Bug#942428: transition: gssdp/gupnp
Package: release.debian.org
Severity: normal
Tags: moreinfo
User: release.debian.org@packages.debian.org
Usertags: transition
Hello!
The gssdp and gupnp transitions are best done together. The transition
has previously been discussed a bit on IRC but I'm filing this for
tracking purposes for getting the final blockers ironed out.
As the dev packages have changed names libg(ssdp|upnp)-1.0-dev to
libg(ssdp|upnp)-1.2-dev the entire transition will need sourceful
uploads. Most reverse dependencies has already been prepared in
experimental. Here are the affected packages and their status:
$ (reverse-depends -r testing -b src:gssdp -l && reverse-depends -r testing -b src:gupnp -l ) | sort -u
caja-extensions - #942343 fixed-upstream <--- NMU?
dleyna-core - fixed in experimental
dleyna-renderer - fixed in experimental
dleyna-server - fixed in experimental
gupnp - fixed in experimental
gupnp-igd - exp FTBFS on armel, tests hangs <--- blocker?
gupnp-tools - fixed in experimental
peony-extensions - no rdeps, unmaintained <--- temporary removal?
rygel - fixed in experimental
upnp-router-control - no rdeps, unmaintained for years <-- permament removal?
The final real blocker as I can see it is the gupnp-igd armel FTBFS.
The problem is 100% reproducible on abel porterbox.
The tests that hangs are the final two:
test_gupnp_simple_igd_invalid_ip
test_gupnp_simple_igd_empty_ip
According to gdb the main thread is hanging in g_object_unref (igd)
that's called last in the test functions.
Since there's no obvious solution to this and I'm not going to invest
(more) time into it nor to track down anyone who cares about armel
issues I'm leaving this here as tagged moreinfo until someone has a plan
for how the transition can happen.
Regards,
Andreas Henriksson
Reply to: