1. SCons
  2. Core
  3. SCons

Wiki

Clone wiki

SCons / BugParty / IrcLog2010-01-19

16:49:02  *      garyo (n=[garyo@209-6-36-50.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com](mailto:garyo@209-6-36-50.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)) has joined #scons 
16:53:33  *      sgk (n=[sgk@c-71-204-129-67.hsd1.ca.comcast.net](mailto:sgk@c-71-204-129-67.hsd1.ca.comcast.net)) has joined #scons 
16:54:27  *      Jason_at_Intel (n=[chatzill@12.18.240.224](mailto:chatzill@12.18.240.224)) has joined #scons 
16:54:37  <sgk>  [GregNoel](GregNoel):  i'll be a bit late getting back to the meeting, ~1705 - 1710 
16:54:48  <Jason_at_Intel>       hello 
16:55:36  <garyo>        Hi folks 
16:55:46  <Jason_at_Intel>       Hi Gary! 
17:00:52  *      You are no longer marked as being away 
17:00:58  <[GregNoel](GregNoel)>     Hi, all.  It's raining in San Diego and I've already had a few power fluctuations today (but no trips yet, knock on wood). 
17:00:59  *      [GregNoel](GregNoel) raps smartly on his head 
17:00:59  <[GregNoel](GregNoel)>     Ow!  In any event, if I suddenly drop off-line, assume I've lost power and I'll get back as soon as I can. 
17:01:51  <garyo>        Hi Greg. 
17:08:44  <garyo>        SO folks, on to the bugs? 
17:09:09  <sgk>  hey all 
17:09:16  <Jason_at_Intel>       Hi steve! 
17:09:21  <[GregNoel](GregNoel)>     Ah, welcome back. 
17:09:23  <garyo>        Hi Steve 
17:09:25  <[GregNoel](GregNoel)>     Shall we get started?  2534 is first up. 
17:09:43  <garyo>        George Foot's comment seems apropos.  Doc p3? 
17:10:03  <[GregNoel](GregNoel)>     No, Steven thought it was a bug last time. 
17:10:33  <garyo>        OK I'm confused, the ssheet note says default type is Entry, but I think it's File. 
17:10:43  <garyo>        Steven? 
17:10:52  <[GregNoel](GregNoel)>     Steven? 
17:10:59  <Jason_at_Intel>       I thought this was to be researched 
17:11:21  <garyo>        I like that: steven, research. 
17:11:30  <sgk>  sorry, catching up on the conversation 
17:11:35  <sgk>  been a hellish day 
17:11:40  <garyo>        sorry 2 hear that 
17:11:49  <garyo>        I have a lot of those these days :-/ 
17:11:55  <sgk>  ooo sick; power outage; etc. 
17:12:29  <[GregNoel](GregNoel)>     My lights just flickered, so I can empathize (not sympathize) 
17:12:21  <sgk>  2534:  think we should give it to me to research 
17:12:25  <garyo>        +1 
17:12:37  <[GregNoel](GregNoel)>     done 
17:12:50  <[GregNoel](GregNoel)>     1910, 2361, 780, 914, 1187, 1745, 1883, 1945: bypass for lack of comments 
17:12:50  <[GregNoel](GregNoel)>     (Yes, Steven, 914 reluctantly) 
17:12:50  <[GregNoel](GregNoel)>     2058 consensus 2.0 p1 stevenknight 
17:12:50  <[GregNoel](GregNoel)>     2070 consensus 2.0 p1 stevenknight 
17:12:50  <[GregNoel](GregNoel)>     2096, I think Gary needs to agree, but otherwise consensus 2.x p4 +sconf_revamp 
17:12:50  <[GregNoel](GregNoel)>     2249 
17:13:27  <Jason_at_Intel>       woo 
17:14:10  <garyo>        2096: I agree, 2.x p4 
17:14:52  <[GregNoel](GregNoel)>     2096, done 
17:13:47  <sgk>  greg re: 914 
17:13:53  <[GregNoel](GregNoel)>     yes? 
17:13:58  <sgk>  how about if we make that future p4 or something so it doesn't get lost 
17:14:07  <sgk>  agree that the general concept of XML for results is good 
17:14:18  <sgk>  but nothing is burning for it right now 
17:14:33  <[GregNoel](GregNoel)>     future p4 is off the radar, but I'll go along. 
17:15:19  <[GregNoel](GregNoel)>     914, other consensus? 
17:15:30  <garyo>        agree w/ 914 
17:15:32  <sgk>  future p3, then 
17:15:37  <[GregNoel](GregNoel)>     done 
17:16:05  <garyo>        My Issue List query doesn't match the spreadsheet today :-/ 
17:16:58  <[GregNoel](GregNoel)>     garyo, a bunch of issues from last time are still there, formerly Steven's research issues. 
17:16:18  <sgk>  2249:  research bdbaddog (since he volunteered) 
17:16:26  <[GregNoel](GregNoel)>     done 
17:16:58  <garyo>        OR: 2249: ask OP if it still happens. 
17:17:34  <sgk>  2249:  that's a reasonable suggestion for the update 
17:17:44  <garyo>        It's a year old and vs_revamp is better. 
17:17:49  <sgk>  yep 
17:18:11  <sgk>  2249:  research bdbaddog, comment suggests asking OP 
17:18:17  <garyo>        +1 
17:18:23  <[GregNoel](GregNoel)>     done 
17:18:39  <[GregNoel](GregNoel)>     2304 
17:18:39  <sgk>  2304:  research SK 
17:18:42  <Jason_at_Intel>       I should comment on 2304, I might have a work around for this one, but i have to finish the testing of this in Parts yet 
17:19:19  <[GregNoel](GregNoel)>     I don't believe that's possible 
17:18:59  <sgk>  Jason_at_Intel:  sounds good 
17:19:17  <garyo>        OK, let's assign it to jason to research? 
