New testsuite failure in today's CVS



David Ronis wrote:
> I just rebuilt against CVS (as of 1/2 hr ago).  I get the following
> testsuite failure:
> 
> ********************** Problem 24 ***************
> Input:
> solve(%, [x, y])
> 
> 
> Result:
> [[x = 2, y = 2], [x = .5202594388652008 %i - .1331240357358706, 
> y = .07678378523787788 - 3.608003221870287 %i], 
> [x = - .5202594388652008 %i - .1331240357358706, 
> y = 3.608003221870287 %i + .07678378523787788], 
> [x = - 1.733751846381093, y = - .1535675710019696]]
> 
> This differed from the expected result:
> [[x = 2, y = 2], [x = .5202594388652008 %i - .1331240357358706, 
> y = .07678378523787777 - 3.608003221870287 %i], 
> [x = - .1331240357358706 - .5202594388652008 %i, 
> y = .07678378523787777 + 3.608003221870287 %i], 
> [x = - 1.733751846381093, y = - .1535675710019696]]
> 
> 70/71 tests passed.

I can confirm this.  Seems to be a round off problem.

But when did solve start returning numeric answers?  I thought it 
returned symbolic answers.  (Perhaps it was always that way and I never 
noticed since this test used to pass.)

Ray