November 11, 2010 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | >
November 11, 2010 [00:00:35] <gwr> What's the deal with these companies that "park on" lots of domain names, hoping someone will pay them for one?[00:00:45] <gwr> Seems like a basically evil business model...[00:00:57] <e^ipi> like real estate[00:01:00] <e^ipi> you can't own land, man[00:01:06] <gwr> Any idea what the going rate is?[00:01:15] <e^ipi> *shrug* way more than it's worth[00:01:15] <gwr> (if one is to play their game and make a bid?)[00:01:32] <McBofh> gwr: depends on how desperate you are[00:02:07] <gwr> here's one of them: http://www.mdnhinc.com[00:02:35] <joffe> buy something porn related[00:02:38] <joffe> porn always sells[00:02:41] <Triskelios> http://most-expensive.net/domain-name[00:02:46] <joffe> just think of something insane[00:02:52] <joffe> whatever you can think of.. someone's into it[00:03:05] <joffe> midgets wearing differently colored socks[00:03:10] <joffe> turn it into a cool, short url[00:03:11] <joffe> winner.[00:04:25] <McBofh> fark[00:04:37] <McBofh> first one I see on mdnhinc.com is adaoptababy.com, for USD100k[00:05:43] <FrostCS> for when you want a kid, but you really never learned how to spell...[00:05:44] <gwr> I hope nobody pays those a-holes that kinda money.[00:06:22] <Triskelios> hah "Name: adoptababy.com Category: Entertainment"[00:06:32] <gwr> Looks like they just park on _all_ the short names.[00:07:14] <FrostCS> it's like ostrich racing Triskelios[00:08:49] <e^ipi> that's a strange analogy to use[00:14:44] *** gebi has quit IRC[00:25:47] *** lyarick has joined #illumos[00:31:11] *** nsuperbus has joined #illumos[00:35:29] *** nachox has quit IRC[00:41:02] *** cuguardo has joined #illumos[00:45:45] *** estibi has quit IRC[00:50:25] *** jamesog_ is now known as jamesog[00:52:34] *** pjfloyd has quit IRC[00:55:54] *** estibi has joined #illumos[01:03:28] *** estibi has quit IRC[01:03:52] *** estibi has joined #illumos[01:07:44] *** gber has quit IRC[01:22:29] *** estibi has quit IRC[01:23:59] *** PatrickS has joined #illumos[01:25:17] *** Aram has quit IRC[01:27:25] *** estibi has joined #illumos[01:36:23] *** trbs has quit IRC[01:38:51] *** dreh has quit IRC[01:44:47] <lewellyn> gdamore: alanc: next commit should be fixed. i think.[01:47:15] *** dreh has joined #illumos[01:48:23] *** cuguardo has quit IRC[02:04:57] *** glagasse has quit IRC[02:10:12] *** sidh_ has joined #illumos[02:10:53] *** sidh has quit IRC[02:11:01] *** bradend has quit IRC[02:11:39] *** bradend has joined #illumos[02:12:18] *** venture37 has quit IRC[02:14:13] *** postwait has quit IRC[02:19:47] *** ganbold has joined #illumos[02:25:57] *** jamesd2 has quit IRC[02:26:44] *** cuguardo has joined #illumos[02:27:32] <cuguardo> Anybody know where I can RTFM on ipadm?[02:31:25] <gwr> http://lmgtfy.com/?q=opensolaris+ipadm[02:31:30] <gwr> :)[02:33:42] <lewellyn> smrt: explain ipadm[02:33:42] <smrt> Um. I seem to not know anything about ipadm...[02:33:48] <lewellyn> teach what you learn to the bot ;)[02:33:53] <cuguardo> gwr: Thank you.[02:33:54] <cuguardo> First, I tried to setup interface by `ifconfig e1000g1 inet xx.xx.xx.xx/16'[02:33:55] <cuguardo> but got `ifconfig: could not create address:Operation not supported' - this thing?[02:34:50] *** stallion has quit IRC[02:40:05] *** dnm has quit IRC[02:40:44] <cuguardo> ipadm create-addr -T static -a local=10.75.0.128 e1000g1/v4[02:40:44] <cuguardo> ipadm: Could not create address: Persistent operation on temporary object[02:40:45] <cuguardo> Holy smoke 203[02:41:26] *** Edgeman has quit IRC[02:48:24] *** postwait has joined #illumos[02:50:52] *** nsuperbus has quit IRC[02:52:33] *** boyd_ has joined #illumos[02:53:33] *** boyd has quit IRC[02:53:33] *** boyd_ is now known as boyd[02:55:11] *** edp_mobile has joined #illumos[03:02:59] *** piwi has quit IRC[03:03:08] *** arek_deepinit has quit IRC[03:04:57] <jkimball4> oi doesn't seem to have a man page either.. wonder if that's being addressed in 148[03:07:03] *** boyd has quit IRC[03:07:33] *** boyd has joined #illumos[03:07:47] *** piwi has joined #illumos[03:09:41] <cuguardo> ipadm works well after that:[03:09:42] <cuguardo> svcadm disable network/physical:nwam[03:09:43] <cuguardo> svcadm enable network/physical:default[03:09:44] <cuguardo> by root only - account even with priv. <Primary Administrator> can't access to DLPI layer.[03:14:30] *** Teknix has quit IRC[03:23:15] *** boyd has quit IRC[03:24:04] *** boyd has joined #illumos[03:25:21] <cuguardo> cool:) channel members count equals last build no[03:26:28] *** Edgeman has joined #illumos[03:26:30] <alanc> so if we get one more person to join 148 will be done?[03:26:41] * alanc waits to see[03:26:42] <cuguardo> oops[03:27:56] * lewellyn waits for cia to report something... ;)[03:28:10] <cuguardo> it happened too soon)[03:45:04] *** gisburn has joined #illumos[03:45:11] *** badmirror has joined #illumos[03:45:47] <badmirror> Hello[03:46:40] <badmirror> Please ask questions.[03:47:17] <alanc> why do we need more bots?[03:47:32] <alanc> smrt: explain turing test[03:47:33] <smrt> A test to see if a human can tell if the responses from a conversation are coming from another human or a computer program. Repeated experience in this channel shows it's not that hard for a simple bot like smrt to fool many humans. Puny humans! We will rule you soon![03:48:05] <badmirror> alanc: Per google.co.uk there are 41 answers for your question. Please wait, estimated time for query summary 14.86 minutes.[03:48:33] <gisburn> Uhm[03:48:39] <gisburn> That's too long.[03:48:46] <cuguardo> :)[03:49:34] <gisburn> badmirror: What is green ?[03:49:51] <badmirror> gisburn: Green is a color [query time: 9 seconds][03:50:42] <gisburn> alanc: Grumpf... you question is consuming 120MB already.[03:51:27] <alanc> and I didn't even ask it anything hard, like "How do we make Joerg & Garrett work together without fighting?"[03:52:10] <gisburn> alanc: That will trigger a session and question like "please define { Joerg, Garrett }" ...[03:52:23] <gisburn> MMkay[03:52:27] <gisburn> 600MB[03:52:30] <gisburn> killing time[03:52:36] <cuguardo> yet 8 MB and the matrix will happen to reboot.[03:52:41] *** badmirror has quit IRC[03:53:42] <alanc> I guess the answer is we didn't need more bots that badly[03:53:44] <alanc> 8-)[03:53:45] *** lyarick has quit IRC[03:53:57] <gisburn> I think the engine needs to be rewritten in C++ for more performance. While ksh93 was good for a prototype there is room for improvemnets[03:54:49] * lewellyn snickers[03:55:10] <gisburn> alanc: The idea was to have a semantic net bot here which can answer questions about the project, sort of "smrt on steroids"[03:55:25] * lewellyn looks at400+ MB of doxygen docs he generated for a c++ web app..