xconq net game: loss of sync

xconq net game: loss of sync

[Jim Kingdon]

Tom and I tried to play xconq tonight.

We got to turn 10 before we got a loss of sync.  I had him save the
game.  It saved on both sides.  I had him email me the one from his
side.

The diff was about 90Kbytes, which is much larger than I would have
thought.  The spurious diffs seemed to be places like
swedish-place-names.  In mine it starts:

(grammar root 100 (root (reject name Aa

(where "Aa" are really a's with circles on top)

and in Tom's it has

(grammar root 100 (root (reject name |Aa|

(note the vertical bars; the Aa seem to be the same in both cases).
Also his lines are broken at about 990 characters.  Windows vs Linux?
Or maybe something about the process of mailing it?  The lines are
broken with "!" at the end of the continued line, which I don't think
is an xconq convention.  The line-break thing does seem to have a lot
to do with the sheer size of the diffs.

Anyway, moving on to what appear to be the real differences.  I've
enclosed the part of the diff which shows some signs of being real.
The differing material levels for Ojo Caliente looks
disturbing to me.  Likewise for the fighter.  Bugs in the supply code?

@@ -1148,7 +1172,7 @@
   (plural-noun "Vietnamese")
   (emblem-name "flag-vietnam")
   (unit-namers "" "" "" "" "" "" "" "" "" "" ""
   "vietnamese-place-names" "vietnamese-place-names")
-  (total-time-used 119)
+  (total-time-used 123)
   (finished-turn 1)
   (advantage-max 15)
   (player 1)
@@ -1218,8 +1242,7 @@
   (plural-noun "Mexicans")
   (emblem-name "flag-mexico")
   (unit-namers "" "" "" "" "" "" "" "" "" "" "" "mexican-place-names"
   "mexican-place-names")
-  (total-time-used 144)
-  (finished-turn 1)
+  (total-time-used 119)
   (advantage-max 15)
   (player 2)
   (default-doctrine 1)
@@ -1244,9 +1267,9 @@
   "40abg4h2f12a"
   "40a2bf3hfb12a"
   "40a2b2f2hgb12a"
-  "38a5b2fhfc12a"
-  "38a6b4fb11a"
-  "39a6bd2fb11a"
+  "40a3b2fhfc12a"
+  "41a3b4fb11a"
+  "42a3bd2fb11a"
   "43a6b11a"
   "43a6b11a"
   "43a6b11a"
@@ -1268,17 +1291,17 @@
   ))
 
