IrcLog2008 06 02

16:23:54 * bdbaddog (n=[email protected]) has joined #scons 16:59:39 * jrandall (n=[email protected]) has joined #scons 17:00:32 <GregoryNoel> Steven and Gary have said they will be late; who else is here for the bug party? 17:00:52 <GregoryNoel> And Gary may not make it at all. 17:01:18 here, but after looking through the current list of bugs, there's not a lot I have to add to them 17:01:40 Had a hard time getting into 2007Q3. Any known problem with that spreadsheet? 17:02:08 <GregoryNoel> No, just the usual. Nobody has figured out the exact magic needed. 17:03:00 Hrm, I had it opened view-only in another tab, maybe that vexed it for some reason. I'll try again later to see if it likes me then 17:03:24 <GregoryNoel> Apparently, that's one no-no. 17:03:55 <GregoryNoel> Could you add that note to the ReadWrite page? 17:04:00 Sure thing 17:04:44 <GregoryNoel> Bill, are you there? Or was that an automatic connection? 17:06:40 <GregoryNoel> Apparently not. Only two isn't a quorum, but we can wait a bit and see if Steven or Gary show up. 17:06:48 Sure thing 17:06:48 * chit-chat while wating for quorum 17:12:54 * stevenknight (n=[email protected]) has joined #scons 17:13:12 hi, who's here? 17:13:23 <GregoryNoel> nobody 17:13:31 damn 17:13:39 aye, 'tis quiet 17:14:11 <GregoryNoel> Gary is caught at work and may not make it. 17:16:43 Hi All. I'm here til about 5:30ish. 17:17:00 hi bill 17:17:09 Good Day! 17:17:31 <GregoryNoel> Hey, Bill. 17:17:13 i just got connected myself, shall we dive into the current issues? 17:17:39 <GregoryNoel> Sure 17:17:38 2073: moot, already fixed 17:17:56 2074: consensus 2.x p2 17:18:14 2076: consensus 1.x p1 17:18:43 <GregoryNoel> Geeze, let me catch up. 17:19:20 <GregoryNoel> 2074, 2076, done 17:19:30 <GregoryNoel> 2077 17:19:37 oh, 2076: we should assign to someone, yes? 17:19:54 <GregoryNoel> Assign Bill 17:20:01 works for me 17:20:10 oh boy. imagine if I wasn't here.. ;) 17:20:17 2077: consensus 1.x 17:20:28 two votes (kind of) for p4, any objections? 17:20:43 <GregoryNoel> Er, 2077 assign Bill; I'll look at 2076 17:20:56 okay 17:20:57 <GregoryNoel> unless Bill wants it 17:21:35 nope. but I'll take a look at 2077. might be 2 weeks as I have trade show next week, before I get a chance. 17:21:45 done 17:21:47 <GregoryNoel> done 17:22:03 2078: 1.x, p2, me (along with other Visual Studio / VC work) 17:22:21 <GregoryNoel> ok 17:22:39 (the reprioritization after 1.0 is released is going to be fun...) 17:22:49 2079: 2.x, greg? 17:23:05 <GregoryNoel> Hmmm... OK, I guess 17:23:14 <GregoryNoel> what priority? 17:23:28 sounds like you have a handle on it 17:23:37 I don't quite grok why a File.Grep() method 17:23:54 it's like Glob() but with regular expressions.. 17:24:00 <GregoryNoel> It's not obvious when to use f.name and str(f) 17:24:09 as opposed to some more generic method that might also grep for Dir, Alias, Value... 17:24:27 <GregoryNoel> No, no, no, it looks at file contents. 17:24:39 ah 17:24:41 <GregoryNoel> Like a scanner. 17:24:50 oh. I thought from the emails, the requestor wanted to grep the file names,not contents. 17:25:10 yeah, like Filter (and FilterOut) in Ant 17:25:25 <GregoryNoel> No, he wanted to scan for 'int main(' to locate the main programs 17:25:28 o.k. never mind just reread. 17:25:44 <GregoryNoel> Maybe those are better names (FilterIn/Out) 17:25:55 well, they imply matching names, not file contents 17:26:02 <GregoryNoel> Ah, true. 17:26:05 yes. sounds clearer, Grep makes me think Glob but Regex. 17:26:13 i guess rather than add a special method (IMHO) 17:26:40 <GregoryNoel> (yes?) 17:26:47 i'm more interested in giving File nodes a read() method 17:26:57 <GregoryNoel> Hmmmm...... 