PlanetsCentral

Greetings!

A user on my BBS who on my recommendation switched to PCC2 from VGAP DOS Client is having an issue recently.

He switched to PCC2 about a 2 months ago, and said it's been working OK for him up until recently (he wasn't able to say exactly when it started happening). Whenever he views a particular ship (either scrolling through the ship listing or using pgup/pgdown to change to it) PCC2 crashes with the following error:

=== Loading... ===
No host configuration file found -- using defaults
Loading Lizard RST file...
Compiling starchart...
=== Loading complete ===
Playing player 2: The Lizard Alliance
game/racename.cc:95: INTERNAL ERROR: player >= 0 && player <= 12

## The program has just crashed. Above are its last words.
## If you send a bug report, please include this information.
Signal number was -1

He said switching back to the DOS client works fine, and he's been using it again for now. He obviously likes PCC2 a LOT more than the DOS client, so would really like to get to the bottom of this issue if possible.

If it matters, the game is running on TimHost Version 3.14b (by request) currently on turn 46.

If it helps to send a copy of an RST file let me know, thanks!!

-=Xhum=-

Hi,

to check whether the problem is in the current data or in the history data, try removing the chartX.cc file (copy game directory, delete that file, check whether issue is still there). If the issue goes away, the problem was in the history data; if you can live with losing history, you have now fixed it...

If the issue does not go away, the problem is still in the current data. One possibility would be that a ship has a strange Primary Enemy. I don't know how this can happen, but sometimes HOST does strange things. The easiest way to fix a bad Primary Enemy is to use Global Actions to set all ships' PE to "none"; there are other possibilities using the scripting language.

You can send me the file(s) so I can help you sanitizing it.

I actually find it interesting that Host 3.14 is still in use. I have a bunch of test cases with it, but I haven't used in a game for a LONG while. It is lacking some features that Host 3.2 has, and neither of the PCCs is guaranteed aware of that.

--Stefan

Thanks for the quick response! I'll try those things and will report back.

The RST was moved into a folder all by itself, and PCC2 still crashed - but doing as you suggested to reset the PE for all ships DID fix the issue! He's thrilled

Thanks again!

The RST was moved into a folder all by itself, and PCC2 still crashed - but doing as you suggested to reset the PE for all ships DID fix the issue! He's thrilled

ok, thanks for the test. I added a filter to the next version of PCC2 that it should no longer crash when this happens.

--Stefan

ok, thanks for the test. I added a filter to the next version of PCC2 that it should no longer crash when this happens.

Thanks!

A different user is having new issue (same game/version) - at a particular starbase whenever they press A to "build torpedos" or when clicking the "beams" tab when building a new ship, the program crashes with this error:

=== Loading... ===
No host configuration file found -- using defaults
Loading Fed data...
Compiling starchart...
=== Loading complete ===
Playing player 1: The Solar Federation
cpluslib/cpluslib/maybe.h:62: INTERNAL ERROR: isKnown()

## The program has just crashed. Above are its last words. ## If you send a bug report, please include this information. Signal number was -1

This error occurs even if placing the RST in an empty folder.

Any help greatly appreciated! Let me know if you would like a copy of the RST file for testing.

I have done a bit more troubleshooting and this is what I've found by looking through nightly backups:

Turn 54 - everything seems fine. The problematic SB (at Planet #232: Lynch) seems fine for this RST file, I cannot create a crashing situation.

Turn 55 - the bug starts to appear. I can click A to access the "build torpedo" screen, and I can click every tab EXCEPT the "Beams" tab in the build ship screen. When clicking the Beams tab, PCC2 crashes with the error described in the above post.

Turn 56 - the bug is now worse - cannot click A to build torpedo, and when trying to build a ship PCC2 crashes instantly (same error as above).