17:19:33  <Jason_at_Intel>       it will probably need a patch to the File object 
17:19:45  <garyo>        ??? 
17:19:45  <sgk>  [GregNoel](GregNoel): re your comment:  agreed we may not be able to build on top of the locked executable 
17:19:51  <[GregNoel](GregNoel)>     rebuilding a program as the same time you're running it is NOT a good idea, no matter where or when. 
17:20:00  <sgk>  but we should handle the error gracefully and not die 
17:20:04  <garyo>        Of course not on Windows, but at least interactive mode could recover. 
17:20:10  <Jason_at_Intel>       I agree with that.. but the system should  not die 
17:20:24  <sgk>  you should be able to stay in interactive mode and build other things 
17:20:33  <sgk>  or rebuild the same executable next time 
17:20:31  <garyo>        ok we all agree.  research jason? 
17:20:31  <[GregNoel](GregNoel)>     On *IX, the file is overwritten, so the code will change out from under the running program.  NOT a good idea. 
17:20:36  <Jason_at_Intel>       Scons dies with the unlink Action() 
17:21:01  <garyo>        Greg: on *IX I do this all the time, multiple times daily.  And yes, sometimes I pay the price. :-) 
17:21:23  <sgk>  agreed, but people do things like this, so the question is how do we want to handle it? 
17:21:26  <[GregNoel](GregNoel)>     consensus? 
17:21:37  <sgk>  research Jason_at_Intel if he's up for it 
17:21:39  <sgk>  otherwise SK 
17:21:47  <garyo>        Recover as best as possible, with build failure msg.  Just the same as "out of disk space." 
17:22:01  <Jason_at_Intel>       basically 
17:22:11  <garyo>        jason, will you take it on? 
17:22:16  <Jason_at_Intel>       Sure 
17:22:19  <[GregNoel](GregNoel)>     OK 
17:22:20  <garyo>        ok, done 
17:22:25  <sgk>  Jason_at_Intel++ 
17:22:39  <[GregNoel](GregNoel)>     2347 consensus 2.x p3 +symlink 
17:22:39  <[GregNoel](GregNoel)>     2536 
17:22:57  <sgk>  i like greg's suggestion of research OP 
17:23:17  <[GregNoel](GregNoel)>     so do I, or I wouldn't have suggested it. {;-} 
17:23:36  <garyo>        Yeah, makes sense 2 me too. 
17:23:39  <sgk>  done 
17:23:49  <[GregNoel](GregNoel)>     done 
17:23:53  <[GregNoel](GregNoel)>     2537 FIXED (thanks, Steven!) 
17:23:53  <[GregNoel](GregNoel)>     2538 I don't like anytime issues assigned to Steven, but I'll go with the flow 
17:24:06  <sgk>  agreed on both counts 
17:24:21  <sgk>  done 
17:24:25  <sgk>  ? 
17:24:36  <[GregNoel](GregNoel)>     consensus? 
17:24:50  <garyo>        ok 
17:24:56  <[GregNoel](GregNoel)>     done 
17:24:59  <[GregNoel](GregNoel)>     2540 FIXED (thanks, Steven!) 
17:24:59  <[GregNoel](GregNoel)>     (We've achieved the minimal target for today; congratulations to us!) 
17:24:59  <[GregNoel](GregNoel)>     2541, 2542, 2545: bypass for lack of comments 
17:24:59  <[GregNoel](GregNoel)>     2546 consensus WONTFIX 
17:24:59  <[GregNoel](GregNoel)>     2547, 2548, 2549: bypass for lack of comments 
17:24:59  <[GregNoel](GregNoel)>     (We've achieved the nominal target for today; I suppose congratulations are due, but it seems rather empty since we'll see so many again next time.) 
17:24:59  <[GregNoel](GregNoel)>     2550, 2551, 2552, 2553, 2554, 2555, 2556 (so close!), 2557: bypass for lack of comments 
17:24:59  <[GregNoel](GregNoel)>     And that's all for today...  Is there any discussion needed about releasing 1.3?  With the checkpoint released today, plan for two weeks, so about January 31st?  (That is, a year late...) 
17:25:18  <garyo>        Steven, I can work that kind of thing in the bg so if it becomes a timesink let me know. 
17:25:53  <sgk>  garyo:  the irc channel? 
17:25:57  <garyo>        sgk: yes. 
17:26:22  <sgk>  cool, i'll take you up on that if it gets thorny 
17:26:28  <garyo>        np 
17:25:45  <garyo>        Greg: I think since this ckpoint is very minimally different from the last except doc, a week is enough.  Just my opinion though. 
17:26:59  <[GregNoel](GregNoel)>     garyo, yes, but only if it's really pushed on the mailing lists to get the maximal feedback. 
17:27:18  <garyo>        I'm itching to get 1.3 out.  I think we've checkpointed it to death.  Greg: great idea re: ML pushing.  I will do that. 
17:27:22  <bdbaddog>     Greetings.. sorry I'm late.. 
17:27:53  <garyo>        Hey Bill!  Good job getting both checkpoints out! 
17:27:40  <Jason_at_Intel>       so what is holding up 1.3 
17:27:52  <Jason_at_Intel>       I am sort of for making it happen 
17:27:36  <sgk>  should we try to look at some of the issues that have two consensus comments? 
17:27:54  <[GregNoel](GregNoel)>     sgk, cherry-pick away... 
17:28:02  <sgk>  if others add "verbal" consensus we won't have to revisit so many 
17:29:09  <[GregNoel](GregNoel)>     that was the point of the fusillade, so if you want to try to get verbal consensus to get rid of them, I'm all for it. 
17:28:18  *      sgk scrolls back in the spreadsheet... 
