rahulbot
commented
8 years ago @sandsfish replies """ Nope, this sounds like a non-perfect but better solution than what we currently have. I agree that a lot of the time it’s a quick, simple difference. The trickier part will be the iteration and editing of queries. My main workflow involves starting and either drilling into queries with more terms, or shifting through time ranges. Questions would be, if it’s a query with a custom name already, does it get updated? Do we keep track of whether the current label was the generated one vs. edited? etc. """
@hroberts suggests:
""" I have argued a few times for better query labels in the dashboard without success. I'm going to try again.
I think the following simple strategy would result in good labels for 90% of use cases:
So for the default query that pops up on the screen, the labels would end up being 'truth' and 'beauty'. If you instead wanted to compare coverage of 'truth' in US MSM during the two halves of 2015, the labels would be '2015-01-01 - 2015-06-30' and '2015-07-01 - 2015-12-31'.
This approach will obviously fall down if more than one field differs between the compared queries. And there would have to some code to truncate long values (including long lists of media sets). And the labels are not as descriptive as they would ideally be. In most cases, we would still want to edit the labels before capturing an image for publication.
But 99% of the time I use the tool, it's just to do some quick queries for my own consumption, and this simple approach would be a big improvement over the 'query a', 'copy of query a', 'copy of copy of query a' labels we end up with now. The goofy query labels are the only remaining aesthetically unpleasant part about the tool for me. """