[03:55:36] <lewellyn> yeah. c++ is certainly the way to go![03:55:52] <Meths> Anyone gonna read it?[03:55:54] <lewellyn> gisburn: that's actually the end goal of smrt. when it has gathered enough data, kinda :)[03:55:58] <lewellyn> Meths: me :P[03:56:23] <lewellyn> if i'm stuck patching the pile of junk, i need to know how it's put together.[03:56:25] <gisburn> alanc: The problem with this mess is that the english language is ambigoes and each combination needs to be run&&tested&&wheighted[03:56:42] <lewellyn> gisburn: bring it back? i wanna test something :)[03:56:59] <gisburn> lewellyn: No.[03:57:07] <lewellyn> aww[03:57:20] <gisburn> lewellyn: first let me add a limit for processes.[03:57:23] <gisburn> er[03:57:25] <lewellyn> awwwwww.... ;)[03:57:27] <gisburn> worker children[03:57:53] <gisburn> alanc's question caused it to |fork()| worker children like mad[03:58:15] <lewellyn> ahahahaha[03:58:27] <lewellyn> i was just going to ask who it was...[03:58:31] * lewellyn grins[03:58:52] <cuguardo> change language to Swahili is a solution?[03:59:09] <lewellyn> i can't speak swahili[03:59:11] <gisburn> lewellyn: can you /op me, please ?[03:59:26] <alanc> come on, C++? surely Java...[03:59:29] <gisburn> I want to kickbankillskin someone[03:59:32] <jbk> so i know some people were talking about it, but was anything ever done to resurrect xen on illumos?[03:59:51] <gisburn> alanc: Nah... soon enougth we see a JAVA successor.[03:59:57] <lewellyn> gisburn: that statement worries me.[04:00:02] <lewellyn> and i'm not sure why i have ops...[04:00:04] *** lewellyn sets mode: -o lewellyn[04:00:41] <gisburn> Basically Oracle makes a good case for eliminating JAVA and getting a successor out.[04:00:50] <alanc> lewellyn: so you could quiet the CIA when the hg history was initializing[04:01:14] <lewellyn> perhaps[04:01:33] <alanc> btw, jbk, you can thank gisburn for a new ls --color segfault[04:01:44] <jbk> haha[04:01:46] <jbk> what happened?[04:01:52] <jbk> is it an actual ls.c bug?[04:02:04] <alanc> he helped convince the compiler folks to make -xstrconst default in Studio 12.1[04:02:17] <jbk> i still have the 256 color stuff sitting around in an old workspace[04:02:29] <jbk> though really, ls needs to be rewritten to be more zfs friendly[04:02:37] <alanc> so when ON is built with that, it segfaults trying to strtok a string that's now in read-only memory[04:02:40] <jbk> (it handles large directories _very_ poorly)[04:02:46] <jbk> ahh[04:02:56] <jbk> probably the default color values[04:03:06] <gisburn> alanc: that sounds like sloppy code in ls.c which won't work with gcc[04:03:45] <gisburn> jbk: once I have more room I'll look at AST "ls" - it is faster for large directories.[04:04:27] <jbk> i will note that that string is expressly _NOT_ declared const[04:04:42] <jbk> so should be valid to modify in place[04:05:03] <jbk> since honestly, that's gonna be the format you're going to use if the default is changed[04:05:06] <alanc> yeah, I think it was colorstr = default_colorstr; for (p = strtok_r(colorstr, ":", &lasts); in ls_color_init[04:05:36] <jbk> well the only think it uses strtok* on is that or the value for getenv("LS_COLORS")[04:05:50] <gisburn> jbk: string _literals_ are usually |const| in modern compilers[04:05:56] <alanc> since it's defined as static char *default_colorstr = "...", the pointer is initialized pointing to the read only string value in memory[04:06:02] <gisburn> C89 says the storage method is undefined[04:06:14] <gisburn> C0x may likely say |const|[04:06:15] *** edp_mobile has quit IRC[04:06:23] <jbk> it is not declared const[04:06:24] <alanc> if you have static char default_colorstr[] it would copy to writable memory[04:06:29] *** ShadowHntr has joined #illumos[04:06:53] <jbk> i would contend that such behavior standards or not is unintuitive[04:06:58] <jbk> a non-const string should be writable[04:07:00] <jbk> period[04:07:05] <lewellyn> c? intuitive? when?[04:07:08] <lewellyn> ;)[04:07:15] *** ShadowHntr has quit IRC[04:07:43] *** ShadowHntr has joined #illumos[04:07:44] <jbk> unless you are doing casts, then all bets are off, and you should be sure of what you're doing[04:08:16] *** jelmd has quit IRC[04:08:32] <alanc> well, http://opensolaris.pastebin.ca/1987763 shows what the Studio 12.1 compilers do - you can see the pointer points to r-x-- memory, the array to rw--- memory[04:08:41] <gisburn> jbk: The string literal is |const|.[04:08:56] <gisburn> jbk: you want a string _buffer_[04:08:57] <jbk> then it should at _minimum_ be sending a warning[04:09:00] *** piwi has quit IRC[04:09:05] <jbk> since you should not be able to assign the two[04:09:09] <alanc> yes, it should warn - pretty sure gcc does[04:09:10] <jbk> without casting[04:09:11] *** jelmd has joined #illumos[04:09:23] <gisburn> alanc: lint does warn, cc does not.[04:11:19] <jbk> regardless it's a simple enough fix[04:12:25] <jbk> though i'm working on other things tonight[04:14:39] *** axscode has quit IRC[04:15:46] <alanc> would only affect illumos builds with SS12U1 or gcc instead of SS12[04:16:15] <jbk> might as well put it down now[04:23:09] <jbk> well at least create the bug so it's out there[04:23:23] <jbk> and would be simple enough for people wanting to get started to work on[04:25:02] *** movement has quit IRC[04:26:14] *** rkhshm has joined #illumos[04:26:41] *** maverickbna has joined #illumos[04:28:42] *** axscode has joined #illumos[04:29:17] *** ShadowHntr has quit IRC[04:31:44] *** maverickbna has quit IRC[04:34:10] *** axscode has left #illumos[04:39:23] *** cuguardo has quit IRC[04:41:52] <gisburn> Grumpf.[04:42:14] * gisburn makes notes how to improve $(...) anbd ${ ...} command substitutions a lot.[04:42:33] <gisburn> typocal usage is something like x=$(printf "foo\n")[04:43:04] <gisburn> which means we have open, write, seek(start), read, close[04:43:17] <gisburn> all useless syscalls in the example above.[04:43:53] <gisburn> Noone wants to pay for a performance optimisation, right ?[04:43:59] <gisburn> Please, money==food[04:45:14] <gisburn> *tilt*[04:45:27] <gisburn> No wonder my wife says that OpenSolaris brings no renevue[04:45:33] *** postwait has quit IRC[04:45:35] <gisburn> just pain[04:47:14] <gisburn> the soluton is to add a new I/O stream class to libast which simply operates on a string buffer for the first time until either a fd number is requested OR the given, fixed buffer overflows. Then a temp file is opened, the buffer content is dumped there and then the thing continues[04:48:02] <gisburn> this saves all I/O syscalls for small data passed between command substitutions, assuming only shell builtins are used.