12:49:09 <jbk> uugh... one of these days i'm going to completely fix coreadm
12:56:12 <gitomat> [illumos-gate] 15366 ATA PASS-THROUGH (16) is broken -- Jason King <jason.brian.king⊙gc>
12:57:45 <gitomat> [illumos-gate] 15810 TSC sync produces unreliable results with caches disabled -- Keith M Wesolowski <wesolows⊙oc>
14:00:58 <gitomat> [illumos-gate] 15827 allow passing of extra arguments to 'pkg update' when running onu(1ONBLD) -- Hans Rosenfeld <rosenfeld⊙gho>
14:16:40 <sjorge> andyf any plans for an omnios-bloody releas with the tsc changes ?
15:19:01 <yuripv> Woodstock: looks like you missed "not" in "I did see any ill side effects" :)
15:19:11 <Woodstock> oops.
15:19:44 <jbk> at the risk of bikeshedding, how horrible would if be if when SMF logged events (e.g. `svcs -L logfile`), it included the timezone it was using (yes everyone should use GMT, reality is of course less perfect)
15:19:51 <jbk> in it's timestamps
15:23:00 <tsoome> I have to say, I do not like GMT;) When I need log, I need local TZ so I can correlate events as needed.
15:41:54 <gitomat> [illumos-gate] 15773 loader: serial devices list should be dynamic -- Toomas Soome <tsoome⊙mc>
15:43:32 <yuripv> tsoome: you can use UTC if you don't like GMT
15:43:43 <tsoome> ou, really:D
15:43:55 <yuripv> ;)
15:46:18 <jbk> well the proposal isn't to force the TZ, but just to add what TZ is actually being used at the time of the logging
15:50:17 <tsoome> thats is definitely nice
15:50:34 <tsoome> less confusion is good.
15:52:36 <mns> I would think that would be default, no ? 
16:00:47 <sommerfeld> jbk: what timezone format are you proposing?   The excruciatingly unambiguous way would be a GMT offset as I'm not sure the abbreviations are globally unique.
16:01:32 <jbk> yeah..
16:01:36 <jbk> just +nnnn at the end
16:04:51 <sommerfeld> strftime %z does this I believe.
16:21:39 <jbk> yeah, i don't think it'd be a massive code change -- i suspect it'd just be a matter of updating this: https://github.com/illumos/illumos-gate/blob/master/usr/src/cmd/svc/startd/log.c#L358
16:22:08 <jbk> to include '%z'
16:34:40 <Woodstock> i wonder how many log parsers that'll break :)
16:36:58 <jbk> well that's a concern.. though _technically_ i don't think SMF logs are meant to be consumed...
16:37:10 <jbk> i suppose we could gate it behind a config option
16:37:24 <jbk> so new installs get it, existing are unchanged unless explicitly updated
16:37:43 <Woodstock> probably a heads up would be sufficient
16:56:23 <sommerfeld> Log file format/contents don't appear to be documented.    A heads up would be sufficient.   If you were going to make the format tunable, probably best to go with a property that specifies a strftime format string rather than an old/new boolean.
17:46:24 <gitomat> [illumos-gate] 14001 dboot: dump_tables off by one error -- Keith M Wesolowski <wesolows⊙oc>
18:10:58 <mns> Making it a tunable property, with the default being what the current setting is, would be best option.  Combined with a heads up.  If anyone is using log parsers, it wont break them.