Branko Čibej wrote: > On 15.01.2015 09:51, rhuij...@apache.org wrote: >> URL: http://svn.apache.org/r1651980 >> Log: >> Resolve the issue identified in r1651963, by properly calculating the depth >> of the location where the moved to information should be stored after >>another >> move. >> >> * subversion/libsvn_wc/wc_db.c >> (delete_update_movedto): Make this assertion maintainer only, like other >> similar checks in wc_db.c. >> (delete_node): Fix depth calculation. > > If the depth calculation is now correct and the assertion doesn't get > triggered, the SVN_DEBUG is doubly useless because it masks bugs. If the > SVN_DEBUG had been there before, we'd never have got the original bug > report in the first place. Either leave the assertion in or rip it out.
We should not hesitate to implement alternative kinds of assertion if they are useful. It seems like what is wanted here is a non-fatal assertion: one that checks a condition, reports if the condition is not met, and then continues execution. Sometimes when I start up a GUI from the command line, I see the windowing toolkit library spewing assertion failure messages on the console: $ kdiff3 Object::connect: No such signal org::freedesktop::UPower::DeviceAdded(QDBusObjectPath) Object::connect: No such signal org::freedesktop::UPower::DeviceRemoved(QDBusObjectPath) ... Presumably we would not wish these to go to the console by default, but to a configurable place (file, gui dialogue box, ...). - Julian