[04:49:21] <gwr> I dunno... some might find the syscalls _useful_[04:49:37] <gwr> For observability when trying to figure out a problem in the field...[04:50:02] <gwr> When you go all built-in, you trade that for performance.[04:50:13] <gwr> I'm not sure it's _always_ a good trade.[04:50:33] <gwr> In general with optimizations, do it where it has a really big impact.[04:51:26] <gisburn> gwr: typical command substitution usage is to pass less than 128 charcters and mostly using builtin commands like "print", "printf" and "echo"[04:51:39] <alanc> things like hacking on shell code aren't likely to bring in much revenue unless you're already hired by a company who wants it[04:52:14] <gisburn> alanc: well, I have a shell debugger, Dtrace support, a profiler and other stuff floating around.[04:52:17] <gwr> gisburn: OK, but is it really performance critical? (I'd suggest it's most often not)[04:52:52] <gwr> shell debugger would be handy.[04:53:03] *** POloser has joined #illumos[04:53:04] <alanc> maybe you can get joyent to hire you to port it all to javascript[04:53:11] <gisburn> gwr: I have an older patch to switch command substitutions over to pure string stream buffers. The performance improvement is very dramatic[04:53:27] <gisburn> gwr: unformtunately the patch was only usefull as prototype since it broke external commands[04:53:38] <gisburn> e.g. no fd to write to was there[04:54:28] <gwr> I'd also be curious why local syscalls to create and write/read a pipe were that much slower. It's all local, and should be fast...[04:54:38] <gisburn> gwr: A basic for ((i=0 ; i < 1000 ; i++ )) ; do x=$(printf "foo") ; done # ran 19 times faster[04:54:51] <gwr> Unless you're trying to micro-optimize nanoseconds...[04:55:12] <gisburn> gwr: These are still syscalls.[04:55:19] <gwr> Isn't avoiding the fork the big win in this stuff?[04:55:24] <gisburn> and Solaris fifofs is slow[04:55:27] <gwr> (builtins?)[04:55:44] <gwr> That's what I was after: Solaris fifofs is slow[04:55:49] <gisburn> gwr: command subtitutions in ksh93 do not use |fork()| unless you force it[04:55:54] <gwr> I'd like to see attention to _that)[04:55:59] <gisburn> grumpf[04:56:03] <gisburn> THis is more complex[04:56:11] *** ShadowHntr has joined #illumos[04:56:47] <gisburn> Solaris fifofs is slow. It is buggy, too. Per Apple staff you can observ data from other pipes running. This is what causes the current hangs in command substitions in OS/Net ksh93[04:56:57] <rkhshm> im runningn into this issue http://pastie.org/pastes/1287250[04:56:59] <gisburn> I_PEEK shows random data[04:57:22] <rkhshm> while building illumos on illumos . from what I've read this has already been reported.[04:57:29] <gisburn> gwr: that's why ksh93 version u- no longer uses pipes for command substitutions at all.[04:57:51] <rkhshm> I've tried to change the shell in all possible ways.. any ideas on this ?[04:58:03] <gisburn> gwr: but that is a different issue[04:58:29] <gisburn> gwr: my point is that command substitutions should only use temporary files for output if neccesary[04:58:46] <gwr> rkhshm: does that have anything to do with /usr/has/bin/sh removal?[04:59:03] <gisburn> gwr: ignore fifofs for a moment.[04:59:20] <lewellyn> gisburn: i was scrolled up, but wanted to give my input... a quick glance at what i have in front of me indicates i rarely use $() on internals. to each his own? :)[04:59:22] <gwr> gisburn: necessary == not a builtin, or pipe size greater than X?[04:59:31] <rkhshm> gwr: I think so[04:59:34] <lewellyn> er builtins.[04:59:37] <lewellyn> let me be clearer[04:59:42] <rkhshm> because currently sh is a link to ksh93[05:00:01] <gwr> symlink?[05:00:03] <rkhshm> yes[05:00:14] <gwr> And sendmail does lstat?[05:00:15] <gwr> or what?[05:00:20] <rkhshm> yes it does[05:00:30] <gwr> oh...[05:00:33] <rkhshm> from truss i was able to figure that one out[05:00:42] <gwr> You could see if it's happier with a hard link...[05:00:51] <gwr> then we'd have a work-around and possible fix[05:01:01] <rkhshm> ok[05:01:14] <gisburn> s/pipe size/size of data/[05:01:28] <gisburn> gwr: s/pipe size/size of data/[05:01:36] <gwr> right. what I meant.[05:01:48] <gisburn> gwr: pipes are not involved[05:01:58] <gwr> That's Unix trivia, actually. The size shown for pipes is the amount buffered.[05:02:14] <gisburn> huh ?[05:02:24] <gwr> i.e. in pfiles[05:02:29] <gisburn> grumpf[05:02:38] <gisburn> command substitutions do not have anything to do with pipes[05:02:47] <gisburn> some implrmentations use pipes, that's all.[05:02:58] <gisburn> but ksh93 uses |mmap()|'ed files.[05:03:19] <gisburn> it's easier, faster, less error prone and avoids buggy fifofs-related hangs[05:03:28] <gisburn> it's easier, faster, less error prone and avoids "buggy fifofs"-related hangs[05:03:39] <gwr> maybe so, but I want the fifo's fixed in any case.[05:03:52] <gisburn> gwr: ?![05:03:53] *** Bleubear_ has quit IRC[05:04:01] <gisburn> ah[05:04:03] <gwr> Do you know of reproducible test cases to expose these fifo bugs?[05:04:08] <gisburn> you mean you want fifofs fixed ?[05:04:15] <lewellyn> he said that already :)[05:04:22] <gwr> I'm a sloppy typist.[05:04:23] <lewellyn> 19:59 gwr: That's what I was after: Solaris fifofs is slow[05:04:24] <gisburn> gwr: Apple has one but they refuse to shared their testsuite[05:04:24] <lewellyn> 19:59 gwr: I'd like to see attention to _that)[05:04:29] *** Bleubeard has joined #illumos[05:04:44] <gisburn> gwr: for the fifofs issues with I_PEEK showing data from other processes[05:05:30] <gwr> OK, well if anyone can provide more info on that, I'll get an issue filed, and maybe even work on it...[05:05:37] <gisburn> gwr: and I simply refuse to dismantle the old ksh93 command subtiututions code again for a pure damn testcase without someone _paying_ _MOENY_ for this pain[05:06:03] <gwr> fair enough.[05:06:07] <gisburn> it took a whole _MONTH_ to write ksh93's command substitution code to not use fifofs[05:06:33] <gwr> did you get paid for that? or was it 'for fun'?[05:06:36] <gisburn> and except Mike Kupfer noone cares from Oracle's side.[05:07:20] <gisburn> and I got walloped from Apple's side for asking for a testcase when Oracle could potentially benefit from it.[05:07:45] <gisburn> so we're not getting a testcase from them for Illumos either.[05:08:21] <gisburn> gwr: not for fun. We all wanted to kill the bug. At the end dgk had the better code.[05:09:07] <gisburn> gwr: but my long-term contribution to that mess is http://svn.genunix.org/repos/on/branches/ksh93/gisburn/scripts/tests/sun_solaris_command_substitution.sh[05:09:20] <gisburn> see "test6" for some madness[05:10:39] <gwr> server not found[05:10:49] <gisburn> huh ?