On 06.01.19 22:34, Bill Cole wrote: > On 6 Jan 2019, at 15:56, Dave Horsfall wrote: > >> Odd... >> >> ozzie:~ dave# port upgrade -p outdated >> ---> Computing dependencies for bitkeeper >> ---> Staging bitkeeper into destroot >> ---> Installing bitkeeper @7.3.3_0 >> ---> Cleaning bitkeeper >> ---> Computing dependencies for bitkeeper >> ---> Deactivating bitkeeper @7.3.2_1 >> ---> Cleaning bitkeeper >> ---> Activating bitkeeper @7.3.3_0 >> ---> Cleaning bitkeeper >> >> Huh? Did the "-p" flag really make that difference? > > I would expect it to. That flag tells 'port' to ignore the failure in > the 'destroot' stage of installation and proceed to archiving the > (presumably bad) staged package and installing it from the somehow bad > destroot directory.
No, that is not correct. The -p flag makes the port command to proceed with the next port, but it does not override errors in a phase. I assume the destroot.cmd worked on the second execution for whatever reason. Maybe the dependencies within the BitKeeper build system are not correct, causing it to assume this already succeeded previously...? > I would NOT expect Bitkeeper to work. I would also expect 'port contents > bitkeeper' to show less than the necessary compliment of files. That would be a good assumption, it is most probably missing some files. Rainer