17:29:08  <sgk>  2547 and 2548:  future p1 +java 
17:29:10  <sgk>  any objections? 
17:29:56  <[GregNoel](GregNoel)>     2547, 2548, I'll go along; any others? 
17:29:37  <sgk>  hang on, table 2547 and 2548, let's finish the 1.3 discussion first 
17:30:36  <garyo>        2547,2548: I'm fine w/ future p1 +java.  sigh. :-) 
17:31:04  <[GregNoel](GregNoel)>     2547, 2548, that's three; done 
17:28:20  <garyo>        Jason: we just need to let Bill's last checkpoint stew for long enough to make sure it's not got horrible bugs. 
17:28:45  <Jason_at_Intel>       ok 
17:29:10  <garyo>        That's why I say 1 week, not the usual 2.  Bill, opinion on how long to let the checkpoint stew? 
17:29:43  <bdbaddog>     We going with 2 weeks again (of stew time)?  I'd be up for shorter, but I don't remember what changes between last checkpoint and this one was, if mostly doc, then 1week is good by me. 
17:30:05  <sgk>  1 week seems good to me 
17:30:12  <garyo>        I put in one significant fix to detect 64 vs. 32 bit OS default.  That's it I think. 
17:30:16  <bdbaddog>     O.k. so 1/24/2010 
17:30:30  <sgk>  i'm anticipating more users and problem reports when we officially release 1.3 regardless of whether it's 1 or 2 weeks 
17:30:46  <garyo>        sgk: I agree, that's another reason to get it out sooner. 
17:30:57  <sgk>  at this point an extra week soaking 1.3 checkpoints isn't the crucial difference between success and failure 
17:30:57  <garyo>        So we can do 1.3.1 sooner :-/ 
17:31:20  <sgk>  right, we should be ready to jump on 1.3 bug reports and prep 1.3.1 
17:31:24  <sgk>  my gut says about a month after 
17:31:44  <garyo>        sgk: maybe, let's see. 
17:31:23  <garyo>        ok, is that consensus?  Next weekend is 1.3? 
17:31:56  <Jason_at_Intel>       I agree... get 1.3 out earlier and let a 1.3.1 fix issues found in 1.3.. this way more people will test it 
17:31:57  <garyo>        But we can start on the 2.0 work immediately. 
17:32:01  <sgk>  i'll be happy to be wrong and go straight to 2.0... :-) 
17:32:12  <[GregNoel](GregNoel)>     yep 
17:32:20  <bdbaddog>     so at 1.3 release we create a 1.3 branch? 
17:32:32  <bdbaddog>     and trunk becomes pre-2.0 ? 
17:32:32  <[GregNoel](GregNoel)>     Er, SVN doesn't work that way. 
17:32:45  <[GregNoel](GregNoel)>     No need to branch until you need it. 
17:32:56  <[GregNoel](GregNoel)>     If you know what I mean. 
17:33:24  <sgk>  either way it'd have the same effect 
17:33:31  <bdbaddog>     so just cp and then branch from that if needed? 
17:33:35  <garyo>        Sure, but we might as well create it when we do 1.3 so people have an official place to work on 1.3.1 rather than making the first person to fix something have to create it. 
17:33:45  <sgk>  i can see an argument for branching right away because I doubt we'll get through a big change like vs_revamp in 1.3 unscathed 
17:33:46  <garyo>        That's how I usually do it. 
17:33:59  <sgk>  what garyo said 
17:34:01  <Jason_at_Intel>       it seems we woudl want to make a 1.3 branch and have truck be 2.0.. and make a 1.3.1 branch we add patches to for a 1.3.1 drop 
17:34:06  <bdbaddog>     +1 for branching 1.3 at 1.3 release. 
17:34:20  <sgk>  okay, let's go for it (scm orthodoxy be damned...  :-)) 
17:34:21  <garyo>        Just cp trunk to tag, then cp tag to branch.  And yes, trunk becomes pre-2.0. 
17:34:25  <bdbaddog>     there'd be a 1.3 release, and a 1.3 branch 
17:34:34  <bdbaddog>     +1 garyo's description. 
17:35:01  <[GregNoel](GregNoel)>     So branch trunk to branches/1.3.fixes or some such?  Sure, why not?  Branches are cheap. 
17:35:30  <[GregNoel](GregNoel)>     ... or branch release?  That should be discussed. 
17:34:55  <bdbaddog>     so then BB does trunk? and/or 1.3? 
17:35:27  <garyo>        what's BB, Bill? 
17:35:32  <bdbaddog>     [BuildBot](BuildBot) 
17:35:36  <garyo>        ah yes. 
17:35:49  <garyo>        Good question. 
17:36:06  <garyo>        Is it realistic to want both? 
17:36:19  <[GregNoel](GregNoel)>     BB already does trunk, checkpoint, release, and branches/*, so what are you saying? 
17:36:40  <garyo>        Really?  I didn't know that.  In that case, just carry on. 
17:37:07  <sgk>  yeah, [GregNoel](GregNoel) added mutliple branch support 
17:37:25  <sgk>  surprised me too, but it tested all of bdbaddog's recent checkpoint patches 
17:39:02  <bdbaddog>     Way to go Greg! 
17:37:31  <garyo>        Cool, is that just in our BB or sent upstream? 
17:38:02  <[GregNoel](GregNoel)>     It's one of the examples in the BB manual, so I think they know about it. 
17:38:13  <garyo>        good. 
17:38:31  <garyo>        OK, sounds like most of a plan! 
17:38:44  <bdbaddog>     :) 
17:38:37  <sgk>  go 1.3! 
17:39:01  <[GregNoel](GregNoel)>     Geaux Saints! 
17:38:58  <garyo>        Yeah!  And thanks bdbaddog for all the grunt work on the checkpoints! 