17:26:58 that looks like normal Python file objects 17:26:59 ahh. I like that even more. 17:27:12 <GregoryNoel> I think I do, too 17:27:14 and then let people manipulate f1.read() using normal Python 17:27:41 <GregoryNoel> Yes, good idea. I'll write it up that way. 17:27:49 okay, thanks 17:27:57 <GregoryNoel> next? 17:28:16 2080: TASK 17:28:32 i forget, how are we marking items like this? 1.0 and just move them along? 17:28:39 i.e., things that can be done any time 17:28:42 <GregoryNoel> How about David as a release team member? 17:28:50 ++ 17:28:58 I think he said he didn't have enough time though. 17:29:03 <GregoryNoel> No, I make up something 17:29:48 ?? 17:29:48 <GregoryNoel> I don't think being on the mailing list would be a problem; I'd appreciate his insight for the spreadsheets. 17:30:07 agreed 17:30:12 sounds good. 17:30:40 <GregoryNoel> "make up something" === try to guess when it would be done; it's what the not-research items should be. 17:30:54 okay 17:31:17 2081: consensus 1.x p2 17:31:29 <GregoryNoel> I can create something for backburner issues, but "backburner" is not a name that delights me. 17:31:51 "backburner" to me would be implied by the priority 17:32:02 since the target milestone is really about timeframe 17:32:07 how about an explicit "anytime" 17:32:09 ? 17:32:20 <GregoryNoel> Hmmm.... I'll look at that 17:32:24 okay 17:32:34 <GregoryNoel> 2081: done 17:33:12 <GregoryNoel> 2082: split between p2 and p4 17:33:15 2082: i meant 1.x 17:33:18 <GregoryNoel> (both 1.x) 17:33:41 so 1.x, and p3? (split the difference) 17:33:47 Looks like just needs some tests to be able to be applied right? 17:33:56 Do we have much coverage on rc files? 17:34:18 not a lot 17:34:23 i was just dealing with rc file today 17:34:34 so i'd put my name on this one, too 17:34:39 <GregoryNoel> works 17:35:03 O.k I"m a pumpkin. I've gotta head to class. 17:35:04 2083: looks like consensus 1.x p2 17:35:08 Good evening to all. 17:35:10 later 17:35:13 * bdbaddog has quit ("Leaving.") 17:36:05 <GregoryNoel> 2083: yeah, but we need to talk about the model. 17:36:16 fire away 17:36:28 or did you mean on the ML? 17:36:25 <GregoryNoel> Maybe not right now, but there needs to be some agreement on how to do it. 17:36:45 okay 17:37:09 <GregoryNoel> ML would be fine; the last time I wrote a suggestion about it, it just died away, and I still don't have any real ideas 17:37:34 yeah, i may be the only one who cares about it in practice 17:37:44 purely because of wanting to do everything that Make does... :-) 17:37:46 <GregoryNoel> No, I do 17:37:58 no, i mean cares whether there is a mechanism that works 17:38:06 i think most people want it to just go away... :-) 17:38:05 <GregoryNoel> The real problem is less-than-clean removals 17:38:17 ah, right 17:39:08 <GregoryNoel> If it were only creating "cleaner" levels, it would be easy, but you want to be able to clean out, say, just the intermediate files 17:39:17 right 17:39:40 <GregoryNoel> I just don't have any good idea for how to do thatt. 17:39:45 so for now: 1.x, p2, and either you or I to lead discussion (even if it's just between the two of us)? 17:39:52 <GregoryNoel> works 17:40:03 either that or "research" since we're still not sure 17:40:23 your choice, 1.x or research 17:40:48 <GregoryNoel> 1.x; that'll force us to look at it at a specific time 17:40:52 good 17:41:02 2084: i'm clueless 17:41:07 <GregoryNoel> 2084, where's Gary? 17:41:17 we could make it research, garyo 17:41:29 just so he doesn't escape completely unscathed by not showing up... :-) 17:41:35 <GregoryNoel> I'll buy that! 17:41:43 done 17:41:49 lol 17:42:20 2085: 1.0, p4 (split difference), me 17:42:28 <GregoryNoel> done 17:42:53 i have doc changes teed up for once i get 0.98.5 out (I hope later this evening, this past weekend was overrun by daughter's birthday) 17:43:12 <GregoryNoel> Happy birthday; daughters are dangerous 17:43:30 oh my goodness yes 17:43:34 <GregoryNoel> how old? 17:43:39 10 17:43:44 <GregoryNoel> ouch! 17:44:05 <GregoryNoel> I remember my niece at ten.... oh, my, are you in for it! 17:44:25 yeah, I'm right on the cusp of going from being cool Dad to the biggest dork in the world 17:44:46 mind you, that last bit isn't much of a stretch... 