On Tue, 9 Dec 2008 15:33:02 -0900, Jeff wrote:
> jspaleta AT
gmail.com
> python-basemap
> python-basemap-0.99-5.fc10.i386 requires libgeos-3.0.1.so
Hmm geos-3.0.3 went out as an update and bumped the library and it
appears that it skipped testing and went straight to stable and was
not marked as security. Joy!
Let's see if I can get python-basemap rebuilt.
Now before I go ballistic over the fact that this skipped testing and
isnt marked as security... was that just a mistake or was that
intended?
It hopefully tells some people that often I'm equal to an early-warning
system. And if I complain that updates ought to spend some time in
updates-testing first [1], this is based on experience. More breakage in
this report, earlier reports, and private reports based on other check
scripts, was also caused by skipping updates-testing. One packager this
time did not know about releng's buildroot overrides and didn't ask
either. So, once again it is both a documentation and communication
problem.
--
[1] And I never suggested that it should not be possible to skip
"testing" for special updates.