17:39:15  <sgk>  okay, and we have consensus on 2547, 2548 
17:39:20  <sgk>  on to more of the issues? 
17:39:34  <bdbaddog>     sure. 
17:39:36  <[GregNoel](GregNoel)>     any others you want to cherry-pick? 
17:40:17  <garyo>        how about 2556? 
17:40:17  <sgk>  2556:  send it back to OP for test case? 
17:40:23  <sgk>  jinx 
17:40:42  <garyo>        OK, 2556 came from a ML discussion.  I'll ask him to paste in the testcase. 
17:41:56  <[GregNoel](GregNoel)>     2556, should we close it and ask him to reopen when he adds the test case? 
17:42:14  <sgk>  2556:  sure, that works 
17:42:14  <garyo>        Nah, I'll just comment on the tkt as is. 
17:42:37  <[GregNoel](GregNoel)>     (is there such a thing as a reverse jinx?) 
17:42:23  <garyo>        I asked him to report it anyway. 
17:42:27  <sgk>  oh, okay 
17:41:17  <[GregNoel](GregNoel)>     done 
17:40:27  <Jason_at_Intel>       2542? 
17:40:40  <sgk>  and 2557:  conensus 2.1 p2 rob? 
17:41:13  <garyo>        +1 on 2557. 
17:41:17  <[GregNoel](GregNoel)>     done 
17:41:29  <sgk>  cool 
17:41:40  <sgk>  i have another topic if we're (nominally) done with issues for this week 
17:41:51  <garyo>        shoot 
17:42:48  <sgk>  next topic:  scrapping runtest.py in favor of using unittest as the harness 
17:43:06  <[GregNoel](GregNoel)>     Ulp. 
17:43:28  <sgk>  turned out to be pretty easy to do 
17:43:51  <sgk>  or at least to prototype 
17:43:57  <[GregNoel](GregNoel)>     How would it work? 
17:44:09  <sgk>  well, that's why i want to discuss it 
17:44:29  <Jason_at_Intel>       you mean the python unit test lib? 
17:44:12  <garyo>        sgk: Can you reuse all the stuff in QMTest/ dir?  There's a lot of value in there. 
17:44:31  <garyo>        (e.g. TestSCons.py, [TestCmd](TestCmd).py) 
17:44:39  <sgk>  garyo:  yeah, that stuff all stays 
17:44:50  <garyo>        ok, good. 
17:44:56  <sgk>  those are basically test fixtures 
17:45:11  <sgk>  what i prototyped was a simple shotgun marriage 
17:45:27  <sgk>  where unittest still calls out (using subprocess) to execute individual test scripts like today 
17:45:35  <sgk>  but it reports the results back using unittest conventions 
17:45:46  <sgk>  so it looks like a python unittest for reporting 
17:45:55  <sgk>  but is (at least for now) still just executing scripts 
17:46:04  <bdbaddog>     does that enable/block parallel testing? 
17:46:23  <sgk>  unfortunately, it looks like unittest itself doesn't enable parallel testing 
17:46:27  <sgk>  so it's orthogonal 
17:46:48  <sgk>  i.e. this change doesn't make parallel testing more or less likely 
17:46:46  <garyo>        So the main benefit is prettier output? 
17:47:14  <sgk>  yeah, this is why I was a little... hesitant... when the idea of unittest seemed to get a positive reaction from you guys a few weeks back 
17:47:19  <sgk>  (iirc) 
17:47:29  <Jason_at_Intel>       Do you have thie prototyped checked in? 
17:47:45  <[GregNoel](GregNoel)>     Hmmm...  I'd have to think about this.  Why don't you add it to SConsTestingRevisions in the wiki? 
17:47:33  <bdbaddog>     upside is less infrastructure to maintain? 
17:47:49  <sgk>  right, runtest.py is all homebrew, this leverages unittest 
17:48:16  <sgk>  but that doesn't necessarily seem like a really compelling reason to switch 
17:48:06  <Jason_at_Intel>       I coudl use it in Parts i think to handle some cases that are hard to handle with a straight unittest code 
17:48:27  <garyo>        runtest.py is only 834 lines, not huge. 
17:48:31  <[GregNoel](GregNoel)>     You'd still need a test runner, but it wouldn't be as complex...  I still need to think about it. 
17:48:55  <sgk>  okay, the code isn't much, I'll add it to the wiki and send out something to the ML for discussion 
17:49:12  <bdbaddog>     sounds good. then we can play with it (time allowing).. 
17:49:16  <garyo>        sgk: my biggest beef with QMtest was managing stdout/stderr when debugging tests (not just running them where you always want it hidden). 
17:49:30  <[GregNoel](GregNoel)>     concur 
17:49:32  <garyo>        Would unittest maybe help? 
17:49:31  <sgk>  garyo:  this doesn't solve that 
17:49:37  <garyo>        oh well. 
17:49:45  <sgk>  you're still running external scripts and having to capture the output 
17:49:51  <sgk>  what I could do about that, though.... 
17:50:05  <sgk>  is expose the Trace() function as a supported part of the API 
17:50:24  <sgk>  that writes to /dev/tty on POSIX and con on Windows 
17:50:34  <sgk>  so you can at least add debug prints easily 
17:50:18  <garyo>        That'd help somewhat. 
17:50:41  <Jason_at_Intel>       unittest makes it easy to set up a structure to run on a give test at a time 
17:51:03  <garyo>        runtest.py can do that too. 
17:51:02  <sgk>  okay, follow-on test discussion: 
17:51:28  <sgk>  all the in-line test code in strings 
17:51:44  <sgk>  which makes it a pain to read the test code 
17:51:51  <sgk>  and complicates going forward to Python 3.0 
17:51:58  <garyo>        plus most of them are big regexes now anyway :-) 
17:52:17  <garyo>        you have a solution?! 