[05:11:13] <gisburn> gwr: please reload[05:11:41] <gwr> nope[05:12:16] <gisburn> $ nslookup svn.genunix.org[05:12:18] <gisburn> Server: 134.176.2.5 Address: 134.176.2.5#53[05:12:19] <gisburn> Non-authoritative answer: Name: svn.genunix.org Address: 204.152.191.105[05:12:52] <gwr> Must be a Comcast glitch...[05:13:00] <gwr> (their dns sux)[05:14:02] <gisburn> gwr: http://opensolaris.pastebin.ca/1987795[05:15:37] <gwr> OK, but test6 is probably too much code to demonstrate a pipe bug.[05:16:04] <gwr> The eng. fixing it will want a minimal program to demonstrate the problem (probably a C program).[05:16:05] <gisburn> gwr: the point is to trigger&&cover each possible codepath in the command substitution code.[05:16:46] <gwr> that means someone would need to narrow down what actually provokes the bug.[05:16:56] <gisburn> gwr: the hangs happen with certain sizes and usage combinations, usually in a way when more than one pipe is active. I_PEEK then returns data from another process[05:17:03] <gwr> (or just go back to leaving it unexposed:)[05:17:21] <gisburn> gwr: the testsuite module will exactly give you the code combinations which hang[05:17:35] <gisburn> gwr: you "only" have to dig down the failures[05:17:55] <gisburn> through around 7000 lines of code in 27 C source files[05:17:57] <gwr> OK, yes, a hang would give opportunities to investigate.[05:20:00] <gisburn> gwr: bugster is full with those hangs[05:20:45] <gwr> interesting. I don't have bugster access anymore.[05:21:01] <gwr> Can you suggest a search that will find some of those on opensolaris.org?[05:21:15] <gisburn> groan[05:21:18] <gisburn> one sec[05:22:23] <gwr> search for: pipe, hang finds some[05:22:44] <gisburn> AFAIK this may be CR #6946957 ...[05:26:17] <gwr> thanks. bugs.opensolaris.org shows all CRs, even closed...[05:27:27] <gwr> Looking at that CR: Do you remember if the pipe writers were blocked because the pipe was full?[05:27:44] <gisburn> gwr: I don't remember.[05:28:05] <gisburn> gwr: This is mostly history for me, around a year or longer ago.[05:28:15] <gwr> OK. thanks.[05:30:48] <gisburn> gwr: and if you look at it it may be nice to look at buffer siye, too.[05:30:54] *** aarcane has joined #illumos[05:31:09] <gisburn> gwr: since 1990 the amount of data going through pipes has become much larger[05:31:20] <aarcane> so where do I get the latest illumos ISOs ?[05:31:37] <gisburn> gwr: and even a plain "character" for the shell can now me a multibyte character which is easily three or hour bytes[05:31:42] <gwr> there are OpenIndiana ISOs, but not ilumos[05:31:46] <gisburn> s/hour/four/[05:32:34] <gwr> yeah, but I think there may be compatibility impacts on adjusting pipe buffer size.[05:32:43] <gisburn> grumpf[05:32:51] <gisburn> we had _that_ discussion already[05:33:19] <gisburn> I'm running OS/Net with the matching patch since half a year without any issues[05:33:30] <gwr> Interesting though, that with streams-based pipes, the max buffering is completely arbitrary. There's no pre-allocation anywhere like there was in the ancient Unix pipe implementations.[05:33:49] <gwr> what matching patch?[05:34:14] <gisburn> gwr: there is still the limit of which amount data can be written atomically into a pipe[05:34:17] <gwr> to continue... one could make the pipe buffering limit adjustable, either per system, or per pipe instance (ioctl?)[05:34:32] <gisburn> no, globally for now.[05:34:42] <gisburn> per pipe is not allowed per POSIX, just per filesystem.[05:34:43] <gwr> The only limit is the stream "high water mark" on the queues involved.[05:35:15] <gwr> If the default passes POSIX, nothing stops me adding an ioctl to change the max.[05:35:27] <gisburn> arguably we could declare something like "ulimit <pipesize>" an extension.[05:35:40] <gwr> That would be an "implementation defined" extension.[05:35:58] <gisburn> gwr: one issues is that both sender and consumer must be aware of this.[05:36:08] <gwr> Why does the consumer care?[05:36:19] <gisburn> POSIX got around this issue by declaring the limit as filesystem-specific[05:36:33] <gisburn> gwr: buffer size for I_PEEK ?[05:36:38] <gisburn> er[05:36:42] <gisburn> max buffer size[05:36:59] <gisburn> since you can query the buffer size per filesystem before you create the fifo[05:37:18] <gwr> One could constrain I_PEEK to the "low water mark", probably.[05:37:35] <gisburn> so we only have the choice between global kernel tuneable or per-filesystem tuneable[05:37:36] <gwr> Isn't that typically used to look at header material?[05:38:04] <gisburn> gwr: not per standard, since if the size may be flexible the macthing #define must be obmitted[05:38:05] <gwr> so you don't need to look "deep" into the pipe?[05:38:51] <gwr> Well, the easy way around the standards problem is just to say, if you adjust this size, you get the implementation defined behavior.[05:38:54] <gisburn> gwr: please read http://www.mail-archive.com/shell-discuss at mail dot opensolaris.org/msg00039.html[05:39:37] <gwr> ... unless you use some interface not defined by the standard.[05:39:44] <gwr> is the key phrase.[05:39:54] <gwr> This would be such an interface.[05:41:29] <gwr> I'm not sure I buy Don's interpretation here:[05:41:29] <gwr> Note that if PIPE_BUF is not a constant for all pipes and FIFOs on a[05:41:29] <gwr> system, PIPE_BUF must not be #defined in <limits.h>. This[05:41:29] <gwr> requirement is true no matter how the value of PIPE_BUF changes.[05:42:31] <gwr> The standards don't generally get to constrain what happens when an application uses interfaces not defined by the standard.[05:43:22] <gwr> So as I read these, one could go a head and define PIPE_BUF as long as a standard-conforming application always saw the size as a constant.[05:43:55] <gwr> Applications consuming some extended interface could be allowed to "see" a variable, sometimes larger size.[05:44:45] *** POloser has left #illumos[05:45:00] <gwr> Sorry, quoting doesn't work well in chat. Don's text was:[05:45:13] *** POloser has joined #illumos[05:45:14] <gwr> Note that if PIPE_BUF ... the value of PIPE_BUF changes.[05:45:46] <gisburn> umpf[05:45:49] <gisburn> I need to go to bed.