thoughts on adding more details to search, gather, and lca gather output.

[ctb]

I've been trying to think about how to allow people (including ourselves :) to dig into results more - in part because I think it helps them understand what's going on.

We're making some progress, e.g. the latest sourmash lca gather output includes some annotation of duplicate hits. But I'd like more details, esp in this situation - right now it's hard to figure out what duplicate hits mean, are, etc.

So I brainstormed a bit today and here are a few thoughts.

• by default, we could enable --details, which would automatically and silently output details to a text file perhaps via a details() logging function.
• this text file could be named something unique (maybe using a uuid?) so as to avoid append conflicts when multiple things are being run in the same directory;
• then this file could contain all of the gory details around what is being loaded, etc;
• maybe put it in a .sourmash-output directory or something?

cc #271 (long running queries UX improve), #258 (output stuff to disk)