17:52:38  <sgk>  yeah, i think so 
17:52:52  <sgk>  we change the unit of test granularity from individual files to directories 
17:53:21  <sgk>  each directory is a test config with the input SConstruct *.c etc. files checked in directly 
17:52:30  <loonycyborg>  sgk: btw how exactly you're planning to upgrade to python3? 
17:52:52  <[GregNoel](GregNoel)>     Hi, Sergey; thanks for being with us again! 
17:52:54  <garyo>        Hi Sergey 
17:52:58  <sgk>  Sergey++ 
17:53:13  <loonycyborg>  That'll require all SConstruct files moved to python3 too. 
17:53:23  <[GregNoel](GregNoel)>     directories, erk.. 
17:53:48  <bdbaddog>     That's the way I usually setup regressions for my clients.. so I'm all for that. 
17:53:49  <garyo>        radical! 
17:54:01  <bdbaddog>     A dir per test. or group of tests? 
17:53:47  <sgk>  the test script(s) within a directory have identifying prefixes 
17:54:08  <sgk>  so you can actually have multiple individual tests re-using the same config (directory) 
17:54:32  <sgk>  the test infrastructure copies the directory contents (except for test scripts and .svn directories) to the temp dir to run the test 
17:54:41  <loonycyborg>  You could as well make a full api overhaul coincide with moving to python3 :P 
17:55:10  <sgk>  loonycyborg:  that's actually worth considering 
17:55:26  <sgk>  at least, maybe use it as a clean way to shed some of the cruftier parts of the current API 
17:55:25  <garyo>        sgk: I get it -- so even sub/a/b kind of things just get checked in. 
17:55:38  <sgk>  garyo:  right 
17:55:48  <sgk>  i think it makes the test configurations more comprehensible 
17:55:52  <bdbaddog>     I've very tired of having to put all the files in strings in the tests.. 
17:56:00  <sgk>  yep 
17:56:12  <garyo>        sgk: I think there are still some cases where files have to be dynamically created etc. but a small %age, so this would clean up a lot. 
17:56:22  <sgk>  right, it also doesn't completely solve things like updating files as an intermediate step 
17:56:55  <garyo>        But what about testing stdout/stderr?  Special files to represent expected output? 
17:57:16  <sgk>  yeah, those could either be checked in as .golden files or something 
17:57:29  <sgk>  or else those can stay in-line if it makes more sense to not clutter the config 
17:57:17  <bdbaddog>     "golden" files in my industry.. 
17:57:39  <Jason_at_Intel>       gold files here :-) 
17:57:44  <sgk>  i think that's less of a problem since they're (typically) not actually code 
17:58:03  <garyo>        Cool.  With regex semantics I assume.  I want to think about this, but it seems very sensible on the face of it. 
17:58:17  <[GregNoel](GregNoel)>     concur w/ garyo 
17:58:32  <sgk>  okay, let's let it sink in for a bit 
17:58:40  <bdbaddog>     +1 for me. I'm all for it. 
17:58:49  <garyo>        I'll pick a few tests randomly to see how they'd look. 
17:58:51  <sgk>  if we want to go forward with it, i'm looking for a read on the right priority 
17:58:59  <[GregNoel](GregNoel)>     I think adding it to the existing wiki page for testing cleanup is the first step. 
17:59:01  <sgk>  we could actually also do it incrementally 
17:59:17  <garyo>        sgk: +1 
17:59:20  <Jason_at_Intel>       incrementally ++ 
17:59:25  <[GregNoel](GregNoel)>     that's why I suggest the wiki page, so we could break down what can be done first 
17:59:36  <sgk>  sure 
17:59:44  <[GregNoel](GregNoel)>     ... and add a discussion page to the wiki page... 
17:59:53  <garyo>        If you can make the test runner run these, that would be a great step. 
17:59:50  <sgk>  my rough idea for incremental is 
18:00:03  *      sgk decides to save that for the wiki page... 
17:59:50  <bdbaddog>     BTW.. are we due for a moin moin update? 
18:00:10  <garyo>        bdbaddog: are we?  I can do that if needed. 
18:00:30  <[GregNoel](GregNoel)>     moin, long overdue; they're at 1.9 already... 
18:00:45  <garyo>        OK, I'll look into it in the next wk or 2. 
18:00:50  <sgk>  garyo:  thnx 
18:01:15  <[GregNoel](GregNoel)>     garyo, please coordinate with me, so we can upgrade at the same time. 
18:01:16  <Jason_at_Intel>       <wishing the Tigris wiki site would be upgraded > 
18:01:34  <garyo>        Greg: will do. 
18:01:09  <sgk>  what should be my top priority right now?  the legal stuff? 
18:01:53  <garyo>        Steven: what are the choices? 
18:02:23  <[GregNoel](GregNoel)>     sgk, I nominate finishing out any 1.3 stuff, then the legal stuff, then the 2.0 stuff 
18:02:48  <sgk>  legal stuff, 1.3 stuff, 2.0 stuff, test infrastructure, performance graph stuff 
18:03:11  <Jason_at_Intel>       greg++ 
18:03:11  <sgk>  actually, i don't have any 1.3 issues on my list at the moment 
18:03:32  <[GregNoel](GregNoel)>     yeah, but something might appear.  set the priorities now. 
18:03:17  <garyo>        ok, get the legal stuff done with.  I don't think you have any significant 1.3 todos. 
18:03:41  <garyo>        [GregNoel](GregNoel)++ 
18:03:39  <sgk>  oh, also string template refactoring 
18:04:03  <[GregNoel](GregNoel)>     what refactoring? 
18:04:23  <sgk>  re-doing subst() and subst_list() for performance 
18:03:53  <garyo>        2526? 
18:04:06  <sgk>  yeah 
18:04:27  <garyo>        My opinion: too late for 1.3, it could introduce new bugs. 
