Problem in editor (bug ?)

[welk]

Using upgrades increments in editor, I experimented a crash with one upgrade : It's the N° 6 in localization file
(#RESEARCH_ID_6)
I renamed it Guard and I affected it to commanders as "manuel upgrade" (not automatic upgrade).

When I "upgrade" a commander that is on map with this upgrade, no problem (1 to 5)

But when I try to low it (example : from 5 to 4, etc), editor crashes and display a windows "Segmentation violation" (see attached image)

What may be the origine of this ? Bug ?

The crash occurs only for this upgrade (actualy, no any crash problem with another upgrade : only this one)
And crash does occur only if I low the parameter (never when I add values)
Exemple : from 1 to 5, no any problem ; but when I want to low from 5 to 4, or 1 to 0, etc, there is the crash (not each time, but approx 80% of time)

Curiously, the crash does occur only with this specific and particular upgrade, and only when I try to low it, and only with the "red side army" (no any pb of crash with commanders of blue side army, when I use exactly the same upgrade)

By security, I verified in editor that blue side and red side armies have exactly the same parameters : seems there is no any apparent reason for this crash


Here is the text of the "exception trace" file in editor :


editor: PANIC: Unexpected harmful signal (Segmentation violation) ...

editor: system execution failed.
Following is the set of recorded exceptions.
NB: The raised panic may have induced completely inconsistent information:

-------------------------------------------------------------------------------
Class / Object Routine Nature of exception Effect
-------------------------------------------------------------------------------
APPLICATION root's creation Segmentation violation:
<000000000B04804C> Operating system signal. Exit
-------------------------------------------------------------------------------
APPLICATION root's creation
<000000000B04804C> Routine failure. Exit
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ entering level 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-------------------------------------------------------------------------------
APPLICATION root's creation Unexpected harmful signal (Segmentation violation):
<000000000B04804C> Eiffel run-time panic. Bye
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ back to level 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-------------------------------------------------------------------------------

editor: dumping core to generate debugging information...

editor: PANIC CASCADE: Unexpected harmful signal (Segmentation violation) -- Giving up...

editor: PANIC: Unexpected harmful signal (Segmentation violation) ...

editor: system execution failed.
Following is the set of recorded exceptions.
NB: The raised panic may have induced completely inconsistent information:

-------------------------------------------------------------------------------
Class / Object Routine Nature of exception Effect
-------------------------------------------------------------------------------
APPLICATION root's creation Segmentation violation:
<000000000B01E04C> Operating system signal. Exit
-------------------------------------------------------------------------------
APPLICATION root's creation
<000000000B01E04C> Routine failure. Exit
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ entering level 1 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-------------------------------------------------------------------------------
APPLICATION root's creation Unexpected harmful signal (Segmentation violation):
<000000000B01E04C> Eiffel run-time panic. Bye
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ back to level 0 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-------------------------------------------------------------------------------

editor: dumping core to generate debugging information...

editor: PANIC CASCADE: Unexpected harmful signal (Segmentation violation) -- Giving up...

[Hubert Cater]

Could you package up the campaign you are experimenting with and send it to me at support@furysoftware.com

Please use a file transfer site like WeTransfer if it is too big for email, or Dropbox or similar.

I would need the campaign file itself, and the subfolder for the campaign as well.

Thanks,
Hubert

[welk]

Hi Hubert, and Merry Christmas

I sent to you a mail with all needed links

The link to download the files is here :
Sent by mail

I added the txt exception file, that reports crash in details (crash did occur only in editor, never in game)


Work is not finished, lot of parameters have nor to be calibrated correctly, thery are not at this time

Files of campaign and files of the mod are in separated directories
The mod will work like that :
1- Separate/independant install of the game
2- Campaign files in the "user" directory"

I had to do that because if not, campaign files have a too high weight. And in addition, "mod" directoy in "user" does work perfect in general, but I experimented some particular sounds problem with the mod (the most part of units had a correct sound, but some units did keep obstinatly the vanilla sounds, even I copy-past the correct sounds in "mods" user directory)

I though because the mod does change heavely the game, it was better to have a separated install to use it


CONCERNING THE CRASH PROBLEM : Curiously, I was not able these days to reproduce the crash.
I am not informaticien, but I suspect now the crash would may have something to do :

1- With my red side army graphs (the crash did occur only with red side army, never with blue army) : I think that because to change the upgrade level of a commander unit in the editor is at the same time to change of graphs column in unit sprites files.

2- Perhaps also with the memory of the used PC : because after restarting my PC 2 or 3 times (2 or 3 days of using) after experimenting the crash, it did no more occured.


I think you should really consider to use your SC engine to produce a Strategic command battles series, because AI makes a very good oponent in small/medium maps (50X0 to 75x75, and perhaps 125x125 to be tested), and battles are very fun to play with adapted sounds and graphs
You would just have to add "formations" and "flank/rear attacks", and to adapt some little things to the "battle scale", and this new scale games battles series would be absolutly perfect. Due to the infinite number of battles in history (from Ancient times to end of 19th century), you would have enough work for 1 or 2 century
In particular, SC engine would be very good for Napoleonic scale (I will also do a Napoleonic version of my mod later)

The most ideal configuration seem be that for excellent and quick AI :
1/Maps : 50x50 to 75x75
2/ 100/150 unit for each side
3/ To avoid to give more than 5 action points to units (same thing for range action)

In addition, you could use the aerial units slots (that are not used in battles) to convert them in new kind of ground units : I succeed to give to the mod 28 kinds of units, but more would be also interessant)

[welk]

In addition : here is a old campaign I had registered : the crash occured with
Link sent by mail

It occured always with red side commanders that are in :
40,16
40,15
45,13
45,14
42,28
42,29
47,28
47,29

When the crash occured, I had set all properties of these commander to 5 (not only defensive guard property), and the crash occured with them only when I tried to low this "defensive guard" property from 5 to 4 or from 1 to 0

But at this time, when I try to reproduce it, doing exactly the same manipulation, with this same campaign and with these specific commanders, the crash does no more occur

The only thing that changed between date of this old campaign and now, is that I restarted my PC for several days