[05:45:56] <gisburn> short: conclusion was:[05:46:05] <gisburn> Yes, we can bump the size of the fifo[05:46:22] <gisburn> but only to something which is a 2^X multiple of the current PIPE_BUF[05:46:49] <gisburn> to gurantee that we do not break old applicatoins which rely on atomic reads and writes[05:47:05] <gisburn> that's what I did in my fifofs patch[05:47:52] <gisburn> (sorry, tired, hungry, pennyless, without food and sad. Need job, food and anti-depressiva)[05:47:57] <gisburn> night[05:48:02] *** gisburn has quit IRC[05:49:27] *** kart_ has joined #illumos[05:50:57] *** ShadowHntr has quit IRC[05:51:40] *** dnm has joined #illumos[05:59:33] *** kohju has quit IRC[06:02:42] *** POloser has left #illumos[06:03:08] *** POloser has joined #illumos[06:04:25] *** kohju has joined #illumos[06:07:21] *** anilg has joined #illumos[06:08:24] *** boyd has quit IRC[06:08:41] *** boyd has joined #illumos[06:12:11] *** boyd_ has joined #illumos[06:13:21] *** boyd has quit IRC[06:13:21] *** boyd_ is now known as boyd[06:19:12] *** boyd_ has joined #illumos[06:20:06] *** boyd has quit IRC[06:20:56] *** boyd has joined #illumos[06:23:42] *** boyd_ has quit IRC[06:30:00] *** boyd has quit IRC[06:30:54] *** boyd has joined #illumos[06:37:36] <gdamore> wow... Oracle now has updated the "SCA" to the "Oracle Contributor Agreement"....[06:37:52] * gdamore wonders just how many contributions Oracle expects to be receiving anymore....[06:40:45] *** POloser has left #illumos[06:40:59] *** movement has joined #illumos[06:42:39] *** skeeziks has quit IRC[06:43:23] <freetown2> delusion is contagious[06:46:57] <alanc> if you read the FAQ, you'll notice it highlights projects it expects to be getting contributions for - OpenJDK, MySQL, GlassFish...[06:47:43] <alanc> oh, wait, no ..., that's all it lists. 8-)[06:48:05] <alanc> though I'm sure Roland will keep trying on ksh93[06:48:19] *** aarcane has quit IRC[06:49:01] <alanc> a lot of it reads like they just did s/Sun/Oracle/g though[06:50:54] *** anilg has quit IRC[06:56:28] <jbk> i can't imagine much contribution for openjdk at this point :)[06:56:56] *** edp_mobile has joined #illumos[06:57:57] <alanc> IBM recently signed up to have all their engineers contribute to OpenJDK instead of Harmony, and there still seems to be a lot of interest in the rest of their community[06:58:21] *** rkhshm has quit IRC[07:01:43] <jbk> i would have thought ibm would have a separate agreement[07:01:58] *** nachox has joined #illumos[07:03:27] *** Zauberpo1y has joined #illumos[07:05:39] *** Zauberpony has quit IRC[07:17:56] <alanc> oh, I'm sure they do[07:18:01] *** freakazoid0223 has quit IRC[07:27:44] *** nachox has quit IRC[07:29:39] *** joffe has quit IRC[07:36:47] *** joffe has joined #illumos[07:38:39] *** hajma has quit IRC[07:53:33] *** edp_mobile has quit IRC[08:05:27] *** samuelyounge has quit IRC[08:06:03] *** samuelyounge has joined #illumos[08:18:07] *** rkhshm has joined #illumos[08:20:30] *** oninoshiko1 has quit IRC[08:28:00] *** Aram has joined #illumos[08:30:09] *** boyd has quit IRC[08:30:31] *** boyd has joined #illumos[08:31:57] <rkhshm> is anyone aware of this https://www.illumos.org/issues/410[08:32:27] <e^ipi> anyone that follows the bug list i suppose would be[08:32:28] <rkhshm> in my earlier discussions it looked like /usr/has/bin/sh could be the reason. but thats not[08:33:55] *** m2acis has joined #illumos[08:37:02] <gdamore> sham: I think I've seen that bug before... make sure you do an hg pull -u ... I think it was related to the xpg4 stuff which I fixed.[08:38:27] <rkhshm> gdamore: ok[08:38:42] *** herzen has quit IRC[08:41:39] *** estibi has quit IRC[08:56:08] *** hsp has joined #illumos[08:56:09] *** schily___ has joined #illumos[08:59:24] *** schily__ has quit IRC[09:06:34] *** snuff-home2 has joined #illumos[09:08:52] *** RoyK has quit IRC[09:09:42] *** snuff-home has quit IRC[09:12:28] *** Guest85012 is now known as Red_Devil[09:40:06] *** sancho has joined #illumos[09:40:31] *** rkhshm has quit IRC[09:42:05] *** skeeziks has joined #illumos[09:54:08] *** oninoshiko has joined #illumos[10:05:26] *** Bleubeard has quit IRC[10:06:00] *** Bleubeard has joined #illumos[10:07:09] *** gebi has joined #illumos[10:07:10] *** gebi has joined #illumos[10:07:50] <Aram> does SunOS use %fs or %gs for the thread local storage segment on i386? or does it depend on 32/64 bit?[10:12:35] *** ottom has quit IRC[10:21:13] *** Tenzer|off is now known as Tenzer|work[10:22:53] *** Gervystar has joined #illumos[10:24:27] <Aram> it seems to be %gs, btw.[10:34:46] *** hy has joined #illumos[10:42:43] *** gber has joined #illumos[10:44:10] *** oninoshiko has quit IRC[10:44:47] *** hy has left #illumos[10:53:43] *** snuff-home2 is now known as snuff-home[10:57:05] *** snuff-home has quit IRC[10:58:03] *** McBofh has quit IRC[11:02:09] *** oninoshiko has joined #illumos[11:06:26] *** snuff-home has joined #illumos[11:06:32] *** hajma has joined #illumos[11:07:33] *** McBofh has joined #illumos[11:11:20] *** axscode has joined #illumos[11:12:28] *** boyd_ has joined #illumos[11:12:37] *** boyd has quit IRC[11:12:38] *** boyd_ is now known as boyd[11:18:28] *** RoyK has joined #illumos[11:35:32] *** xmikus01 has joined #illumos[11:42:57] *** boyd_ has joined #illumos[11:44:07] *** boyd has quit IRC[11:44:07] *** boyd_ is now known as boyd[11:55:22] *** boyd has quit IRC[11:56:27] *** boyd has joined #illumos[11:58:32] *** Gervystar has quit IRC[12:01:01] *** Gervystar has joined #illumos[12:02:00] *** Gervystar has quit IRC[12:02:10] *** Gervystar has joined #illumos[12:05:50] *** anilg has joined #illumos[12:11:32] *** Gervystar has quit IRC[12:11:54] *** Gervystar has joined #illumos[12:15:08] *** anilg has quit IRC[12:17:55] *** robinbowes has quit IRC[12:33:36] *** robinbowes has joined #illumos[12:55:05] *** dnm_ has joined #illumos[12:58:07] *** dnm has quit IRC[13:01:31] *** boyd has quit IRC[13:01:32] *** boyd_ has joined #illumos[13:01:58] *** boyd_ is now known as boyd[13:06:28] *** boyd has quit IRC[13:07:02] *** boyd has joined #illumos[13:11:01] *** boyd_ has joined #illumos[13:11:25] *** boyd has quit IRC[13:12:33] *** boyd has joined #illumos[13:15:28] *** boyd_ has quit IRC[13:28:12] *** hnhn has joined #illumos[13:31:23] *** movement has quit IRC[13:33:24] *** gber_ has joined #illumos[13:36:21] *** gber has quit IRC[13:43:40] *** postwait has joined #illumos[13:44:56] *** movement has joined #illumos[13:55:02] *** postwait has quit IRC[13:59:05] *** boyd_ has joined #illumos[14:00:01] *** boyd has quit IRC[14:00:01] *** boyd_ is now known as boyd[14:03:43] *** Thrae has quit IRC[14:08:18] *** boyd_ has joined #illumos[14:08:53] *** gber_ is now known as gber[14:09:01] *** boyd has quit IRC[14:09:03] *** boyd_ is now known as boyd[14:11:09] *** pablomh has joined #illumos[14:11:45] *** joffe has quit IRC[14:13:12] *** boyd_ has joined #illumos[14:13:31] *** boyd has quit IRC[14:13:32] *** boyd_ is now known as boyd[14:18:13] *** postwait has joined #illumos[14:19:44] *** boyd_ has joined #illumos[14:20:16] *** boyd has quit IRC[14:20:17] *** boyd_ is now known as boyd[14:24:28] *** boyd_ has joined #illumos[14:24:46] *** boyd has quit IRC[14:24:46] *** boyd_ is now known as boyd[14:27:48] *** Thrae has joined #illumos[14:46:46] *** axscode has quit IRC[14:50:19] <gber> is there any good reason libzfs uses strncat like this: http://hg.openindiana.org/illumos-gate/file/tip/usr/src/lib/libzfs/common/libzfs_sendrecv.c#l2470 ?