18:04:36  <sgk>  using the same generatl technique of string.Template in the Python lib 
18:04:46  <sgk>  oh, definitely not that for 1.3 
18:05:10  <sgk>  don't want to derail that train 
18:05:06  <bdbaddog>     I think for 1.3, only doc bugs and regressions.. 
18:05:20  <garyo>        ok, so make 2526 2.1 then. 
18:05:51  <bdbaddog>     are you on to another spreadsheet? 
18:06:08  <garyo>        (No, I just brought up all Steven's issues.) 
18:06:15  <bdbaddog>     ahh o.k. 
18:05:56  <garyo>        So I'm with Greg: any 1.3 issues, then legal stuff, then 2.0 stuff. 
18:06:50  <sgk>  okay, i think i'm set then 
18:06:57  <garyo>        excellent. 
18:06:58  *      sgk grits his teeth in anticipation of talking to lawyers 
18:07:37  <garyo>        On a totally unrelated note, my Nexus One is the most awesome pocket computing device ever made.  Google ftw! 
18:08:30  <sgk>  garyo:  i'm totally digging mine.  the keyboard interface is good for you? 
18:08:59  <garyo>        sgk: yes, no problems at all. 
18:07:41  <[GregNoel](GregNoel)>     I'd also reconsider the 1.3 all-doc issues (upgrading the user's guide) and see if they can be done for 2.0, but have Steven parcel them out to sucke, ah, people of his choice. 
18:08:14  <garyo>        Good point Greg -- we need to move the rest of the 1.3 issues forward.  Can you do that? 
18:08:34  <[GregNoel](GregNoel)>     For everything except the doc, yes, I have a policy I apply. 
18:08:41  <garyo>        Now that the doc procedure is improved I should be able to help there. 
18:08:56  *      sgk remembers that he has to write up that README... 
18:09:35  <garyo>        re: doc readme: yes please. 
18:09:57  <sgk>  k, that's probably P(-1) for me then 
18:10:03  <sgk>  the legal stuff is all P0 
18:10:08  <[GregNoel](GregNoel)>     (For what it's worth, the policy is to move the issues forward to the next point release, but bump their priority by one so they'll get more consideration at the next level.) 
18:10:29  <garyo>        Sensible. 
18:10:30  <sgk>  sounds good 
18:11:02  <[GregNoel](GregNoel)>     Is applying for a group also p0? 
18:11:15  <sgk>  i suppose so 
18:11:28  <[GregNoel](GregNoel)>     I'm in agreement then. 
18:11:25  <garyo>        Are we going to get kicked out or something if we don't apply? 
18:11:51  <garyo>        If so, then I agree.  This channel works nicely. 
18:12:02  <sgk>  garyo:  i don't think we're going to get kicked out 
18:12:17  <sgk>  but we're restricted in what we can do administratively 
18:12:25  <[GregNoel](GregNoel)>     garyo, no, we're grandfathered, since our channel was formed 6+ years ago, and the policy is only 4 years old. 
18:12:46  <[GregNoel](GregNoel)>     but it would be good to abide by the policy. 
18:12:52  <sgk>  it was a bit of a hassle to transfer ownership from TTimo to me because we don't fully exist under the new rules 
18:13:05  <garyo>        Makes sense to get it done then. 
18:13:23  <[GregNoel](GregNoel)>     can at least start the process to show good faith. 
18:13:42  <sgk>  BTW, [GregNoel](GregNoel):  i looked around at some other channels 
18:14:00  <sgk>  it seems like using | as the separator between sections of the topic is more the norm than - 
18:14:10  <[GregNoel](GregNoel)>     OK, wilco 
18:14:51  *      sgk has changed the topic to: SCons | building software, better | [http://www.scons.org](http://www.scons.org) | next bug triage party is Tue 2 Feb 2010 17h00 US/Pacific 
18:14:55  <garyo>        Greg, guess your power held out today! 
18:15:01  <[GregNoel](GregNoel)>     so far... 
18:14:50  <[GregNoel](GregNoel)>     Next party on Groundhog's day? 
18:15:15  <sgk>  excellent!  we can have the next bug party over and over and over again... 
18:15:24  <garyo>        :-) 
18:15:49  <[GregNoel](GregNoel)>     sgk, ++ 
18:15:36  <[GregNoel](GregNoel)>     Ten points to anybody who can spell where the Groundhog's Day official celebration is held... 
18:15:47  <sgk>  Punxatawney PA 
18:15:47  <garyo>        punxatawney 
18:16:04  <garyo>        sgk: you win 
18:16:05  <[GregNoel](GregNoel)>     jinx, you guys are too good. 
18:16:20  <garyo>        I was a spelling bee kid in my youth 
18:16:41  <Jason_at_Intel>       :-) I was not :-) 
18:17:03  <[GregNoel](GregNoel)>     It's not far from where my wife grew up, so she can frustrate me by spelling it...  all I can manage is punx* 
18:16:39  <bdbaddog>     And a the groudhog's name? 
18:16:44  <sgk>  Phil 
18:16:44  <garyo>        phil 
18:16:59  <bdbaddog>     :) 
18:17:03  <Jason_at_Intel>       look at the trivial buffs 
18:17:09  <[GregNoel](GregNoel)>     yes, Phill 
18:17:16  <sgk>  damn, the amount of grey matter garyo and i have devoted to this stuff is scary 
18:17:22  <[GregNoel](GregNoel)>     (two ll's I think) 
18:17:33  <garyo>        two Ls: I doubt it. 
18:18:13  <[GregNoel](GregNoel)>     oops, the wife agrees. 
18:18:18  <[GregNoel](GregNoel)>     mea culpa. 
18:18:35  <garyo>        anyway, bdbaddog: can you manage the release cycle?  If so, how can I help? 
