IETF
roll@jabber.ietf.org
Tuesday, April 26, 2016< ^ >
mcr/credil has set the subject to: IETF96 ROLL WG meeting
Room Configuration
Room Occupants

GMT+0
[08:47:23] eintopf joins the room
[08:51:33] <eintopf> mcr: I will try to send pull-requests. I never did that on github
[10:47:35] eintopf leaves the room: Disconnected: Replaced by new connection
[10:47:35] eintopf joins the room
[12:36:50] <mcr> thanks!
[12:37:14] <mcr> or just git trees to pull from; github part is optional.
[12:37:34] <mcr> I don't mind patches to the mailing list; but git pull requests are much simpler.
[12:47:29] <eintopf> yea, it was just clicking a button
[12:47:47] <eintopf> but testing travis stuff failed somehow
[12:57:03] <mcr> what is the travis report that failed?
[12:57:12] <mcr> travis can be really picky.
[12:58:36] <eintopf> https://travis-ci.org/mcr/unstrung/builds/125788924#L2411
[12:59:26] <eintopf> I think it's the same like in your pull-request: https://travis-ci.org/mcr/unstrung/builds/95932272#L2346
[13:01:00] <mcr> /home/travis/build/mcr/unstrung/programs/sunshine/sunshine --help 2>&1 | tee ../OUTPUTS/sunshine-help.raw | sed -f pathnames.sed | diff - sunshine-help.txt
2,3d1
< interface lo has IPv6 disabled; ignored
< interface eth0 has IPv6 disabled; ignored
[13:03:13] <mcr> oh, yes.  so travis now builds things on VMs with IPv6 disabled.  I don't understand their logic, it's stupid, and they have moved to having new infrastructure which has no IPv6.
[13:03:26] <mcr> I can't understand thinking it was a good idea.
[13:06:20] <eintopf> github itself has also no IPv6 support
[13:07:13] <eintopf> but it's not the reason that the VMs should not have IPv6 support
[14:24:44] <mcr> travis moved from amazon/rackspace VMs to a combination of containers and Google Engine VMs.   The Google VMs have no IPv6. It's ridiculous.
[14:34:20] <cgundogan> For RIOT, we moved away from travis several weeks ago .. it really started to annoy us, or maybe we were not using it correctly
[14:46:26] <mcr> I think that I'm getting to that point as well.  I would boot up UserModeLinux instances to run tests under, since they are really cheap and let me control the kernel side code, but without a better way to debug what is going on, it will just be a time sink.
[17:38:16] eintopf leaves the room: Disconnected: closed
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!