I would need to be convinced, the absolute intensity depends on instrument, sample concentration etc. Yours, Steffen

Well ... maybe people could input absolute intensity but we would have to relativize it and search on that (which would mean the underlying search would be the same as already exists)?
Is the current search dependent on input of relative intensity only?

Hmm, but we need the absolute intensity of the base peak to recalculate to absolute values. I guess, the search algorithm could be the same with some extension. Relative intensity search is the status. I am also not conviced about the absolute intensites. If they would be stable and independent from sample conditions (they are not as @sneumann said), we would have less problems with similarity search at all. I suggest to keep the status, recalculation to relative values is no magic.

Well - if people input absolute intensities, calculating relative intensities is trivial too, based on max(I) of the provided list. There's little point searching on absolute intensities, but one could provide a service to accept absolute intensities, but search relative behind the scenes, this is what any algorithm does anyway

Okay, but this needs some adjustment in the pipeline after peaklist import. I think it is doable and we should think from the service view, not from the freak view.