[00:46] <babbageclunk> thumper: fix for the watcher tests: https://github.com/juju/juju/pull/10573
[00:46] <babbageclunk> I don't really understand why it works though - shouldn't the call to StartSync in the NotifyWatcherC have the same effect?
[01:55] <thumper> babbageclunk: do you want me to explain why?
[01:56] <thumper> babbageclunk: and there is a better method
[02:02] <babbageclunk> thumper: yes please!
[02:03] <thumper> babbageclunk: otp now, see PR
[02:03] <babbageclunk> ok, that's clear - thanks!
[02:03] <babbageclunk> updating now
[02:40] <wallyworld> anastasiamac: any chance to look at those PRs? i've tweaked the use of ctx.out in the first one
[02:41]  * anastasiamac looking
[02:53] <anastasiamac> wallyworld: almost :) i think it's painfully close... commented
[02:54] <wallyworld> anastasiamac: ty, will look real soon after meeting
[02:54] <anastasiamac> nws. m not in rush \o/
[03:00] <thumper> babbageclunk: https://github.com/juju/juju/pull/10573 wasn't approved, so your merge didn't happen
[03:00] <thumper> I have approved it now
=== exsdev0 is now known as exsdev
[03:45] <wallyworld> anastasiamac: using a struct for args results in different output to the spec (the Arguments header has a ":" instead of free text). And there's still the need to print the action description as free text
[03:46] <wallyworld> maybe the extra ":" doesn't matter
[03:47] <anastasiamac> wallyworld: yeah i would not worry about ':"
[03:47] <anastasiamac> :D
[03:48] <wallyworld> i also like the extra \n at the end
[03:48] <wallyworld> otherwise it's hard to see where the output ends and the prompt starts
[03:49] <anastasiamac> but inconsistent with all other commands, especially show- ones
[03:49] <wallyworld> the other ones are wrong then :-)
[03:52] <anastasiamac> haha
[03:53] <wallyworld> anastasiamac: there's still the matter of some free text being needed for the description. that still requires fprintf
[03:54] <wallyworld> in which case may as well use fprintf for the "Arguments" header as well
[03:56] <wallyworld> unless we insist on the use of "Description:"
[03:56] <anastasiamac> wallyworld: m wondering if u r really thinking this or just not keen to change the PR...:) fwiw m happy with it to land as is but i think it has these rough edges that could have been avoided from the start....
[03:56] <anastasiamac> wallyworld: i do not think that "DEscription:" is bad :)
[03:56] <wallyworld> what rough edges? just sticking to the spec
[03:56] <wallyworld> the spec calls for description as free text
[03:57] <wallyworld> i can see why that is not a bad thing as it's user readable
[03:57] <anastasiamac> like too much fmt.Printf directly to stderr and stdout - one of the things i want to tackle next and it'd be (and laready is a nightmare) coz of the rogue direct writes to stderr/stdout
[03:57] <wallyworld> it just so happens the printing of the action args is "yaml-like"
[03:57] <anastasiamac> already*
[03:58] <anastasiamac> want to talk it over rather than type?
[03:58] <wallyworld> sure
[03:59] <anastasiamac> stdup?
[03:59] <wallyworld> ok
[04:23] <babbageclunk> thumper: ah, right - thanks!
[04:42] <wallyworld> anastasiamac: https://github.com/juju/juju/pull/10568 should be ok now, has the rename
[04:43] <anastasiamac> yes i saw the update and m looking already :D
[04:43] <wallyworld> you are on the ball
[04:48] <anastasiamac> haha yes u do keep me on my toes :)
=== hml_ is now known as hml
[13:46] <rick_h> hml:  looks like the build came through this time so woot woot
[13:46] <hml> rick_h: rgr
[14:02] <achilleasa> hml or stickupkid: can either of you please take a look at https://github.com/juju/packaging/pull/7?
[14:07] <stickupkid> achilleasa, looking