I have uploaded the Player1 RST files for the above 3 turns, as well as the submitted .TRN files (for turns 54 and 55) to this folder:

https://planetscentral.com/file.cgi/xhumeka/bug20220303/

The only other thing I'll mention is that there was a weird error thrown in the ERROROLD.LOG file, but it was only thrown for turn 56:

**********************
03-01-2022 04:01:18
Turn Number: 54

**********************
03-02-2022 04:01:20
Turn Number: 55

**********************
03-03-2022 04:01:20
Turn Number: 56

DATA UNDERFLOW ERROR! in player 1 .TRN file

I already told those playing this game that I won't be running another game on such an old Host since these weird problems keep popping up, but any help you can provide to keep this game going until completion would be grealy appreciated!! Thanks Stefan!

Turn 56 - the bug is now worse - cannot click A to build torpedo, and when trying to build a ship PCC2 crashes instantly (same error as above).

Thanks for the files. First analysis so far: you have -1 Heavy Blasters on that base. PCC2 gets confused because -1 means "unknown", but on the other hand, if you have a base, there shouldn't be any unknown values.

This message from host

**********************
03-03-2022 04:01:20
Turn Number: 56

DATA UNDERFLOW ERROR! in player 1 .TRN file

essentially says the same thing, and there actually is a command with negative beams (Blasters, this time) in the turn 55 turn file.

I vaguely recall that some host versions have trouble when you combine Super Refit and ship building: say, you have 6 beams on the base, and want to build a ship with 6 beams, so the build order is valid. Now, Super Refit happens before ship building and takes away one beam, leaving the base with one too few, which causes this very problem.

The best solution would be to edit-out this problem using Plangineer (and not touch this base otherwise this turn to not cause data errors). I don't have a good client-side solution of the form "here's your fixed RST", because this error needs to be fixed host-side, and unlike a bad Primary Enemy we cannost just edit it out. If we do, host will see that the number of beams increased without you paying for it.

--Stefan

The best solution would be to edit-out this problem using Plangineer (and not touch this base otherwise this turn to not cause data errors). I don't have a good client-side solution of the form "here's your fixed RST", because this error needs to be fixed host-side, and unlike a bad Primary Enemy we cannost just edit it out. If we do, host will see that the number of beams increased without you paying for it.

--Stefan

Great, thanks so much! I'll try this out for tonight's turn and will report back the results, thanks again!

-=Xhum=-

Quick update - I tested the fix on a copy of the game folder (just ran host with no .trn files after running plangnr and changing the -1 to a 0) and the new .RST file seems to work properly with no crashing in PCC2!!

Thanks again so much for the speedy fix - now I just have to make sure the player does not do anything with the SB as you suggest to ensure they aren't hit by the Tim Continuum!

Quick update - I tested the fix on a copy of the game folder (just ran host with no .trn files after running plangnr and changing the -1 to a 0) and the new .RST file seems to work properly with no crashing in PCC2!!

I also tested a little: given a base building a Nocturne (4 beams), and an Outrider trying to refit (1 beam), I get pretty weird and different errors with HOST 3.0f up to 3.15b....

    ERROR! Impossible Hull specs! Player #  0
    ERROR! Impossible Hull specs! Player #  12
    ERROR! Impossible Hull specs! Player #  12
    ERROR: Player 1 Tryed to build an ship without an engine

(what? there is no player #0 or player #12 in my universe)

...but no problems with HOST 3.21 and later, or PHost 3.2 and later.

I haven't yet hit "your" exact problem yet, but I would conclude that combining ship building and Super Refit on Host 3.15 is not a good idea.

--Stefan

I haven't yet hit "your" exact problem yet, but I would conclude that combining ship building and Super Refit on Host 3.15 is not a good idea.

That helps a lot, I'll pass on the info - it would be nice to avoid this problem going forward so knowing what's (perhaps) causing it is a huge help!

Success! The user submitted her turn file last night (avoiding making any changes on the problematic starbase) and everything seems fine today - no Tim Continuum attack and the crashing problem has been fixed.

Thanks so much Stefan!

-=Xhum=-