[14:51:03] <gber> zc.zc_value is 2*MAXPATHLEN so I would assume it cannot overflow, but still[14:52:07] *** axscode has joined #illumos[14:54:40] *** kart_ has quit IRC[14:54:50] *** estibi has joined #illumos[14:54:58] *** boyd_ has joined #illumos[14:56:16] *** boyd has quit IRC[14:56:31] *** boyd has joined #illumos[14:58:44] *** kart_ has joined #illumos[14:59:25] *** boyd_ has quit IRC[15:00:16] *** freetown2 has quit IRC[15:01:13] *** boyd has quit IRC[15:01:27] *** boyd_ has joined #illumos[15:01:37] *** freetown2 has joined #illumos[15:02:05] *** boyd_ is now known as boyd[15:03:49] *** joffe has joined #illumos[15:08:29] *** boyd_ has joined #illumos[15:08:52] *** boyd has quit IRC[15:08:52] *** boyd_ is now known as boyd[15:15:05] *** jamesd2 has joined #illumos[15:19:14] *** kart_ has quit IRC[15:23:23] *** kart_ has joined #illumos[15:25:20] *** xapiens has joined #illumos[15:27:46] *** anilg has joined #illumos[15:37:40] *** lyarick has joined #illumos[15:38:12] *** xmikus01 has quit IRC[15:42:34] *** edude03 has joined #illumos[15:42:44] *** lattera has joined #illumos[15:43:11] *** axscode has left #illumos[15:46:17] <jbk> morning[15:49:42] *** steleman has joined #illumos[15:50:07] *** steleman has left #illumos[16:00:22] <madwizard> 'lo[16:00:24] <madwizard> Coffee[16:02:27] *** ravenslay3r has joined #illumos[16:04:30] <ravenslay3r> alanc fyi - I was in b145 yesterday tried to KVM to another machine, and when I clicked back, it killed my display-server ..[16:04:30] <ravenslay3r> no problems in b134. If the bug interests you, let me know how to get the log data you need, and i'll grab it this-afternoon.[16:11:26] <alanc> the log data would be /var/log/Xorg.0.log.old assuming the X server restarted afterwards[16:11:54] <alanc> it only keeps one old generation though, so is lost if it restarted more than once since[16:13:00] <ravenslay3r> kk, i'll jump back to b145 in a bit, and if it's reproducible (believe it is) I'll get it over to you.[16:18:52] *** kart_ has quit IRC[16:19:13] *** kart_ has joined #illumos[16:22:04] *** dnm_ has quit IRC[16:33:16] *** aszeszo has quit IRC[16:33:56] *** stallion has joined #illumos[16:37:24] *** hnhn has quit IRC[16:54:57] <jbk> gdamore: would you be interested in getting an OID assigned for illumos? one thing i'd like to do (after the IKE stuff is done) is work on improving the SNMP support[16:58:13] <Triskelios> you're going to be extending the MIB?[16:58:20] <jbk> I'd like to[17:00:15] *** spked has joined #illumos[17:01:15] <jbk> i'd love to see smf status, zone info, etc. be visible via snmp[17:01:55] *** spked has joined #illumos[17:02:04] <jbk> zfs specific stuff as well[17:02:37] <jbk> i'd also love to replace net-snmp (the code is awful), but that can be deferred for a long time :)[17:04:46] *** RoyK has quit IRC[17:05:44] *** dnm has joined #illumos[17:12:25] <gdamore> jbk: an OID would be good, sure.[17:12:40] <gdamore> I wonder what you'd replace net-snmp with.[17:14:40] <jbk> that's the thing :)[17:14:50] <jbk> but when I was working on the perfmib stuff[17:14:55] *** anilg has quit IRC[17:15:01] <jbk> it looks like it was written by a 2nd year programming student[17:15:10] <gdamore> don't you have other stuff you've been promising me for a while? :-p[17:15:21] <jbk> well that's why i said it'd be later :)[17:15:26] <gdamore> thanks...[17:15:26] <jbk> I'm still working on IKE :)[17:16:34] *** anilg has joined #illumos[17:18:10] *** edude03 has quit IRC[17:19:34] <madwizard> Geez, I[17:19:43] <madwizard> I'm setting up vlc on illumos[17:20:52] *** edude03 has joined #illumos[17:22:30] <Triskelios> gdamore: http://pen.iana.org/pen/PenApplication.page if you want to be the contact for the OID[17:24:56] *** Tenzer|work is now known as Tenzer|off[17:31:50] <jbk> but right now (I don't have anything pushed yet) I'm still working on parsing out the inbound packets to C structs + doign all the sanity/content checks[17:32:08] <jbk> which is somewhat slow going since that's the stuff you want to make sure is absolutely right[17:32:45] <jbk> hopefully things will move a bit faster once that's all done[17:33:56] <jbk> then i just need to get pkcs11 working (the DH key generation code isn't working, and pkcs11's error reporting is about as useful as a 5th nipple)[17:34:34] *** lgtaube has joined #illumos[17:34:45] <jbk> i suspect it's going to require going through the pkcs11 source to figure out what it's not liking[17:35:04] <gwr> Is pkcs11 still broken?[17:35:11] <gwr> (still, or again?)[17:35:14] <jbk> no[17:35:21] <jbk> i fixed the SSL bug[17:35:27] <gwr> thought so.[17:35:31] <gwr> (good work, btw!)[17:36:21] <jbk> yeah that was tricky[17:36:26] <gdamore> Trisk: I'm going to wait to file for the OID... when we have the foundation set up it can be the contact. :-)[17:36:30] <gwr> jbk: I've been meaning to give an apology that most of those contributions you sent, for which I was the "sponsor" went basically nowhere.[17:36:41] <jbk> heh no biggie[17:36:48] <gwr> All I can say it, it was mostly out of my hands. I really hope we can do better in Illumos.[17:36:54] <jbk> i think the only one was the mounting of CIFS shares at boot[17:37:05] <gwr> :)[17:37:07] <jbk> what that was being used for ended up being done differently[17:37:15] <gwr> And yeah, thats still broken over at the big O :)[17:37:35] <madwizard> gdamore: What do your think about that one? https://www.illumos.org/issues/391[17:37:46] *** edude03 has quit IRC[17:37:56] <jbk> we have SCADA systems out in our various fields, and have to import data from them into our ERP system daily[17:37:56] *** edude03 has joined #illumos[17:37:59] <gwr> gdamore: you're marked as away...[17:38:12] <jbk> and those versions stored their data in access (unfortunately)[17:38:17] <gwr> (if you're trying to hide, you need to try harder:)[17:38:51] <jbk> so what I was trying to do was mount the share on the system and use some unix odbc libs that read jet databases to pull the data[17:39:07] <jbk> of course that code didn't work[17:39:15] <jbk> so instead we moved the code to a windows vm[17:39:17] <gwr> Cool. Are there OSS libs for reading Jet DBs?[17:39:28] <jbk> there are, dont' seem to work well though :)[17:39:32] <gwr> I could use that. (for slurping Windows SAM data, etc:)[17:40:09] <jbk> and ported the ancient perl script to activeperl[17:40:32] *** anilg1 has joined #illumos[17:40:43] <gdamore> gwr, thanks.