On Mon, Feb 27, 2017 at 04:09:09PM +0000, Guenter Milde wrote: > On 2017-02-27, Kornel Benko wrote:
> > I think, Scott's question is because of > > export/doc/es/EmbeddedObjects_pdf5_texF and > > export/doc/es/EmbeddedObjects_dvi3_texF Yes. > I understand. However, as > > >> Test #2102: > >> SUSPENDED.UNRELIABLE.WRONG_OUTPUT_export/doc/es/EmbeddedObjects_dvi3_texF > >> Test #2109: > >> SUSPENDED.UNRELIABLE.WRONG_OUTPUT_export/doc/es/EmbeddedObjects_pdf5_texF > > belong to the 12 passing tests, I see no reason to invert them. > > So, to repeat my answer to Scott: > > No, this is not equivalent: > > * there is a difference > * the difference is intended. > * It may be that you get different results. If these two tests fail at your > site, feel free to invert (with proper comment, under the proper subtag -- > depending on the failure reason or under TODO, if there is no time to > explore now). Thanks for the explanation. I actually did not think about whether the changes made sense. I was just surprised that there *were* changes, but I think that's because I did not correctly understand your commit message. I thought it meant that you were just sorting. I understood this to mean that you did not intend for any changes in functionality to be made. Another reason I asked is that I've often tried to combine multiple lines into a one-line regex intending for it to be equivalent and I realize later that I made a mistake in the regex. Now when I do that, I try to remember to check with ctest -N before and after to make sure that my change was indeed equivalent. I'll run the full tests and report my failures and us ctesters can compare results and decide together if something should be changed. In any case, thanks for your work on the tests! Scott
signature.asc
Description: PGP signature
