tutorial 22 does not give right solution about hermes-legacy

[Zhonghua]

I remember this issue during the summer, but I did not figure out this at that moment.

[l-korous]

Actually, it works just fine for implicit Euler, but for Crank-Nicolson scheme, there is something wrong when calculating the second time step.

[mhanus]

Do we still have this example?

[l-korous]

Hi Milan, I think it is P03-03-newton-heat-ie (but Crank-Nicolson is not used any more).

Right now we (so far mostly me) are in the process of changing all examples to the new WeakForm etc. style in Pavel Karban's "weakform" branch. So maybe if there is a will to re-implement this example, it can already be done in this branch (that is supposed to become master as soon as all examples are converted).