Last night, Google Squared – which I think was billed in some circles as Google’s answer to Wolfram ‘not intended to be a Google killer’ Alpha appeared on the public web (announced here: Square your search results with Google Squared).
Pop in a query, and you’ll get a list of results as a grid, with the main result listed in the most left hand column and properties of that result in the following columns. So for example, here’s my Google Squared search for population european countries:
The ability to add columns (as well as delete columns) means you can customise the results grid and save it (although not name it?)
(One thing to be wary of is that you can add extra rows ‘get more results’ style before you save the grid, but once saved, it looks like you then need to add additional row elements in the left hand column by hand.
So that’s all very well and good, and maybe useful – or not. It’s easy enough to use, and while there doesn’t appear to be a public API yet, I wouldn’t have thought it would be too hard to support e..g.a CSV export from a query or saved square.
Anyway, I duly tweeted the launch last night, and then found this response this morning:
Hmmm….that sort of reminds me:
I’m going to assume you’ve read that post in what follows, rather than repeat myself, except for this:
In order to do a “bulk fact lookup”, enter a list of row label entity values into a spreadsheet, a set of attribute (fact) values as column headings, enter an appropriate ‘relativised GoogleLookup formula” into the first table data cell, and then drag it across the set of table data cells: voila – a bulk query made in a single shot across multiple entity and attribute values:-)
So, hmmm, Google squared is just a generalisation and simplification of this? That is, where I was relying on the existence of particular Google Lookup functions, Google suqard tries to populate a column on the basis of a query term in the column head? So I wonder? Is =googlelookup being used in Google Squared and working hard with arbitrary search term’s; is it being automatically relativised to take a second query term from the left hand most column of a Google Square as the basis for a property search defined by each column header? Did I, in fact, invent somthing very much like Google Squared some time ago?! ;-)
PS the spreadsheet solution has th benfit of offering an API, CSV outputs etc.
PPS wouldn’t it be nice if you could seed a column header with a Google Visualization API query language query?;-)