Possible CRITICAL bug (lost garrison)

[gwheelock]

I had a case where my Wurzberg garrison vaporized after loading the GB turn file:

This may be related to a strange end case that occured IN MECKLENBERG:

History :

1) Prussia DOWs Mecklenberg, Britain gets control
2) Prussia moves in kills >1< of the 2 garrison factors & take the other one POW.
3) I (France) am at war with both Prussia & Britain; move in & knock the
Prussian corp out.
4) Mecklenberg has a lapse of war with Prussia; POW is returned TO BRITAIN
5) Britain runs reinforcement; cannot place Mecklenberg POW because of
my occupation.
6) I load Britain turn; my garrison IN WURZBERG disappears.

Could you have either a bad pointer variable or a database record reference
that isn't set correctly & is confusing the WURZBERG garrison with the
Mecklenberg POW? If this is so; I think this may be a CRITICAL bug because
it could potentially vaporize any force on the board (such as Nappy & the full
#1 French corp)


Here is the screenshot of the setup before applying the British turn :

[gwheelock]

Here is screenshot AFTER the British

[gwheelock]

Here is my turn files from BEFORE the British

[gwheelock]

And here is the GB turn file that causes the problem

[Jimmer]

Sometimes I cause the STRANGEST end-cases.

[zaquex]

ORIGINAL: gwheelock

...Could you have either a bad pointer variable or a database record reference
that isn't set correctly & is confusing the WURZBERG garrison with the
Mecklenberg POW? If this is so; I think this may be a CRITICAL bug because
it could potentially vaporize any force on the board (such as Nappy & the full
#1 French corp)

I'm not claiming to be an expert and I have no knowledge about the EiANW db, still this is not the first bug or thread that seems to indicate that there is a db problem. Either as you say that its a pointer/reference or that one or more tables has an unexpected number of rows. Bugs I'm thinking about is where corps and garrisons seem to change nationality or a player suddenly find himself playing a minor!?!? And all these problems seem to have appeared in 1.01 beta

[gwheelock]

I am an expert (25 years professional programming experience) & while I'm not familer with the
internals of EiANW; all of these sound like the type of thing that you could get if you do updates
to data records with your record pointer pointing to the wrong place.
&nbsp;
That is why I listed it as a "critical" level bug - the severity level that I use for program bugs are :
&nbsp;
1) Critical - This is a bug which not only keeps a major component from working correctly; but damages
the underlying data set.&nbsp; This means that just fixing the PROGRAM will NOT correct all of the problems
produced by this bug because the data has been corrupted. (ie. in this case the Wurzburg garrison
will not reappear)
&nbsp;
2) Serious - This&nbsp;is a major componet of the app not working correctly or at all.&nbsp; Significant imparement
to the usability of the program.
&nbsp;
3) Problem - Some component not working correctly but either a workaround exists or the item is not
critical to the app's function.
&nbsp;
4) Minor - These are cosmetic bugs (spelling; etc)
&nbsp;
Guy

[fvianello]

We had a similar problem in our PBEM game, and it was caused by one of the players using version 1.0.0. Moreover, for some reason MEcklenburg is a very sensitive region (probably a data problem somewhere)