On Monday 19 July 2004 12:34 am, Guillaume Melquiond wrote:
That's strange. I wouldn't have proposed a patch without first running the regression tests.
I hope it isn't compiler-dependent :( Which compiler are you using? Also, are you checking the program exit code? The test doesn't actually print anything, it just returns -1 on failure. I've changed the test to complain loudly if something does fail.
So because it was failing, I was looking for a patch. And I only submitted such a patch when there was no more failure in adjacency_matrix_test (what would have been the benefit otherwise?). Moreover, I hope you agree with me that the previous code was wrong. Adding a positive value to an end() iterator is obviously an error.
I didn't mean to sound accusing; my apologies if I came across that way. The pre-patch code is clearly wrong; I'm just not sure how to fix it at the moment.
Consequently, I'm not sure the failure is caused by my patch. Maybe another patch somewhere (not necessarily in the Graph library) did break something. For example, I see that 196 lines were changed in adjacency_matrix_test.cpp yesterday. So maybe...
I looked at the changes to adjacency_matrix_test.cpp; they're only formatting changes ("cvs diff -b" shows no differences). Reverting the patch to boost/graph/adjacency_matrix.hpp does "fix" the problem for me (i.e., adjacency_matrix_test passes now). I can dig into this later today or this evening, unless you have a chance; I don't understand the representation well enough to put in a quick fix now. Doug