Unable to terraform

[HarryMudd]

Morning!
I hope this is the correct place for a bug report? I tried looking through the forum to see if this has already been reported but I was getting too many results from my searches.
I'm playing an ECM 5 session. I've attached a saved game and in it, I'm trying to terraform Desert Planet Deneb 3 in sector C1.
However, I just noticed that nothing is happening. The planet symbol shows as 'terraforming in progress' but the planet type letter remains red (usually the next turn it shows yellow 'in progress').
Is this a known issue?
Brian

[thunderchero]

Morning!
I hope this is the correct place for a bug report? I tried looking through the forum to see if this has already been reported but I was getting too many results from my searches.
I'm playing an ECM 5 session. I've attached a saved game and in it, I'm trying to terraform Desert Planet Deneb 3 in sector C1.
However, I just noticed that nothing is happening. The planet symbol shows as 'terraforming in progress' but the planet type letter remains red (usually the next turn it shows yellow 'in progress').
Is this a known issue?
Brian

not really this section is for installation reports/issues
mod issues should go in mod topic, in this case ECM topic
viewforum.php?f=30

as for your issue, system with issue Deneb? if so that fleet has some stuck order just redeploy fleet and give terraform order again.

[HarryMudd]

OK, sorry for the wrong section.
I tried redeployment. Originally I had 1 destroyer & 3 colony ships grouped together when I issued the order.
When I realised there was problem I set an 'Avoid' order, then redeployed one of the colony ships out of the group and tried terraform again. No joy.
Any thoughts?
Brian

[thunderchero]

redeploy all ships, until all ships say "avoid" then close deploy screen

worked for me

[HarryMudd]

redeploy all ships, until all ships say "avoid" then close deploy screen

worked for me

I did that before I posted, no joy. However, I since moved everything to an adjacent sector, then moved back 1 colony ship and it worked!
Odd, but hey!
I must say, the ECM is a really good update, once I got used to the changes from the vanilla version I was used to. I keep getting distracted from what I should be doing for too long!
Brian

[Spocks-cuddly-tribble]

I did that before I posted, no joy. However, I since moved everything to an adjacent sector, then moved back 1 colony ship and it worked!

It works in the same turn but is a bit messy (e.g. one CS to Packled, Packled to CS fleet, all CSs to ex-Packled fleet).

Vanilla glitch with temp fleet IDs after re-deploying, combined with different/conflicting orders and ship types, can produce some really weird bugs.

Removing a colony ship from the fleet the display gliches out with empty frame and changes the Pakled ship into a colony ship.
Disolving the colony ship fleet also shows four colony ships and even removes one colony ship next turn:

redeploy_glicth.png (264.84 KiB) Viewed 1157 times

Back in the day, playing new-ish vanilla BotF, I corrupted some nice vanilla saved games beyond repair, by re-deploying odd gost fleets with hidden/bizarre cross order data. I think Flocke provides a sav file repair function in UE to auto-fix most of the issues. But, no offense, this is a strange cheat game, so not worth Flocke's time to investigate the glitch further.

@ thunderchero: The fleets where re-deployed after loading a save before ending turn, and then saved again (aka save scumming).
Same glitch issue can happen to the build queue (i.e. build/buy order not executed, like the terraforming here). Or worse aka crash.

Odd, but hey! I must say, the ECM is a really good update, once I got used to the changes from the vanilla version I was used to.

Glad you like it, I hope to send thunderchero the files for the pending update this evening, as I had some nice test games:

ECM_234.png (498.57 KiB) Viewed 1157 times

[thunderchero]

I think Flocke provides a sav file repair function in UE the auto-fix most of the issues.

due to Asteroid and Asteroid Belt Dilithium patch UE will corrupt (remove resources) in saved games with this patch.
viewtopic.php?p=57639#p57639

[Spocks-cuddly-tribble]

due to Asteroid and Asteroid Belt Dilithium patch UE will corrupt (remove resources) in saved games with this patch.
viewtopic.php?p=57639#p57639

Since asteroids are just unused cosmetics in ECM this is no big issue. Otherwise we just have to avoid editing ECM saves with UE for the time being.
Unless Flocke adds an option to ignore unknown feature bits in [systInfo+70] instead of deleting them.

[justanick]

Otherwise we just have to avoid editing ECM saves with UE for the time being.

It is also possible that the problem is related to auto-fix only, so apart from that editing savegames from ECM could be reasonably safe.

[Flocke]

I think Flocke provides a sav file repair function in UE the auto-fix most of the issues.

due to Asteroid and Asteroid Belt Dilithium patch UE will corrupt (remove resources) in saved games with this patch.
viewtopic.php?p=57639#p57639

Should be fixed, please compile latest code and confirm.

edit: I tested above save. There indeed seems to be a weird issue with the terraform order, which is not flagged by the integrity checks yet. I'll have a look. Good thing is that UE didn't mess anything when saved, but the game can still be loaded and continued.

[Flocke]

I did that before I posted, no joy. However, I since moved everything to an adjacent sector, then moved back 1 colony ship and it worked!
Odd, but hey!

A very odd bug indeed!

