Coordinates of old examples stay the same when loading new examples by the APT editor

[mgieseking]

When I for example copy this file in the APT Editor

// Adapted tafel 2 example: the enviroment first decides on their own
// Cannot be won by the system anymore, since the system cannot prevent the env
// to enter the right arm anymore and so if the env take t4, in this case t5 could 
// not be used and thus the system blocks if it don't uses t3. 
// Deleting t4 or the env2 in the preset of t5 would fix the problem.
.name "exampleTafel"
.type LPN
.description "Example tafel 28.09.2014"
.options
condition="SAFETY"

.places
Env[env="true"]
Env1[env="tr"]
Env2[env="tr"]
Env3[env="tr"]
Env4[env="tr"]
Env5[env="tr"]
s1
s2
s3
s4
s5
s6
s7[bad="asdf"]
s8

.transitions
t1
t2
t2n
t3
t4
t5
t6
t7
t8

.flows
// left arm
t1: {Env} -> {Env1}
t2n: {Env} -> {Env5}
t2: {Env5,s1} -> {Env2,s4}
t3: {s3} -> {s7}
t4: {Env2} -> {Env3}
t5: {Env2,s5} -> {Env4,s8}
t6: {s2,s4} -> {s5,s6}
t7: {s8} -> {s8}
t8: {Env1} -> {Env1}

.initial_marking {Env,s1,s2,s3}

and freeze the nodes (or do anything which saves the coordinates). Then going back to the APT editor and loading the example without the coordinates again, then it already has the same coordinates as the example before.

This is also happens for a subset of the nodes when I for example add some. Or have another example, but the names of the nodes are equal (maybe the name of the net has to be the same too, I don't know).

[mgieseking]

Another, I guess related, misbehavior is when

1. loading the page
2. deleting place p2
3. creating a new place

The place gets also the id p2 and is already initially marked.

[annyanich]

Another, I guess related, misbehavior is when

1. loading the page

2. deleting place p2

3. creating a new place

The place gets also the id p2 and is already initially marked.

Has this misbehavior been fixed by mistake sometime in the last couple of weeks? I recall having reproduced it when you reported it, but now, when I try it, the new place does not end up with an initial marking on it.

[mgieseking]

Hm maybe it seems like that. At least I'm currently also not able to reproduce the error. (The other bug is still around, I'm currently exploiting it for not having to model a net each time from the beginning for some scaling examples ;))