« first day (2901 days earlier)      last day (1881 days later) » 

06:09
@LeonidShifrin Thanks for all the comments on the EntityFramework question. When I logged in and saw seven comments my first thought was, "Boy, I must have really fouled up that response" :) I have made some minor revisions to avoid implying that compilable query operators can only be used for filtering. But I hope you will forgive me for not trying to incorporate more of your observations into an already TL;DR answer :)
06:24
@LeonidShifrin See above.
 
9 hours later…
15:31
@WReach Though might be out of context for the answer here: mathematica.stackexchange.com/questions/211956/…, I never knew about the Import["! command 2&1", "Text"] thingy. Can you explain this syntax a bit?
 
2 hours later…
17:23
@PlatoManiac As described in Streams and Low-Level Input and Output, if we prefix an input stream name with an exclamation mark then the stream data will be the output that results from executing the command that follows the exclamation mark. We can use any input operation: Read, ReadList, Get, Import, etc.
I use Import["!command", "Text"] to get the output as a single sting without any further interpretation.
As for the syntax "2>&1", this is not a Mathematica feature but rather a command shell feature supported by many shells (both on Unix and Windows). It indicates that the stderr stream (channel 2) is to be redirected to the stdout stream (channel 1). This way, any error messages get captured by the Import as well.
Such redirection is a bit hacky since stderr and stdout messages might be interleaved somewhat randomly. But it works reasonably well in interactive contexts where we are just interested in whether and how things might have failed.
 
2 hours later…
19:42
@WReach Your post is awesome, and I would never say otherwise. It is also very important for us at this pretty early stage of adoption of this framework to have such strong proponents who can contribute high quality answers. For this particular one, I was thinking of giving one but I clearly see that it would've been far inferior to yours.
@WReach The comments I made were in now way aimed at diminish it, just to clarify a few points.

« first day (2901 days earlier)      last day (1881 days later) »