18:19:00  <bdbaddog>     as far as building and pushing the packages, no prob. 
18:19:16  <garyo>        OK, I can help w/ release announcements etc. 
18:19:22  <bdbaddog>     might need some help on the create 1.3 branch commands. 
18:19:32  <garyo>        np, just ask. 
18:19:45  <Jason_at_Intel>       Steve: code review? good time to talk ? 
18:20:30  <sgk>  Jason_at_Intel:  let's set up something for next week 
18:21:11  <Jason_at_Intel>       Steve.. I have a check in to sync with teh public drop yet... but that sounds good 
18:21:24  <Jason_at_Intel>       do we want anyone else to be part of this? 
18:21:30  <Jason_at_Intel>       such as Gary or Greg? 
18:21:58  <garyo>        If I can get out there I'd love to.  I don't get to travel much these days... hopefully spring will slow down @ work a little. 
18:22:05  <sgk>  garyo/GregNoel/bdbaddog:  any of you interested in participating in discussion of how to integrate Parts ? 
18:22:16  <bdbaddog>     for 3.0? 
18:22:27  <Jason_at_Intel>       or 2.x 
18:22:31  <sgk>  not sure, that needs discussion 
18:22:34  <bdbaddog>     :) 
18:22:37  <Jason_at_Intel>       depending on what people think 
18:22:39  <sgk>  what it should ultimately look like, time frame, etc. 
18:22:45  <garyo>        I am, especially the toolchain part of it. 
18:22:55  <garyo>        (or merge it w/ other toolchain ideas) 
18:22:59  <sgk>  right 
18:23:16  <bdbaddog>     is there a wiki page to bring me up to speed? 
18:23:19  <sgk>  k, i'm thinking we should structure it like a subproject for now? 
18:23:24  <Jason_at_Intel>       can i set up a phone conference for this.. talk is better for me 
18:24:05  <Jason_at_Intel>       I can set up one through work 
18:24:13  <Jason_at_Intel>       as this is a work task for me :-) 
18:23:45  <bdbaddog>     I hate talking.. IM is my friend. 
18:24:05  <[GregNoel](GregNoel)>     I hate talking AND typing.. Mail is my friend. 
18:24:13  <bdbaddog>     GN: :) 
18:23:31  <garyo>        skype is good 
18:24:17  <sgk>  i'm okay w/phone or skype so long as someone volunteers to take good notes 
18:24:27  <sgk>  the best thing about IM to me is the log 
18:24:33  <garyo>        That's a really good point. 
18:24:45  <Jason_at_Intel>       and that is fine.. but teh review will be slow and long 
18:25:01  <sgk>  we may want a mix of the two 
18:25:07  <garyo>        We need to prepare first so we can go faster. 
18:25:19  <sgk>  (garyo:  good point re: prep) 
18:25:10  <sgk>  how about start w/voice or skype 
18:25:35  <sgk>  since the first conversations will be a little more brainstorm-y 
18:26:02  <sgk>  and (maybe) transition to irc if we start settling on things and only have to iron out details 
18:26:01  <Jason_at_Intel>       this sound reasonable... get our ducks in a row... 
18:26:16  <garyo>        ok w/ me 
18:26:18  <sgk>  over time, i mean, not within one session 
18:26:46  <[GregNoel](GregNoel)>     ([YouTube](YouTube), anyone? {;-}) 
18:26:59  <sgk>  okay, so the action item is that Jason_at_intel and I set up a time for a voice / skype chat next week 
18:27:07  <bdbaddog>     I'd say write something up on the wiki, give us a week on email list, then conf call. 
18:27:16  <sgk>  open invitation to the conf. call 
18:27:24  <garyo>        Sounds good. 
18:27:34  <sgk>  w/approval of course 
18:27:42  <sgk>  and we'll figure out the record-keeping by then 
18:27:51  <sgk>  and try to have some agenda ahead of time 
18:27:58  <bdbaddog>     k. 
18:28:13  <sgk>  based on that, i'm thinking it makes more sense for Jason_at_intel and i to organize next week 
18:28:22  <sgk>  with a target of the conf. call the week following 
18:28:23  <Jason_at_Intel>       SO steve you will give me an e-mail? 
18:28:38  <[GregNoel](GregNoel)>     Maybe agenda in a spreadsheet, so we can note/comment in advance? 
18:29:07  <sgk>  [GregNoel](GregNoel):  good idea re: the spreadsheet, that's working well here 
18:28:52  <sgk>  Jason_at_Intel:  send me something on Monday to get the ball rolling 
18:28:53  <garyo>        Jason: it's still @ parts.tigris.org/... right? 
18:29:02  <Jason_at_Intel>       yep 
18:29:15  <garyo>        I have a note I'm supposed to review [ConfigToolsandChainNotes](ConfigToolsandChainNotes) there 
18:31:13  <Jason_at_Intel>       ya... reading the document  i have would be nice start 
18:31:22  <Jason_at_Intel>       even if it a little out of date ina few places 
18:29:07  <Jason_at_Intel>       big update coming 
18:29:29  <Jason_at_Intel>       major reporting /coloring/logging overhaual 
18:29:29  <sgk>  Jason_at_intel:  is Parts attracting many users? 
18:29:50  <Jason_at_Intel>       I get odd mails here and there .. have a few bugs 
18:29:43  <garyo>        sgk: I'd suggest a Wave but not everyone has it or knows how to use it 
18:30:13  <[GregNoel](GregNoel)>     A wave would be good, actually, maybe for the comments and notes? 
18:30:05  <Jason_at_Intel>       It is beta.. or not 1.x yet... so people are holding off 
18:30:03  <bdbaddog>     I have wave. 
18:30:08  <bdbaddog>     can invite everyone else.. 
