Changes to --csv

[Jon Haddad]

Hey all,

I just wanted to drop a line that there's a slight change to the way we handle CSV reporting. Previously, a directory with a timestamp was created and 3 files were created in that directory.  

Now, --csv takes a filename parameter.  The read, writes and error statistics are written to this file.  This change was made after running a few dozen tests where it became difficult to keep track of which test results corresponded to what test.  This will hopefully make it easier by allowing you to do something like --csv row_cache_enabled.csv.

Hopefully this isn't too much of a hassle for folks.  It's one of the last breaking changes i intend to make.  We're approaching our first release (1.0!) and I think we've mostly settled on the existing feature set, we're just doing a little cleanup, documentation and bug fixes.

We've just gone through the 1.0 milestone on Github.  If anyone's looking to contribute, these two are good first issues to get involved:

https://github.com/thelastpickle/tlp-stress/issues/75

https://github.com/thelastpickle/tlp-stress/issues/89

Thanks everyone, we're looking forward to releasing 1.0.  Please get us your bugs and feedback.

Jon