Test with even larger values, move code to right place

[spkrka]

No description provided.

[coveralls]

Coverage increased (+0.5%) to 83.53% when pulling 455b612 on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+0.62%) to 83.66% when pulling 455b612 on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+6.44%) to 89.47% when pulling 03ab85191c4352277b662d21e3d6166ed4801df5 on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+6.44%) to 89.47% when pulling 03ab85191c4352277b662d21e3d6166ed4801df5 on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+1.73%) to 84.76% when pulling c32c66b on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+2.77%) to 85.8% when pulling 316bdd7ace81000a1d38703d10c10d12e472ff3b on krka/tests into dd05c19 on master.

[coveralls]

Coverage increased (+2.78%) to 85.81% when pulling 648ede68d02c5da17bc2af72539c2e8f9edc0636 on krka/tests into dd05c19 on master.

[spkrka]

@protocol7 Maybe I put too many changes in the same PR now... but it would be nice to have some review on the API change - it might leak too many abstractions i guess, but i couldn't think of a cleaner way to expose a more complete connection state

[coveralls]

Coverage increased (+2.78%) to 85.81% when pulling 1361c21 on krka/tests into dd05c19 on master.

[protocol7]

+1, merge away!