URL:
<http://gna.org/bugs/?24435>
Summary: Savecompat conversions assume consecutive player
numbers
Project: Freeciv
Submitted by: cazfi
Submitted on: Fri 19 Feb 2016 11:55:05 PM EET
Category: general
Severity: 3 - Normal
Priority: 5 - Normal
Status: None
Assigned to: None
Originator Email:
Open/Closed: Open
Release:
Discussion Lock: Any
Operating System: None
Planned Release: 2.5.4, 2.6.0, 3.0.0
_______________________________________________________
Details:
I were checking why a 2.5 scenario from
http://forum.freeciv.org/f/viewtopic.php?f=24&t=762#p4538
does not load to newer branches.
The error messages loading gives are about the fact that there's no player10
(there's other holes in the array later on, but that's the first one and where
things break), player numbers are not consecutive.
The main savegame?.c modules are prepared to that since they do
player_slots_iterate(pslot) {
if (NULL == secfile_section_lookup(loading->file, "player%d",
pslot_id)) {
continue;
}
However, in the savecompat conversion functions the usual way to update
players is
for (plrno = 0; plrno < nplayers; plrno++) {
There are some cases where it *does* check that section "player%d" exist, but
I suspect those too go wrong in that the iteration anyway ends at plrno ==
nplayers despite some slots up to plrno have not been used -> do not count
toward nplayers.
With that scenario file it's shown that S2_6 and TRUNK are affceted. I suspect
that S2_5 is too, if loading savegame from older versions.
_______________________________________________________
Reply to this item at:
<http://gna.org/bugs/?24435>
_______________________________________________
Message sent via/by Gna!
http://gna.org/
_______________________________________________
Freeciv-dev mailing list
[email protected]
https://mail.gna.org/listinfo/freeciv-dev
