Re: [Gnu-arch-users] Features command for arch

[James Blackwell]

Adrian Irving-Beer <address@hidden>
>> I hope I'm understanding the discussion so far, but it sounds like
>> 'features' is a cheap way to announce capabilities that are
>> programmatically (but expensively) provable.
>> If that's the case, couldn't your regression tests be your calculator?

Tom Lord wrote:
>
> But nobody who is advocating for a "features" feature has (that I've
> seen) proposed anything so specific as tying "features" to regression
> tests.

I didn't think of it! But I like the sound of the idea.  I like it a
lot.

Ok. Here's a quick thought: 

1. rename the features idea to included-tests
2. front end guys can check against a list of the included test cases
3. If the test case they want doesn't exist, then they know "it doesn't
   work".
4. If the test case they want *does* exist, then they know "it does 
   work"
5. If there is behavior they want to distinguish that they currently
   can't (because there isn't a test case), then they can hack on the
   test cases and submit a new test case for review. 


How's that? 


-- 
James Blackwell          Try something fun: For the next 24 hours, give
Smile more!              each person you meet a compliment!

GnuPG (ID 06357400) AAE4 8C76 58DA 5902 761D  247A 8A55 DA73 0635 7400