+* faire un autre thread pour les envois d'infos
+ -> c'est stupide d'attendre la fin des calcules s'il y en a beaucoup
+ surtout en bookkeeping...
-* fix deadlock bug with ./loba cluster1000.xml -N26 -i2
+* review receive with timeout.
-* implement loba_* algorithms (start with some trivial one)
+* verify bookkeeping version.
+
+* add several metrics
+ - message exchanges : number/volume of sent/received data/ctrl messages
-* fix process::run (see inline comments)
+* add options -j/-J : minimum number of iterations?
-* find a better
+* add a variant to (not) change neighbor load information at send?
-* add some statistics about load (im)balance at the end of the simulation
+* implement loba_* algorithms (start with some trivial one)
* for automatic process topology,
-> implement some random initial distribution of load
* add synchronized mode
-* translate README file ?
+* translate README file?
