Target Again

I have another one:

In MR !259 (closed) , you @schaefed suggest to remove the target parameter since it can be emulated via copy - i beg you not to do that! While writing the tutorials, the target parameter comes in so handy and natural, and on the other hand, it is so annoying to have to insert a copy line, whenever generating a variable, i think it would be a great loss really for the intuitivity of saqc usage.

Also there is another inconsistency that i found hard to explain in tutorial. For data processing functions (like polynomial fit/interpolation...), defaultly, the processed column (=field) gets overridden. But with generic.process it is the other way around - the field parameter acts as kind of a target parameter and you do not have to generate a copy before calculation - that is hard to sell as consistent.

Then i do not know, how you plan to incorporate the fields extension consistently, in case it is not interpreted as a vectorized application, but as a multivariable application, as discussed in #141 (closed) . (that would be something like: generic.process(fields=['a', 'b'], func=lmbda x,y: x-y)... dont we need a target parameter here anyway, or would both, a and b be overridden in that case?

I think we should discuss that soon, since it influences the dramaturgy of the cookbooks i am working on!