1. SCons
  2. Core
  3. SCons

Wiki

Clone wiki

SCons / BugParty / IrcLog2008-12-17

17:29:34  *      sgk_ (n=[stevenkn@67.218.110.27](mailto:stevenkn@67.218.110.27)) has joined #scons 
17:31:20  <sgk_> hey greg 
17:31:39  <Greg_Noel>    Hi, Steven. 
17:32:55  <sgk_> wiki down in general, or is it just me? 
17:33:20  <Greg_Noel>    Not just you. 
17:33:21  <Hydrant>      I noticed that too 
17:33:45  <sgk_> damn, i've grown dependent on the links in the [BugParty](BugParty) page 
17:34:19  <Greg_Noel>    I can post some... 
17:34:27  <sgk_> thanks, that'd help 
17:34:37  <sgk_> i've had no chance to pre-review, unfortunately... :-( 
17:35:17  <sgk_> is there anything i can/should look at over at pair.com w.r.t. trying to resurrect the wiki? 
17:35:34  <sgk_> (says the complete novice...) 
17:35:51  <Greg_Noel>    Dunno.  I was wondering if Gary did something. 
17:36:07  <Greg_Noel>    (The spreadsheet is [http://spreadsheets.google.com/ccc?key=p1VsJ6ACdIhAs91YGwSnBYw&hl=en](http://spreadsheets.google.com/ccc?key=p1VsJ6ACdIhAs91YGwSnBYw&hl=en)) 
17:37:02  <Greg_Noel>    Sohail, you here for the bug party? 
17:37:33  <sgk_> Greg_Noel:  thanks, starting in 
17:40:24  <sgk_> 1910:  me, research 
17:40:27  <sgk_> figure out if it's a real bug 
17:40:31  <Greg_Noel>    concur 
17:40:48  <sgk_> (yeah, I know, we don't quite have a quorum, but what the hey...) 
17:41:00  <sohail>       Greg_Noel, no, just hanging out sorry 
17:41:14  <Greg_Noel>    (Maybe Gary will show up eventually) 
17:41:39  <Greg_Noel>    sohail, if you have an opinion, don't hesitate to speak up. 
17:42:21  <sohail>       Greg_Noel, will do 
17:43:31  <sgk_> 2249:  weird weird weird 
17:43:36  <sgk_> needs some research 
17:43:47  <Greg_Noel>    yeah, for sure 
17:43:47  <sgk_> smells like a possible drive letter issue to me 
17:43:47  <Greg_Noel>    I hadn't thought of drive letters. 
17:44:41  <sgk_> put my name on 2249 too 
17:44:52  <sgk_> i'm actually buried in other Windows stuff now anyway 
17:44:55  <Greg_Noel>    ok, done.  I'm glad it's not me! 
17:45:40  <sgk_> :-) 
17:46:05  *      sgk_ curses the network lag on the shuttle... 
17:48:10  <sgk_> 2084:  looks to me like a future thing (2.x or later) and not terribly high priority 
17:48:18  <sgk_> David has a workaround for his situation 
17:48:25  <sgk_> and no one else has been beating down the doors for it 
17:48:55  <Greg_Noel>    Priority? 
17:49:54  <sgk_> p4 works for me 
17:49:57  <Greg_Noel>    2084, future, draft pick, p4, done 
17:50:36  <sgk_> 2190:  would this be easy to drop into the current framework? 
17:51:35  <Greg_Noel>    No worse than many others.  Is the short-term tradeoff worth the time we could spend on a replacement? 
17:51:54  <Greg_Noel>    And there are workarounds. 
17:51:55  <sgk_> yeah, probably not 
17:52:13  <sgk_> future, mark it with some keyword for Configure replacement 
17:52:21  <Greg_Noel>    works for me 
17:52:29  <sgk_> i.e., agree w/Greg  :-) 
17:52:55  <Greg_Noel>    Any ideas for a keyword? 
17:53:10  <sgk_> configure ? 
17:53:49  <Greg_Noel>    I'd rather not; too vague; could be something for existing configure. 
17:53:59  <sgk_> good point 
17:54:18  <Greg_Noel>    Oh, and a priority is needed 
17:54:34  <sgk_> p4 since there are workarounds 
17:54:38  <sgk_> or p3 
17:54:08  <sgk_> configure_revamp ? 
17:55:00  <Greg_Noel>    Longish.  I'd go for something like ConfigNG first. 
17:55:34  <Greg_Noel>    p3 works for me for now.  I'd like to see it sooner, but there's so much to do. 
