Steve M. Robbins wrote: > Hello again, > > On Sat, May 31, 2008 at 02:57:27PM -0500, Steve M. Robbins wrote: >> On Sat, May 31, 2008 at 02:44:12PM -0500, Steve M. Robbins wrote: > >>> So bjorn.haxx.se says: >>> >>> Checking cppunit >>> >>> * trying to update cppunit from 1.12.0-3 to 1.12.1-1 (candidate is 98 >>> days old) >>> * Updating cppunit makes 2 non-depending packages uninstallable on >>> i386: libsp-gxmlcpp-dev, libsp-gxmlcpp1c2a >>> o binary package libsp-gxmlcpp-dev is part of source package >>> sp-gxmlcpp >>> + sp-gxmlcpp has the same version in unstable and testing >>> (1.0.20040603-3) >>> o binary package libsp-gxmlcpp1c2a is part of source package >>> sp-gxmlcpp (explained above) >>> >>> >>> This leaves me with three questions: >>> >>> 1. How can cppunit make a non-depending package uninstallable? >>> >>> 2. The gxmlcpp packages are already the same version on testing and >>> unstable, so if there would be a problem moving cppunit to testing, >>> the same problem must already exist for unstable. What is the >>> problem? >> OK, so I discovered the answer to #2: it is indeed uninstallable in >> sid; see below. I think gxmlcpp just needs to be rebuilt to cope with >> the cppunit SONAME change (cppunit lib package is now >> libcppunit-1.12-1). > > Thanks to Philipp Kern for scheduling binNMUs of gxmlcpp in > response to my message. > > I see that they are now all finished and installed. Yet the PTS for > gxmlcpp doesn't show the +b1 packages in unstable. And the haxx.se > excuses for cppunit still read as above; i.e. no mention of the +b1 > version, and still the claim that they have the same version in > unstable and testing. > > What's up with that?
The PTS shows source packages, you have to look at packages.debian.org to see the binary packages. cppunit and the binNMUs of sp-gxmlcpp have to transition to testing together otherwise cppunit would still make the non binNMUed sp-gxmlcpp binary packages uninstallable => I have added a hint. Cheers Luk -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