GWTForce & GTForceList at first sight looked fine and deleting all the orders and turn results as well as deleting all the AI tasks didn't help. What helped is to reset current task force mission in UE and re-assign it ingame. What also helped is to redeploy ships into the Pakled transport fleet like SCT described. This way I could also reproduce the display bug where the Pakled ship turns into a colony ship.

Redeploying ships into a new fleet however makes things worse. No matter whether 2 or all 3, when I give terraforming order agein, instead of terraforming they move down to tantalus. They even move 2 sectors each turn while they only have a speed of one sector. They must be in real hurry to get away from the Bandi.

messed up movement.jpg (822.61 KiB) Viewed 275 times

Checking GWTForce on this, this messed up movement indeed is caused by overlapping temporary task force ids. This case the three colony ships have same temporary task force id like the scout ship. I dunno whether it is stored somewhere. From this bug I fear not. Possibly one should not redeploy ships when there already are redeployed task forces stored to the loaded savegame.

Like documented in viewtopic.php?p=55981#p55981 the romulan temporary task force ids start with B1 E0 (-20000) and are counted down by -1 each new task force. With the redeployment above it is however ignored that there already exist two of them, but the counter starts over with B1 E0.

Nonetheless, this temporary id mess IS NOT responsible to the reported bug. The actual cause I identified to be the given task in GTForceList instead.
By GTForceList the three colony ships in deneb were already ordered to terraform 3rd planet on turn start, so the order must have been assigned the turn before. Then however there should have been terraform progress! Given there has been no progress before, the responsible trek.exe routine seems to skip any progress calculation, which seems somewhat reasonable.

deneb.jpg (34.54 KiB) Viewed 273 times

When I reset previous turn order to avoid, the terraforming works perfectly. I additionally verified this by doing the reverse and applying the former turn terraforming to another colony ship, that worked beforehand. It caused same problem. Terraforming of previous turn must be unset and simply can either be fixed by resetting the task force orders with UE, by carefully redeploying ships without introducing new temporary task forces, or by giving another order and waiting a turn:

fix-broken-terraform.jpg (109.4 KiB) Viewed 275 times

What I failed at is to reproducce this bug ingame without hex editing.
The order usually is immediately unset when terraforming completes.

One new thing I confirmed is that for terraforming, the mission data of GTForceList / GWTForce at 0x30 - 0x33 lists the terraformed planet number instead of the target sector.

[Spocks-cuddly-tribble]

Possibly one should not redeploy ships when there already are redeployed task forces stored to the loaded savegame.

Yes, I propagate this for quite some time now. Worst case is that changed or phantom ships can crash the game in later turns.

But even without redeploying there is many issues with changing things in loaded games before ending turn. And the ordInfo might affect this as well?
E.g. just changing stuff like research/intel settings or system build orders sometimes causes some ships to stop unrelated tasks like just moving. Sometimes they even reverse course.

given task in GTForceList (...) previous turn order (...) must be unset

Similar issue can happen with system build queues. No more progress until you remove all items and sacrifice the invested industry in this case.

The order usually is immediately unset when terraforming completes.

Some glitches e.g. finishing simultaneous terraforming the same planet with multiple fleets just cancels the first fleet order. All other fleets maintain the pointless task until next turn.

for terraforming, the mission data of GTForceList / GWTForce at 0x30 - 0x33 lists the terraformed planet number instead of the target sector.

This might explain why just clicking a taskforce can stall terraforming (albeit the order terraform is still displayed),
cf. my false libertions fix due to TT mission data corruption: viewtopic.php?p=59286#p59286

[Flocke]

Some glitches e.g. finishing simultaneous terraforming the same planet with multiple fleets just cancels the first fleet order. All other fleets maintain the pointless task until next turn.

Ah well, that would be a plausible cause. Anyhow, I added another integrity check to fix the stuck terraforming issue by resetting the former turn mission. Works like a charm with the provided game6.sav

[Spocks-cuddly-tribble]

Ah well, that would be a plausible cause.

No, that's just a confusing UI problem and does NOT cause this topic issue!
I still think this was initially caused by invalid temp IDs somehow pointing to wrong GTForceList previous turn order and/or OrdInfo.

overlapping temporary task force ids... I dunno whether it is stored somewhere. From this bug I fear not.

Word [58FD24 + player empire ID *6] needs to be stored when saving games (e.g. unused alienInfo area) and restored from there when loading games
This means two small extra subs for save/load game with skip MP games check to avoid sync issues.
sub_4495E0 temporary_task_force_IDs // sub_4495C0 reset_temp_TF_IDs is called during each TURN_PROCESS.

[Flocke]

I still think this was initially caused by invalid temp IDs somehow pointing to wrong GTForceList previous turn order and/or OrdInfo.

Very unlikely. The affected task force doesn't have a temporary ID, plus the game is at turn 18 I think and there are only low task force ids in use.
The only way I could imagine, is that the fleet already moved with terraforming order into the sector. But then even the planet index would have to match, and I guess it was manually selected, else it would have been noticed and reported.

But whatever has been causing it, the issue can be easily fixed even without patching the savegame. No data mess.

Word [58FD24 + player empire ID *6] needs to be stored when saving games (e.g. unused alienInfo area) and restored from there when loading games

Nice! Where's the patch?