Skip Menu |

This queue is for tickets about the POE CPAN distribution.

Report information
The Basics
Id: 25879
Status: resolved
Priority: 0/
Queue: POE

People
Owner: Nobody in particular
Requestors: pause [...] erichacker.com
Cc:
AdminCc:

Bug Information
Severity: Important
Broken in: 0.9989
Fixed in: (no value)



Subject: IPv6 tests fail in Hell (WinXP) with IPv6 enabled.
C:\Perl\cpan\build\POE-0.9989>prove t\30_loops\00_base\wheel_sf_ipv6.pm -v -b t\30_loops\00_base\wheel_sf_ipv6....Your vendor has not defined Socket macro AF_INET6, used at t\30_loops\00_base\wheel_ sf_ipv6.pm line 22 BEGIN failed--compilation aborted at t\30_loops\00_base\wheel_sf_ipv6.pm line 39. dubious Test returned status 2 (wstat 512, 0x200) FAILED--1 test script could be run, alas--no output ever seen Also, due to the dependency: Compilation failed in require at t/30_loops/50_tk/wheel_sf_ipv6.t line 18. C:\Perl\cpan\build\POE-0.9989>perl -v This is perl, v5.8.8 built for MSWin32-x86-multi-thread (with 25 registered patches, see perl -V for more detail) Copyright 1987-2006, Larry Wall Binary build 817 [257965] provided by ActiveState http://www.ActiveState.com Built Mar 20 2006 17:54:25 Running Socket6 v 0.19 installed through PPM from either trouchelle.com or bribes.org. I don't know which. I do know that there was a bug in Want once due to the PPM at ActiveState. IPv6 is up: netsh interface ipv6>show int Querying active state... Idx Met MTU State Name --- ---- ----- ------------ ----- 8 2 1280 Disconnected Teredo Tunneling Pseudo-Interface 7 0 1500 Connected LAN2 6 0 1500 Connected LAN1 5 0 1500 Connected VMware Network Adapter VMnet1 4 0 1500 Connected VMware Network Adapter VMnet8 3 1 1280 Connected 6to4 Tunneling Pseudo-Interface 2 1 1280 Connected Automatic Tunneling Pseudo-Interface 1 0 1500 Connected Loopback Pseudo-Interface To get past this I changed line 18 to: elsif ($^O =~ qr(win)i ) { and was able to pass the non-network tests.
I'm trying to resolve this a different way. I've wrapped the use of AF_INET6 in a block eval, and the tests should be skipped if that use fails. That ought to be more flexible than just skipping outright based on $^O. I don't have an environment suitable to test this fix. I'm just trusting it'll work. Please reopen this ticket (or what the heck, open a new one if you want) ... if the problem persists in svn or after the next CPAN release. Thanks for a tidy, detailed report. The fix has been committed as revision 2195.