[#67346] Future of test suites for Ruby — Charles Oliver Nutter <headius@...>

I'll try to be brief so we can discuss all this. tl;dr: RubySpec is

19 messages 2015/01/05

[ruby-core:67557] Re: [ruby-cvs:56375] usa:r49225 (trunk): * test/ruby/test_numeric.rb (TestNumeric#test_coerce): fixed wrong message.

From: "U.Nakamura" <usa@...>
Date: 2015-01-13 04:05:03 UTC
List: ruby-core #67557
Hi,

In message "[ruby-core:67555] Re: [ruby-cvs:56375] usa:r49225 (trunk): * test/ruby/test_numeric.rb (TestNumeric#test_coerce): fixed wrong message."
    on Jan.13,2015 06:30:25, <normalperson@yhbt.net> wrote:
> Oops, r49225 is a 64-bit specific change and my original r49224
> was only tested on 32-bit.

Oh, flonum...


> How about making error messages consistent across platforms?

Agree.
But, shouldn't we show the value instead of the class name?
About a value which can be expressed as a literal, showing the value
itself is more helpful than showing the class, in my opinion.


Regards,
-- 
U.Nakamura <usa@garbagecollect.jp>

In This Thread

Prev Next