SHA1-based copy checking breaks when there are expired sources in the target
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Launchpad itself |
Fix Released
|
Critical
|
Steve Kowalik |
Bug Description
I can copy-package.py from cocoplum to copy a package from a private PPA to a public one and it broke
lp_archive@
2010-05-04 21:38:01 INFO Creating lockfile: /var/lock/
2010-05-04 21:38:08 INFO FROM: kubuntu-ninjas: lucid-RELEASE
2010-05-04 21:38:08 INFO TO: kubuntu-ppa: lucid-RELEASE
2010-05-04 21:38:08 INFO Copy candidates:
2010-05-04 21:38:08 INFO kde4libs 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs-bin 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs-bin 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-data 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-dbg 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-dbg 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-dev 4:4.4.3-
2010-05-04 21:38:08 INFO kdelibs5-dev 4:4.4.3-
2010-05-04 21:38:08 INFO libplasma3 4:4.4.3-
2010-05-04 21:38:08 INFO libplasma3 4:4.4.3-
Traceback (most recent call last):
File "/srv/launchpad
script.
File "/srv/launchpad
implicit_
File "/srv/launchpad
self.main()
File "/srv/launchpad
self.mainTask()
File "/srv/launchpad
self.
File "/srv/launchpad
copy_
File "/srv/launchpad
self.
File "/srv/launchpad
sha1 = library_
AttributeError: 'NoneType' object has no attribute 'sha1'
OOPS-1586C2359 as an example.
Changed in soyuz: | |
status: | New → Triaged |
importance: | Undecided → Critical |
assignee: | nobody → Steve Kowalik (stevenk) |
Changed in soyuz: | |
status: | Triaged → In Progress |
tags: | added: oops |
description: | updated |
Changed in soyuz: | |
status: | In Progress → Fix Released |
milestone: | none → 10.05 |
This is going to break any copies where the target archive has expired sources with the same name. In that case we will have to relax the check -- the SHA1 data is actually gone from the DB! That's pretty bad, but I don't think there's much choice.