18:30:09  <sgk>  garyo:  are you actively using wave? 
18:30:18  <garyo>        no, nobody else I know has it 
18:30:21  <sgk>  hmm 
18:30:29  <garyo>        a one-user wave is kinda lonely 
18:31:10  <[GregNoel](GregNoel)>     "Wave without a shore" 
18:31:03  <sgk>  i haven't had a good excuse to really try to use it 
18:31:37  <bdbaddog>     I don't see any of you guys in my wave contact list. 
18:32:04  <garyo>        how do I add you or vice versa? 
18:32:09  <Jason_at_Intel>       Is Wave any good? 
18:32:16  <bdbaddog>     I'm on as bdbaddog.at.gmail.com 
18:32:17  <garyo>        don't know yet 
18:32:29  <bdbaddog>     try adding me as a contact under wave. 
18:32:29  <Jason_at_Intel>       it seemed sort of in the works 
18:32:41  <garyo>        just did it 
18:32:52  <garyo>        I'm gary.oberbrunner.at.gmail.com 
18:32:54  <sgk>  bdbaddog:  have you been using it regularly?  how's the stability for you?  in which browser? 
18:33:04  <bdbaddog>     haven't been, Firefox.. 
18:33:41  <bdbaddog>     Grey, Jason, Steven, do you all have wave accounts? 
18:34:08  <sgk>  i do on my google account but not my external gmail account 
18:34:18  <[GregNoel](GregNoel)>     Grey===Greg?  No, not a permanent one 
18:35:02  <bdbaddog>     [GregNoel](GregNoel): oopsie... 
18:34:26  <Jason_at_Intel>       I need to be invited it says 
18:34:39  <bdbaddog>     email accouts for invites? 
18:35:00  *      sgk just realizes that his macbook pro is still running Firefox 2.0.x...  :-/ 
18:35:01  <Jason_at_Intel>       jasonk512.at.aol.com 
18:35:19  <bdbaddog>     aol.. really? ;) 
18:35:52  <garyo>        aol omg! 
18:35:28  <Jason_at_Intel>       it for the junk mail 
18:35:26  <bdbaddog>     just sent an invite 2 u. 
18:35:29  <sgk>  sweet 
18:35:40  <bdbaddog>     Steven - do you have an acct? 
18:35:54  <sgk>  bdbaddog:  sure, add sguireknight.at.gmail.com 
18:36:20  <bdbaddog>     says "user does not have a wave account" 
18:36:32  <sgk>  oh, sorry, still need invite 
18:36:53  <bdbaddog>     done. 
18:36:57  <bdbaddog>     sgk: sent. 
18:37:08  <bdbaddog>     though it says the invites won't be sent immediately. 
18:37:30  <garyo>        I guess I don't have any invites to give out :-( 
18:37:34  <bdbaddog>     [GregNoel](GregNoel): Did u neeed one? 
18:37:35  <[GregNoel](GregNoel)>     Try inviting gregnoel.at.tigris.org and see if it sends me one (eventually)... 
18:37:37  <bdbaddog>     I have 30 left. 
18:37:54  <bdbaddog>     [GregNoel](GregNoel): added to invite list.. 
18:37:58  <bdbaddog>     29 left. 
18:38:19  <garyo>        cool, this should be fun.  we'll learn something anyway. 
18:38:27  <bdbaddog>     k. I'm outta here for tonight. will be interesting to see if wave is useful for this. 
18:38:41  <garyo>        yup, me too, have to find out who won in MA special election. 
18:39:01  <bdbaddog>     [GregNoel](GregNoel): Let Steven and/or  I know the dates you'll be up if you'd like to get together when you're up.. 
18:39:22  <[GregNoel](GregNoel)>     bdbaddog, wilco 
18:39:03  <Jason_at_Intel>       so have i been added? 
18:39:07  *      sgk gets a shoeshine while waiting for the invite to show up 
18:39:47  <bdbaddog>     Gnight all! 
18:39:56  <sgk>  bdbaddog:  good night, thanx for the checkpoints 
18:40:14  <Jason_at_Intel>       bddaddog: night! 
18:40:23  <garyo>        see you folks on the ML... 
18:40:26  *      [GregNoel](GregNoel) has new mail... 
18:41:18  <[GregNoel](GregNoel)>     Hey, Rob is still around; he just corrected that spelling error 
18:41:40  <garyo>        [GregNoel](GregNoel): good to know. 
18:41:54  <garyo>        anyway, time for bed I guess, here on the benighted east coast. 
18:42:18  <garyo>        take care folks, see you soon 
18:42:25  <Jason_at_Intel>       if you get a chance gary.. read that document :-) 
18:42:25  *      garyo (n=[garyo@209-6-36-50.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com](mailto:garyo@209-6-36-50.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com)) has left #scons 
18:42:46  <sgk>  good night.  Geaux Saints! 
18:43:22  <Jason_at_Intel>       Night all.. till next time 
18:43:50  <[GregNoel](GregNoel)>     No invitation...  I'll keep an eye out. 
18:43:51  *      Jason_at_Intel has quit ("[ChatZilla](ChatZilla) 0.9.86 [Firefox 3.5.3/20090824101458]") 
18:43:57  <[GregNoel](GregNoel)>     G'night all. 
18:46:39  *      You have been marked as being away 
18:49:33  *      loonycyborg has quit ("Zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz") 
18:49:45  *      sgk (n=[sgk@c-71-204-129-67.hsd1.ca.comcast.net](mailto:sgk@c-71-204-129-67.hsd1.ca.comcast.net)) has left #scons 
18:51:22  *      bdbaddog (n=[bdeegan@adsl-71-131-11-120.dsl.sntc01.pacbell.net](mailto:bdeegan@adsl-71-131-11-120.dsl.sntc01.pacbell.net)) has left #scons 

Updated