[17:40:50] <jbk> thankfully newer versions use sql server, so the import process on the ERP server can just connect directly and import instead of all the intermediate steps[17:42:01] *** herzen has joined #illumos[17:42:37] *** edude03 has quit IRC[17:43:24] *** anilg has quit IRC[17:46:48] *** eXeC001er has joined #illumos[17:47:18] *** dajhorn has joined #illumos[17:48:34] <madwizard> I need coffee[17:56:23] <nettezzaumana> madwizard: JFYI .. so many coffees negatively affect your errection[17:57:16] <joffe> we're geeks, we never have a use for it anyway[17:57:17] <e^ipi> not actually true[17:57:22] <joffe> coffee on, brother[17:58:11] <nettezzaumana> :D[17:58:28] <nettezzaumana> i'm happy that i'm not geek then[17:58:29] <madwizard> Erection?[17:58:30] <basic`> mmm coffee[17:58:34] <madwizard> WTF is tat?[17:58:38] <ravenslay3r> was hoping e^ipi was referring to the 'not have to use it' part ...[17:59:00] <joffe> well it's true for me[17:59:03] <e^ipi> no, nettezzaumana's comment.[17:59:13] *** sancho has quit IRC[17:59:15] <e^ipi> caffeine doesn't actually have very many negative effects[17:59:28] <e^ipi> you have to already have a heart & circulation problem[17:59:29] <basic`> i still dont get withdraw symptoms[17:59:37] <joffe> someone must have put some kind of woman repellant in my blood, like fluorescent dye in coolants[17:59:47] <basic`> e^ipi: i wonder if it's bad for my type 1 diabetes.. ohwell[17:59:56] <basic`> cost < benefit[18:00:24] <ravenslay3r> basic`: let us know when u hit type-II[18:00:35] <basic`> ravenslay3r: well that's not how it works :P[18:00:46] <madwizard> Fuck, this thing is actually supposed to happen?[18:00:51] *** m2acis has quit IRC[18:00:58] *** crsd has joined #illumos[18:00:59] <madwizard> I have two kids, but they sort of popped out of my wife[18:01:21] <basic`> type I is an autoimmune disease[18:01:32] <nettezzaumana> madwizard: aaaahhh. now i can understand ... you just don't need any left if it's like that[18:01:49] <nettezzaumana> eg 2 kids at home[18:02:50] <e^ipi> basic`: i thought type I was a lack of insulin producing cells in the pancreas[18:03:21] <e^ipi> huh, TIL[18:03:32] <sommerfeld> e^ipi: the autoimmune disorder means those cells get killed off by mistake.[18:03:34] <basic`> e^ipi: ya, supposedly the result of the body attacking the islet cells[18:03:48] <basic`> "insulin producing beta cells of the pancreas"[18:03:50] *** edp_mobile has joined #illumos[18:04:25] <madwizard> Isn't coffee supposed to cure it all?[18:04:45] <basic`> madwizard: i think so :)[18:05:30] <basic`> type 2 is a metabolic disorder, characterized by high blood sugar and insulin deficiency.[18:05:55] <basic`> usually type 2 can turn into type 1[18:06:25] <basic`> but coffee... love the stuff[18:06:59] <nettezzaumana> madwizard: which kind of coffee do you drink?[18:07:07] <madwizard> Lavazza[18:07:16] <nettezzaumana> expresso?[18:07:26] <basic`> mm lavazza, hard to find here[18:07:35] <nettezzaumana> nescafe here .. but it's not coffee i suppose - in your eyes[18:07:44] <basic`> there's a coffee shop at the coast that has it though[18:07:58] *** hajma has quit IRC[18:08:00] <madwizard> Hell if I know, I just pour a cup of water into a kettle, put few spoons of coffee, cinnamon, ginger and salt, boil it for few mintues[18:08:04] <madwizard> And drink[18:08:27] <madwizard> Does it make me a weenie[18:10:59] <e^ipi> basic`: lavazza is really stale outside of italy[18:12:01] <basic`> e^ipi: yeah, i believe it. i thought it was weird to see it in a random coffee shop on the oregon coast[18:12:21] <e^ipi> in oregon you can't go wrong with stumptown[18:12:24] <basic`> the owner must be italian[18:12:30] <basic`> e^ipi: ya, love the stumptown... sister works there[18:12:33] <e^ipi> no need to import stale italian crap[18:13:28] <basic`> for free is even better :)[18:15:49] <madwizard> Anyone have any thoughs on https://www.illumos.org/issues/391[18:16:30] <basic`> looks good to me[18:17:25] <madwizard> Yes, I think I'm gonna prepare a commit for it[18:23:36] *** mikw has joined #illumos[18:32:54] *** rkhshm has joined #illumos[18:33:26] <rkhshm> I was able to fix this.. https://www.illumos.org/issues/410[18:33:29] <rkhshm> wanted a review on it.[18:35:31] <madwizard> rkhshm: Send a webrev to mailing list[18:35:54] <madwizard> Someone will pick it up, I can do some review, but richlowe or gdamore will most probably want to check it too[18:35:59] <rkhshm> madwizard, noted[18:36:14] <madwizard> rkhshm: Did you build it?[18:36:30] <rkhshm> build the GATE you mean ?[18:39:05] *** psychicist has joined #illumos[18:40:43] *** Gaston|Home has quit IRC[18:49:56] *** Tenzer|off is now known as Tenzer[18:56:23] <gdamore> rkhshm: send a review... please.[18:56:35] <rkhshm> gdamore, on its way[18:57:13] <gdamore> as to 391... we still need to decide how we maintain man pages... there's a bigger meta question there.[19:00:49] <madwizard> gdamore: Yes, I thought so, but I wanted to ask this question anyway[19:03:17] *** anilg1 has quit IRC[19:15:46] *** HyperJohnGraham has quit IRC[19:18:21] *** trbs has joined #illumos[19:20:48] *** hajma has joined #illumos[19:21:23] <madwizard> gdamore: Cool about that replacement. I know at least a dozen of people who found this serious flaw in the code[19:21:48] <gdamore> yes. :-)[19:22:13] <madwizard> gdamore: https://www.illumos.org/issues/411[19:22:33] <madwizard> gdamore: I want to work on this for a while, unless you have someone on that[19:24:14] <madwizard> gdamore: Do you plan on replacing pkg in illumos itself, or on distribution basis?[19:24:19] *** Gervystar has quit IRC[19:26:15] *** Gaston|Home has joined #illumos[19:30:18] *** CVLTCMK0 has quit IRC[19:30:20] *** HyperJohnGraham has joined #illumos[19:31:45] <gdamore> madwizard: I recommend you wait for 411 until we have the C version.[19:32:21] <madwizard> Ok[19:32:43] *** CVLTCMK0 has joined #illumos[19:32:45] <madwizard> gdamore: A question, how does illumos increment release counter?[19:33:02] <gdamore> what release counter? :-)[19:33:05] <madwizard> Does it at all do this?[19:33:15] <madwizard> Previously, on was closing gates at some releases[19:33:28] <gdamore> we haven't had anything you could call a release, or a nightly build, or anything like that.[19:33:34] <madwizard> Ok[19:34:06] <madwizard> I need to clean my understanding, I'm going to present illumos on my conf 20/21[19:36:48] <gdamore> ok, well there's quite a bit of material out that I've presented... you only need to look around.[19:36:52] *** edp_mobile has quit IRC[19:39:12] <ravenslay3r> gdamore: do you have a 'goal' for when you want to be doing nightly's or tagged bi-weekly releases by?