William S Fulton
2016-03-12 12:45:43 UTC
Hi Simon
The SWIG test-suite has been failing on Travis for 4 days now due to your
recent Scilab changes. As a temporary workaround I've changed .travis.yml
to accept failures in Scilab. Having failures for this long isn't up to
standard and consequently I've revoked your commit privileges to the SWIG
repository. I will restore them once you have submitted a patch to fix the
errors and understand that in the future, you will do your best to avoid
and fix Travis failures. The master branch is also expected to be in a
'ready to release' state. Can you fix this in the next couple of days, if
not please revert the changes. The email address used in the commits isn't
valid either... can you sort that out please?
I have also turned on warnings as errors as without this we can't ensure
the quality of the wrappers is good enough. Scilab was the only tested
module that never had this fully turned on. Hence can you also please
address all the remaining C++ warnings in a 2nd patch?
Finally, I don't understand why 54b84b8fa1789958e2f2ede69edd7b9e9af88dfa is
necessary. This looks like a backwards compatibility issue that is not
explained nor documented in the changes file. We normally avoid breaking
backwards compatibility unless absolutely necessary.
Thanks
William
The SWIG test-suite has been failing on Travis for 4 days now due to your
recent Scilab changes. As a temporary workaround I've changed .travis.yml
to accept failures in Scilab. Having failures for this long isn't up to
standard and consequently I've revoked your commit privileges to the SWIG
repository. I will restore them once you have submitted a patch to fix the
errors and understand that in the future, you will do your best to avoid
and fix Travis failures. The master branch is also expected to be in a
'ready to release' state. Can you fix this in the next couple of days, if
not please revert the changes. The email address used in the commits isn't
valid either... can you sort that out please?
I have also turned on warnings as errors as without this we can't ensure
the quality of the wrappers is good enough. Scilab was the only tested
module that never had this fully turned on. Hence can you also please
address all the remaining C++ warnings in a 2nd patch?
Finally, I don't understand why 54b84b8fa1789958e2f2ede69edd7b9e9af88dfa is
necessary. This looks like a backwards compatibility issue that is not
explained nor documented in the changes file. We normally avoid breaking
backwards compatibility unless absolutely necessary.
Thanks
William