[12:36] <Odd_Bloke> smoser: I'm seeing problems upgrading cloud-init on yakkety: http://paste.ubuntu.com/23177787/
[12:36] <Odd_Bloke> smoser: Known issue?
[12:36] <Odd_Bloke> smoser: (It looks like the two versions in question there are the xenial-updates and xenial-proposed versions, so we might see this in xenial as well)
[13:07] <smoser> Odd_Bloke, it causes error ?
[13:08] <Odd_Bloke> smoser: Ah, no, not now I look at it properly.
[13:09] <Odd_Bloke> smoser: It was just the last package to be installed before other failures were reported.
[13:09] <Odd_Bloke> And I can't read, apparently. :p
[13:09] <smoser> ok. it is definitely a result of me having dropped the list of 'dirs' from cloud-init that were in the packages.
[13:09] <smoser> because i was advised that it wasnt needed.
[18:29] <harlowja> @smoser qq
[18:29] <harlowja> for package building, whats the recommended way to not have `X.Y.Z+xxx.gHASH` show up
[18:29] <harlowja> and just force X+Y+Z (even if there are additions)
[18:29] <smoser> you dont like that ?
[18:29] <smoser> hm..
[18:29] <harlowja> its ok for dev
[18:30] <harlowja> but we have some add-on changes (cc_godaddy)
[18:30] <smoser> hm.
[18:30] <harlowja> pbr has a way to force a explicit version
[18:30] <harlowja> maybe we can do the same
[18:30] <smoser> i dont think i considered that ata ll.
[18:30] <harlowja> seems like it partially exists in read_version
[18:30] <smoser> right
[18:30] <harlowja> read_version has
[18:30] <harlowja> use_long = '--long' in sys.argv or os.environ.get('CI_RV_LONG')
[18:30] <harlowja> use_tags = '--tags' in sys.argv or os.environ.get('CI_RV_TAGS')
[18:30] <harlowja> which are sorta close, ha
[18:30] <harlowja> just no turn-off-the-thing
[18:30] <smoser> but it *is* useful to have the hash avialable
[18:30] <smoser> even from where you started i think.
[18:31] <harlowja> ya, maybe i should just tell the guy that doesn't like the dirty name to be quiet, lol
[18:34] <harlowja> :-
[18:34] <harlowja> :-p
[18:37] <harlowja> perhaps still have the json write
[18:37] <harlowja> but have the version reported still be something that can be set?
[18:38] <harlowja> i think the json still is useful
[20:23] <smoser> mgagne, can you verify https://bugs.launchpad.net/cloud-init/+bug/1605749 in xenial ?
[20:52] <mgagne> smoser: yes, will rebuild image with latest package and see what happens
[23:22] <harlowja> @smoser do u know if https://git.launchpad.net/cloud-init/tree/setup.py?id=0.7.7#n196 is really needed anymore
[23:22] <harlowja> sucky part for that one is ./brpm doesn't know about it being added on :-P
[23:22] <harlowja> but i can't remember why we are still adding that in, lol
[23:27]  * nrezinorn waves at harlowja 
[23:27] <harlowja> sup
[23:28] <nrezinorn> lol its jim :)
[23:28] <harlowja> hi jim
[23:29] <harlowja> smoser this jim is saying brpm not working cause of cheetah being dynamically included ;)
[23:29] <harlowja> which makes sense :-P
[23:29] <nrezinorn> what i said was, and i paraphrase "fix it josh"  lol
[23:30] <harlowja> lol
[23:30] <harlowja> then i was like, man those cloud-init people like patches
[23:30] <harlowja> (especially from new people)
[23:30] <harlowja> lol
[23:30] <nrezinorn> lol
[23:33] <harlowja> (and then we ended up here)
[23:33] <harlowja> in happy land
[23:46] <harlowja> also argparse is weird
[23:55] <harlowja> cause apparently we are seeing that if its listed in requirements, and cent7 says yup its installed (it lies since its built-in) that python later complains (saying its not there)
[23:55] <harlowja> which is a lie
[23:55] <harlowja> nrezinorn found that one, i guess i never hit cause i always used a machine that had a older cloudinit on it before