Confusing automatic ordering of counters in ct, scans etc.
Until now, I did not put much attention to the way how counters are ordered in BLISS. I was already happy getting counter values. But with increasing numbers of counters it becomes annoying that they are automatically ordered, i.e. by controller names and names. This could be acceptable in short counter lists where all counters would fit on a single screen, but it becomes unusable in larger counter lists, e.g. Ncounters>50. In long lists specific counters cannot be found easily. Missing counters can only be found by searching through all counter names. Ordering counters by controllers is probably useful for debugging, but when everything is working other aspects become more important. When I insert counter 'W' between counter 'B' and counter 'A' I expect that it will appear in this order in the output: (counter 'B', counter 'W', counter 'A') and that its output position is not automatically altered by a hidden controller name.
The implementation of this "feature" seems to be deeply embedded in BLISS. It is hard to find where it is done and why it is done.
I propose the following:
- Displaying the counters in the order as given in the measurement group, e.g. like they are written in the yaml file.
- Everybody who needs or detests this feature should have the possibility to switch it individually on or off, e.g. with automatic_counter_ordering_on/off.
- Alternatively, a tuple with counter names in the required order for displaying and saving would already help.
To be clear, I do not ask only for the BLISS output, I would also like that counters appear in the same order in the output files. This gives external programs (e.g. pymca) the possibility of displaying them in the same order and it would be faster of finding a specific counter.