17:55:57  <sgk_> right 
17:56:38  <sgk_> 2190:  future, p3, +configure_revamp Keyword 
17:56:46  <Greg_Noel>    done 
17:57:08  <sgk_> 2196:  agree with both Russel and you in principle 
17:57:16  <sgk_> but it's a big architectural change 
17:57:30  <Greg_Noel>    Yeah, that's why I think future. 
17:57:37  <sgk_> agree w/future 
17:58:12  <Greg_Noel>    On the other hand, there are some features in Python 2.2 that may help, so I could be convinced otherwise. 
17:58:00  <sgk_> I'm working in the background on a list of big architectural changes to tackle in 2009 
17:58:12  <sgk_> e.g. DAG with real edge objects 
17:58:27  <Greg_Noel>    Hmm... 
17:58:57  <Greg_Noel>    Why would you need the arcs articulated? 
17:59:01  <sgk_> i still say future, with the intent of discussing it at a prioritization session for big architecture work 
17:59:14  <Greg_Noel>    so, needs a keyword 
17:59:22  <sgk_> right 
17:59:37  <sgk_> "architecture" or "arch_revamp" ? 
17:59:48  <Greg_Noel>    rearchitecture? 
17:59:53  <sgk_> +1 
18:00:15  <Greg_Noel>    so future, p2, done 
18:00:20  <sgk_> done 
18:00:23  <sgk_> re: DAG edges 
18:00:41  <sgk_> one, i've finally caught up with your good advice about being able to do independent topo sorts and the like 
18:00:56  <sgk_> more immediate and practically, i'd like to be able to walk the DAG in the other direction 
18:01:04  <sgk_> i think that would help developer builds 
18:01:30  <sgk_> because you could say "I know foo.c is the only thing that changed, build just its targets" 
18:01:37  <Greg_Noel>    It's possible to build the transverse graph as you go, but it's slower. 
18:01:57  <sgk_> but you'd have to walk it to do so, yes? 
18:02:27  <sgk_> maybe i'm naively assuming that separating the edges 
18:02:46  <sgk_> would provide benefit from being able to use more sophisticated algorithms 
18:03:12  <sgk_> than our current tweaked-into-behaving-kinda-how-we-want walk 
18:04:31  <Greg_Noel>    I can agree with that, but building the transverse graph is cheap relative to what else you have to do.  It's probably not a lot different than building it as you go. 
18:05:47  <Greg_Noel>    And you still have to be able to create arcs for implicit dependencies, so the targets of foo.h may not be known until you walk up. 
18:04:40  <sgk_> okay 
18:04:52  <sgk_> gary just fixed the wiki -- deprecation warnings 
18:05:00  <sgk_> sent email that he won't be able to make it 
18:05:40  <sgk_> onward? 
18:05:54  <Greg_Noel>    yeah 
18:07:06  <sgk_> 2261:  2.x p3 
18:07:50  <Greg_Noel>    BTW, disclaimer for this evening: We've had almost four inches of rain in the last couple of days, and the local power grid seems to be unstable.  We've had three power trips today, all short enough so that the only thing that rebooted was our firewall, but if I suddenly stop, figure that I've lost power. 
18:08:14  <sgk_> Greg_Noel:  okay, thanks for the heads up 
18:08:19  <sgk_> good luck... 
18:07:59  <sgk_> mpcomplete is a googler whose initials are "MP" 
18:09:33  <Greg_Noel>    Tell him he needs to change his name: MP-complete has been demonstrated to be NP-hard. 
18:09:51  <sgk_> :-) 
18:10:23  <sgk_> 2268:  2.x p3 draft pick 
18:10:31  <Greg_Noel>    2261, okay. 
18:11:20  <Greg_Noel>    2268, ok, may need to revisit. 
18:11:57  <Greg_Noel>    Agree that stack trace is not a good thing. 
18:12:39  <sgk_> 2269:  agree, wontfix 
18:12:44  <Greg_Noel>    done 
18:12:56  <sgk_> probably should document the API if we expect him to reach into the Executor, though 
