2015-01-06 7:18 GMT+09:00 Charles Oliver Nutter <headius / headius.com>:
> I'll try to be brief so we can discuss all this. tl;dr: RubySpec is
> valuable, MRI tests are valuable, we need to better utilize both of
> them.

I like a smaller test framework.
mspec is larger than our minitest4-based one.
Ruby implementation is sometimes broken.
Larger test framework tends to cause a problem (such as SEGV) in the
test framework itself.
It is difficult to invesitigate the problem for larger test framework.
I'm tired for this situtation until Ruby 1.8 which use old test/unit
(ancestor of test-unit gem).
I'm happy in this aspect since Ruby 1.9.

RubySpec uses "should" style which is not recommended by RSpec now.
http://myronmars.to/n/dev-blog/2012/06/rspecs-new-expectation-syntax
I guess people familiar with "should" syntax will decrease.

Changing a tool is not enough for readabiliy.
Who wrote hard-to-read tests in a tool will still write hard-to-read
tests in another tool.
Some kind of feedback for hard-to-read tests should be considered to
address readability.
It is a long way to go, regardless of a tool.
If such feedback is not designed, changing a tool causes more trouble
than benefits.
-- 
Tanaka Akira