 (side 2 (unit-views
-  (fighter 2 (null) 1 39 19 0 1 10 23 fighter)
   (infantry 2 (null) 1 42 25 0 1 10 21 soldiers)
   (city 2 Chihuahua 1 43 20 0 1 10 12 city20)
-  (infantry 2 (null) 1 45 18 0 0 10 38 soldiers)
+  (infantry 2 (null) 1 24 16 0 0 10 38 soldiers)
   (infantry 2 (null) 1 43 22 0 1 10 34 soldiers)
   (town 2 Durango 1 43 23 0 1 10 10 town20)
   (bomber 2 (null) 1 24 16 0 0 10 28 4e)
   (town 2 El Sueco 1 43 24 0 1 10 7 town20)
-  (bomber 2 (null) 1 44 19 0 0 10 26 4e)
+  (bomber 2 (null) 1 45 19 0 0 10 26 4e)
   (town 2 Ojo Caliente 1 44 19 0 1 10 17 town20)
-  (infantry 2 (null) 1 25 13 0 0 10 40 soldiers)
+  (fighter 2 (null) 1 44 19 0 1 10 23 fighter)
+  (infantry 2 (null) 1 25 15 0 0 10 40 soldiers)
   (town 2 Veracruz 1 45 19 0 1 10 8 town20)
   (infantry 2 (null) 1 44 20 0 1 10 33 soldiers)
   (bomber 2 (null) 1 25 12 0 0 10 35 4e)
@@ -1295,7 +1318,7 @@
   (adjective "independent")
   (color "black,black,white")
   (emblem-name "none")
-  (total-time-used 1)
+  (total-time-used 6)
   (finished-turn 1)
   (advantage-max 15)
   (controlled-by 0)
@@ -1365,7 +1388,7 @@
   (plan Passive 2))
 (fighter 25 12 1 (# 30) (image-name "fighter") (nb 1) (m 15 3)
   (acp 0) (am 3)
-  (plan Passive 8 (tasks (move-to 4 0 25 11 0 0 0))))
+  (plan Passive 8 (supply-alarm false) (tasks (move-to 4 0 25 11 0 0
0))))
 (bomber 25 13 1 (# 31) (in 2) (image-name "4e") (nb 1) (hp 1) (cp 10)
(m 0 0)
   (acp 0) (acp0 0)
   (plan Passive 2))
@@ -1399,7 +1422,7 @@
 (city 43 20 2 (n "Chihuahua") (# 12) (image-name "city20") (nb 1) (m
900 400)
   (acp 0)
   (plan Passive 0 (tasks (build 6 0 0 38 1 2 43 20))))
-(town 44 19 2 (n "Ojo Caliente") (os 0) (# 17) (image-name "town20")
(nb 6) (m 485 200)
+(town 44 19 2 (n "Ojo Caliente") (os 0) (# 17) (image-name "town20")
(nb 6) (m 500 200)
   (acp 0)
   (plan Passive 4 (tasks (build 6 0 0 40 1 2 44 19))))
 (infantry 42 25 2 (# 21) (image-name "soldiers") (nb 1) (m 5 6)
@@ -1408,9 +1431,9 @@
 (infantry 46 23 2 (# 22) (image-name "soldiers") (nb 2) (m 5 6)
   (acp 0) (am 1)
   (plan Passive 4 (wait true)))
-(fighter 39 19 2 (# 23) (image-name "fighter") (nb 1) (m 13 3)
-  (acp 0) (am 6)
-  (plan Passive 8 (tasks (move-dir 5 0 4 0))))
+(fighter 44 19 2 (# 23) (in 17) (image-name "fighter") (nb 1) (m 2 3)
+  (acp 5) (am 1)
+  (plan Passive 8 (wait true) (supply-alarm false) (supply-is-low
true)))
 (bomber 46 21 2 (# 25) (in 11) (image-name "4e") (nb 1) (hp 1) (cp
10) (m 0 0)
   (acp 0) (acp0 0)
   (plan Passive 2))

Re: xconq net game: loss of sync

[Eric McDonald]

Jim Kingdon wrote:

> Oh, a bug in the supply code is sort of what I was insinuating before,

It could also be due to the fact you guys may not have saved at exactly 
the same point, and some automatic things (plan executions) continued 
even after you got out sync, and thus some parts of the game diffs are 
present for that reason.

> but there is at least one other possibility: whether I really ended up
> with the same version of xconq as Tom.  I did a CVS update and applied
> Eric's patch ( http://sources.redhat.com/ml/xconq7/2004/msg01014.html
> ).  So assuming that is what Eric did, this probably isn't it. 

This is exactly what I did before building the Windows executables.

>Insert groveling about how I shouldn't really gumble
> so much unless I'm volunteering to fix it).

;-)

Eric

Re: xconq net game: loss of sync

[Jim Kingdon]

Oh, a bug in the supply code is sort of what I was insinuating before,
but there is at least one other possibility: whether I really ended up
with the same version of xconq as Tom.  I did a CVS update and applied
Eric's patch ( http://sources.redhat.com/ml/xconq7/2004/msg01014.html
).  So assuming that is what Eric did, this probably isn't it.  But I
thought I'd mention it just in case.

(Insert random boilerplate grumbling about how fragile xconq's
protocol is, compared with a freeciv-style one (as I understand the
latter, anyway).  Insert groveling about how I shouldn't really gumble
so much unless I'm volunteering to fix it).