1358 rename Data to data

[postelrich]

closes #1358

rename Data to data and add deprecation warning to Data.

[llllllllll]

is there a reason to make these classes rather than just using the superclass directly?

[kwmsmith]

We originally thought it would be good to have a subclass to make it easily filterable, but now I'm with you, and see no really good reason not to just use DeprecationWarning directly.

[postelrich]

The idea was to so that we can set warnings.simplefilter to only display the warning once but I couldn't get it to work. I created the derived class so that I wouldn't effect how other warnings are displayed, especially since deprecation warnings are ignored by default.

[llllllllll]

How do people feel about the name, bound_symbol or something? data feels too close to what resource is.

[kwmsmith]

How do people feel about the name, bound_symbol or something? data feels too close to what resource is.

As you know, we discussed this at length in another issue thread.

We bikeshedded the naming question again among 4-5 people before starting this PR. We decided to stay with data:

• It's a very common end-user entry point, so we want to keep it short;
• Given the target end user--Data Scientist, Business Analyst, Data Analyst, etc.--bound_symbol was thought to be too far removed from their working vocabulary;
• data, compute, and expression are the three main concepts we use when talking about blaze, so it's a big plus to have those three concepts reflected in the API directly.

[postelrich]

@llllllllll I agree that datais too generic of a name (especially as part of a data science library/distribution) and not descriptive enough of what it's doing. But I also agree with @kwmsmith against naming it something too technical.

At the risk of starting another lengthy discussion, I propose that if we were to change the name that it be simple. So something like blaze.ignite because you're firing up a powerful connection to your data. Or really simple like blaze.on or blaze.open.

But we shouldn't delay progress, rather just keep the discussion open and make a decision before we get to a 1.0.

[llllllllll]

@kwmsmith Sorry, I forgot that we had come to a conclusion in the other thread.

[kwmsmith]

Closing this PR, see #1431.