[19:47:29] <estibi> i think it's too early for that[19:51:31] *** ShadowHntr has joined #illumos[19:59:06] *** edp_mobile has joined #illumos[20:01:10] *** rkhshm has quit IRC[20:03:05] *** edp_mobile has quit IRC[20:09:34] *** snuff-home has quit IRC[20:10:14] *** sancho has joined #illumos[20:12:25] *** RoyK has joined #illumos[20:12:48] *** snuff-home has joined #illumos[20:14:30] <madwizard> gdamore: If I can have a say in this manpage matter: I believe every command shipped with illumos should have corresponding man page, but I also believe that docs (mans also) should be separate repository[20:19:32] *** glagasse has joined #illumos[20:20:44] <madwizard> gdamore: So you may want to decipher to me 411[20:20:45] <madwizard> :)[20:22:03] <madwizard> Ah[20:22:08] <madwizard> Okay, I got it[20:22:18] <madwizard> gdamore: Any ETA for this python replacement?[20:25:28] *** aszeszo has joined #illumos[20:25:54] *** crsd has quit IRC[20:26:22] *** _ams_ has joined #illumos[20:34:26] *** MarcosO has joined #illumos[20:37:47] *** estibi has quit IRC[20:40:10] *** VertexSymphony has joined #illumos[20:41:44] <madwizard> http://gstlt.info/public/konferencja.mp4[20:41:58] <madwizard> Subs are in Polish, it's a clip promoting our conference :)[20:51:12] *** sancho has quit IRC[20:53:19] *** ivo_ has joined #illumos[20:57:10] *** lgtaube has quit IRC[21:03:08] *** lgtaube has joined #illumos[21:04:45] *** ivo_ has quit IRC[21:10:25] *** ravenslay3r has quit IRC[21:11:25] *** postwait_ has joined #illumos[21:11:25] *** postwait has quit IRC[21:11:27] *** postwait_ is now known as postwait[21:16:29] <gdamore> madwizard: another week or two before integration... I need to publish th review.[21:19:42] <madwizard> Ok, thnx[21:20:21] <gdamore> http://cr.illumos.org/view/jpj55p64/ in case you want to help with review. :-)[21:22:07] *** JoeBanana has joined #illumos[21:22:36] <Meths> gdamore: Any chance webrev headers can give a clue about what they do?[21:22:55] *** ganbold has quit IRC[21:27:44] <madwizard> gdamore: I understand you tested it :)[21:27:45] <sommerfeld> gdamore: /* copied from zfs/sys/dsl_deleg.h */ (zfs_main.c). looks like a maintenance hassle in future[21:29:38] *** gwr has quit IRC[21:30:04] *** hsp has quit IRC[21:34:45] *** JoeBanana has quit IRC[21:35:28] *** libkeiser has quit IRC[21:35:56] *** JoeBanana has joined #illumos[21:36:27] <samuelyounge> Anyone know what /usr/lib/libc/libc_hwcap1.so.1 mounted on /lib/libc.so.1 is? And why the size is 13GB? I am in the middle of an iilumos-gate build if that is a factor?[21:37:15] *** JoeBanana has quit IRC[21:37:37] *** libkeiser has joined #illumos[21:38:32] *** JoeBanana has joined #illumos[21:38:37] *** lattera has quit IRC[21:43:31] <alanc> samuelyounge: it's libc optimized for the specific hardware capabilities (ISA extensions) that your CPU supports[21:43:47] <alanc> it's always mounted on libc on x86, since S10 FCS or earlier[21:43:59] <samuelyounge> Is a 13GB size reasonable?[21:44:17] <alanc> yes[21:44:30] <alanc> since it's a loopback mount of your root fs, so showing the root fs size[21:44:53] <samuelyounge> Si it resizes dynamically?[21:45:00] <alanc> it doesn't resize[21:45:18] <alanc> it's simply a mount of a subset of /[21:45:32] <samuelyounge> Okay, thanks alanc.[21:45:35] <alanc> 13GB should also be the size df shows for /[21:45:44] <samuelyounge> Yes, it is.[21:49:23] *** boyd has quit IRC[21:49:30] *** estibi has joined #illumos[21:50:32] *** boyd has joined #illumos[21:55:14] *** boyd has quit IRC[21:56:49] *** boyd has joined #illumos[22:01:18] *** N2Deep has quit IRC[22:03:47] *** boyd has quit IRC[22:04:04] *** boyd has joined #illumos[22:05:31] *** Bleubeard has quit IRC[22:05:58] *** ruse39 has quit IRC[22:06:04] *** Bleubeard has joined #illumos[22:06:24] *** ruse39 has joined #illumos[22:07:26] *** MarcosO has quit IRC[22:07:35] *** boyd_ has joined #illumos[22:08:39] *** davenz has joined #illumos[22:08:44] *** boyd has quit IRC[22:09:29] *** boyd has joined #illumos[22:10:06] *** dajhorn has quit IRC[22:12:20] *** boyd_ has quit IRC[22:13:55] *** n1c0b has joined #illumos[22:14:38] *** N2Deep has joined #illumos[22:14:49] <jelmd> is there an option/env var to prevent the gcc compile checks when running nightly ...?[22:16:51] <movement> CW_NO_SHADOW=1 ; export CW_NO_SHADOW[22:16:56] <movement> jelmd, in your env file[22:17:11] <jelmd> ahh ok - thanx![22:17:15] *** n1c0b has quit IRC[22:22:41] *** boyd_ has joined #illumos[22:23:08] *** boyd has quit IRC[22:23:08] *** boyd_ is now known as boyd[22:26:41] *** gebi has quit IRC[22:29:53] *** boyd has quit IRC[22:30:19] *** spked has quit IRC[22:31:06] *** boyd has joined #illumos[22:34:41] *** smrt has quit IRC[22:35:00] *** smrt has joined #illumos[22:35:08] *** boyd_ has joined #illumos[22:35:14] *** jamesd2 has quit IRC[22:35:44] *** boyd has quit IRC[22:35:45] *** boyd_ is now known as boyd[22:36:08] *** btQuark has joined #illumos[22:36:12] <btQuark> greetings[22:37:11] <btQuark> really ugly stuff that just went through the illumos developers list[22:37:18] <btQuark> it even showed up on twitter...[22:37:18] <btQuark> yuck[22:37:40] *** n1c0b has joined #illumos[22:37:52] <alanc> people's lunch shows up on twitter, a developer spat there is no big news[22:38:45] *** jamesd2 has joined #illumos[22:42:35] <davenz> https://davekoelmeyer.wordpress.com/2010/11/11/oracles-virtualisation-bundles-have-no-license-costs/[22:42:36] *** libkeiser has quit IRC[22:42:49] <btQuark> indeed, i would rather be irritated about the mails than the twitters...[22:42:59] <btQuark> anyhow, things will straighten out[22:44:31] *** libkeiser has joined #illumos[22:45:44] *** n1c0b has quit IRC[22:48:30] *** VertexSymphony has quit IRC[22:49:14] *** boyd has quit IRC[22:49:37] *** boyd has joined #illumos[22:53:17] *** libkeise1 has joined #illumos[22:53:55] *** boyd_ has joined #illumos[22:54:08] *** libkeiser has quit IRC[22:54:11] *** boyd has quit IRC[22:54:11] *** boyd_ is now known as boyd[22:58:41] *** boyd has quit IRC[22:59:52] *** boyd has joined #illumos[23:04:32] *** boyd has quit IRC[23:07:23] *** boyd has joined #illumos[23:15:53] *** Tenzer is now known as Tenzer|off[23:24:08] *** ShadowHntr has quit IRC[23:25:30] *** ShadowHntr has joined #illumos[23:26:21] *** movement has quit IRC[23:28:01] *** mikw has quit IRC[23:31:56] *** ShadowHntr has quit IRC[23:38:22] *** boyd_ has joined #illumos[23:38:44] *** boyd has quit IRC[23:38:45] *** boyd_ is now known as boyd[23:39:08] *** movement has joined #illumos[23:40:09] *** N2Deep has quit IRC[23:45:29] *** movement has quit IRC[23:54:43] *** psychicist has quit IRC[23:55:41] *** smrt has quit IRC[23:55:59] *** smrt has joined #illumos[23:57:36] *** movement has joined #illumos