the leaders available to choose are commanding subs. I will be off doing more testing to see what I find.

[rtrapasso]

OK - i can get the results you speak of, but i am wondering if when the turn is run that the cloned leaders are cleaned up...

Anyway, i have forwarded it to the programmers.