[00:25:29] --- Russ has left: Disconnected
[01:20:26] --- lars.malinowsky has become available
[05:49:55] --- steven.jenkins has left
[06:09:30] <jakllsch> i wish i was more excited about doing everything all at once
[06:10:01] <jakllsch> i have a feeling i'm going to have to push like 5 more patchsets before this gets merged
[07:12:49] --- steven.jenkins has become available
[07:23:33] --- lars.malinowsky has left
[07:47:56] <jakllsch> oh well, on to patch set 4 :-)
[07:52:16] --- deason_gmail has become available
[08:00:16] --- reuteras has left
[09:07:59] --- lars.malinowsky has become available
[09:27:23] --- Simon Wilkinson has become available
[10:24:10] --- rra has become available
[10:31:05] --- Simon Wilkinson has left
[13:02:46] <jakllsch> shadow: would using sys_setgroups be preferable to duplicating it?
[13:06:33] <jakllsch> i hate it when i can't explain to myself why something is the way i made it
[13:12:56] --- jaltman/FrogsLeap has left: Replaced by new connection
[13:12:58] --- jaltman/FrogsLeap has become available
[13:16:02] <shadow@gmail.com/barnowl754E0B64> > shadow: would using sys_setgroups be preferable to duplicating it?

well, why does kauth_cred_setgroups not work? and if it really
doesn't, why can osi_{s,g}etgroups use kauth methods
[13:23:02] <jakllsch> sys_setgroups does more than just kauth_cred_setgroups
[13:24:50] <shadow@gmail.com/barnowl754E0B64> ok. consider commenting it that that's true. 
[13:37:57] <jakllsch> setpag() needs to work as non-root, right?
[13:39:04] <deason_gmail> for PAGs to "work" yes
[13:40:21] <jakllsch> but setgroups(2) still must be privileged?
[13:40:57] <jakllsch> anyway.
[13:42:15] <deason_gmail> yes...? we're not trying to set a pag just from userspace
[13:43:25] <deason_gmail> (but I'd expect at this point a stable client is more desirable than PAGs working)
[13:45:43] <shadow@gmail.com/barnowl754E0B64> yes, setpag() needs to work for all.
[13:46:14] <shadow@gmail.com/barnowl754E0B64> and yes, no PAGs on macos, so, well, people can live without.
[13:46:19] <shadow@gmail.com/barnowl754E0B64> at least at a start.
[13:58:21] <jakllsch> any comment i try to conjure seems painfully obvious
[13:59:58] <shadow@gmail.com/barnowl754E0B64> /* results visible via kauth_cred_getgroups. also does other work */
[14:20:43] <phalenor> hmm, would the data corruption bug not show up when using memcache?
[15:00:32] <rra> It only shows up if you're retrieving files larger than chunksize, and yeah, it's possible that it's disk cache only since it was related to fixes for reading past the end of files.
[15:02:17] <phalenor> we're running pre5 with the patch, and that seems fine
[15:02:26] <rra> Yeah, pre5 with the patch is fine.
[15:44:31] --- deason_gmail has left
[15:59:54] --- jaltman/FrogsLeap has left: Disconnected
[20:27:27] --- jaltman/FrogsLeap has become available
[22:51:06] --- lars.malinowsky has left
[22:52:14] --- reuteras has become available
[23:00:32] --- rra has left: Disconnected
[23:18:39] --- Russ has become available