18:13:45  <Greg_Noel>    str(Executor) probably gets him what he wants; I'd rather not tie down the API 
18:14:31  <sgk_> Greg_Noel:  good 
18:14:35  <sgk_> i like it 
18:14:56  <sgk_> 2270:  me, 2.x, [VisualStudio](VisualStudio) keyword 
18:15:01  <sgk_> actually, 1.3 
18:15:08  <sgk_> i'm integrating vs_revamp 
18:15:34  <sgk_> i don't think this part is affected by that, but i'll double check and do the right thing 
18:15:49  <Greg_Noel>    ok, more issues for you to resolve. 
18:15:55  <sgk_> (or maybe vs_revamp keyword...?  whatever's appropriate) 
18:16:50  <Greg_Noel>    Probably vs_revamp, but I don't know the distinction. 
18:17:13  <sgk_> either way 
18:17:10  <Greg_Noel>    Priority? 
18:17:15  <sgk_> p3 
18:17:23  <Greg_Noel>    done 
18:17:56  <sgk_> 2270:  me, 1.3, p3 
18:18:04  <sgk_> part of integrating again 
18:18:15  <sgk_> sorry 
18:18:19  <Greg_Noel>    good; not me 
18:18:20  <sgk_> 2271:  me, 1.3, p3 
18:18:31  <Greg_Noel>    right 
18:20:03  <sgk_> 2272:  ouch 
18:20:18  <Greg_Noel>    you got that right. 
18:20:29  <sgk_> good analysis 
18:20:42  <Greg_Noel>    yeah, we should recruit him 
18:20:55  <sgk_> good idea 
18:21:21  <sgk_> 2.x p3 draft choice 
18:21:28  <sgk_> and see if we can sign him up as the draft choice? 
18:22:04  <Greg_Noel>    ok, with some reference to TaskmasterNG, that may complicate this even further. 
18:22:15  <sgk_> right 
18:23:00  <sgk_> 2273: 
18:23:20  <sgk_> anytime, anyone 
18:23:42  <Greg_Noel>    anytime only works if there's someone assigned 
18:23:56  <sgk_> good point 
18:24:06  <Greg_Noel>    I think we should put out a RFE 
18:24:15  <sgk_> E == editor? 
18:24:16  <Greg_Noel>    Request For Editor 
18:24:16  <sgk_> concur 
18:24:46  <sgk_> 2273:  anytime, Greg_Noel? 
18:25:08  <Greg_Noel>    It's DOS, so I'd probably botch it. 
18:25:15  <sgk_> okay 
18:25:20  <sgk_> 2273:  anytime, stevenknight 
18:25:47  <Greg_Noel>    maybe we should give it to Gary as punishment for not being here. {;-} 
18:25:55  <sgk_> i LIKE it... :-) 
18:26:17  <Greg_Noel>    done 
18:26:15  <sgk_> ~5 minutes to bus stop 
18:26:40  <Greg_Noel>    Too bad; I had some other things I wanted to discuss. 
18:26:41  <sgk_> 2274:  1.x p3, see if arve comes up with the patch he mentions 
18:27:16  <Greg_Noel>    Maybe we should see if Gary&others could handle a special meeting? 
18:28:05  <Greg_Noel>    2274, ok 
18:28:13  <sgk_> yes re: special meeting 
18:28:24  <sgk_> 2275:  2.x p3 stevenknight +[VisualStudio](VisualStudio) keyword 
18:28:38  <sgk_> if not part of vs_revamp, part of other stuff I'm working on for Google Chrome 
18:28:55  <Greg_Noel>    Looks like #ERROR! to me... {;-} 
18:29:09  <Greg_Noel>    2275, done 
18:29:32  <Greg_Noel>    (put a space in front of the plus) 
18:29:55  <sgk_> 2276:  stevenknight 
18:30:01  <sgk_> yeah, i thought i fixed it already 
18:30:39  <sgk_> 2277:  stevenknight +[VisualStudio](VisualStudio) 
18:30:47  <sgk_> last minute or so 
18:31:01  <Greg_Noel>    2276, okay, I'll presume you'll close it. 
18:32:12  <sgk_> 2276:  i'll close 
18:31:10  <Greg_Noel>    2277, done 
18:32:18  <sgk_> 2278:  recruit swig guru 
18:32:19  <Greg_Noel>    2278 can wait, but I think I'll contact him and see if he can be recruited. 
18:32:25  <sgk_> +1 
18:32:47  <sgk_> okay, gotta go 
18:32:47  <Greg_Noel>    Surprisingly, that's all 
18:32:55  <sgk_> many thanks for the work 
18:33:01  <Greg_Noel>    I'll drop a line about another meeting; cul 
18:33:15  <sgk_> l8r 
18:33:17  *      sgk_ has quit ("Leaving") 

Updated