17:45:20 anyway, 2007 q2? 17:45:32 <GregoryNoel> er, q3? 17:45:41 oh, right, q3 17:45:48 i was working ahead a little on q2 17:46:20 <GregoryNoel> 1869 17:46:51 ? 17:46:55 i have 1687 as the first? 17:47:01 <GregoryNoel> fixed 17:47:08 ah 17:47:39 1689: consensus 1.x, 17:47:55 p2? 17:48:14 <GregoryNoel> Another one that needs some discussion after a bit of research, but 17:48:27 <GregoryNoel> p2 is a reasonable time to do it. 17:48:27 right 17:48:43 <GregoryNoel> OK, done 17:49:08 assign to...? you (maybe ParseConfig), me (I might know what's going on), leave blank for now? 17:49:41 <GregoryNoel> blank, actually issues@scons 17:49:55 okay 17:50:09 <GregoryNoel> I don't think it was backtick 17:50:27 maybe not 17:50:04 1690: research, me (Visual Studio stuff) 17:50:40 <GregoryNoel> 1690, done 17:50:52 1691: documentation, 1.0, me 17:51:14 <GregoryNoel> done 17:51:28 1692: research, me (Visual Studio again) 17:51:29 <GregoryNoel> may need to follow up to see what the message was 17:51:40 <GregoryNoel> 1692, done 17:51:52 1693: consensus 1.x p2 17:52:08 <GregoryNoel> done 17:52:09 good manageable bug for someone else to take 17:52:15 <GregoryNoel> yes 17:52:28 1697: research, me (Visual Studio) 17:52:47 <GregoryNoel> okay 17:53:16 <GregoryNoel> 1701, ditto 17:53:17 1701: research, me (Visual Studio) 17:53:19 right 17:53:20 <GregoryNoel> done 17:53:34 it isn't the pipes thing, it has to do with how we look in the #*@&(#$ registry for various info 17:53:56 1702: same... 17:54:02 <GregoryNoel> 1702, ditto 17:54:10 man, there's a lot of Visual Studio cruft piling up 17:54:26 I'm really itching to get in there and clean this stuff up 17:54:50 1703: 17:54:52 <GregoryNoel> Do you want a keyword for it? I can set it up, but you'll have to assign them all. 17:54:56 not sure about my 1.x p3 17:55:04 keyword: yes 17:55:17 "VisualStudio" seems logical 17:55:30 <GregoryNoel> do you mean 1704? 17:55:39 oh, yes 1704: 17:56:00 1704: seems like if it were really crucial more people would have asked for it 17:56:07 i only know of this one patch 17:56:19 on the other hand, it kind of goes along with what Russel was saying on the ML today 17:56:33 about how SCons really doesn't have much traction in the Java community 17:56:35 <GregoryNoel> There was something on the mailing list about JAR() recently, maybe today? 17:56:46 <GregoryNoel> oops, you already said that 17:56:47 yeah, Russel's threads 17:57:00 let's leave it p3 17:57:06 since there's already a patch 17:57:07 <GregoryNoel> OK 17:57:25 if we ever are going to do better with Java, it can't hurt to have this already supported 17:57:32 <GregoryNoel> Maybe draft a Java specialist to keep us on track 17:57:50 <GregoryNoel> Maybe Russel? 17:57:58 maybe 17:58:09 he tends to appear and reapper in fits and starts 17:58:14 disappear i mean 17:58:47 <GregoryNoel> I'll write him about creating a wiki page with what's needed for Java support 17:58:56 hmm, i thought i recalled there was someone else who showed up on the ML with some Java knowledge a month or two ago 17:59:09 maybe i'm making that up 17:59:22 well, it can't hurt to ask, anyway 17:59:24 <GregoryNoel> No, I have his name 17:59:37 <GregoryNoel> I'll ask them both 17:59:43 good idea re: wiki page 17:59:49 sounds good 18:00:35 <GregoryNoel> anyway, what did we decide about 1704? 18:01:04 <GregoryNoel> 1.x, p2, you? 18:01:21 done 18:01:58 1705: 1.x, jim ... p3? 18:02:05 <GregoryNoel> or p2 18:02:09 Aye - I've got a patch in that fixes it 18:02:24 jrandall++ 18:02:32 <GregoryNoel> bravo! 18:02:47 <GregoryNoel> p2 then? 18:02:48 thanks. 18:02:52 yeah, p2 18:02:55 <GregoryNoel> done 18:03:23 1706: 1.x, but now i'm not sure of priority 18:03:50 <GregoryNoel> I'll look at it, maybe p4? 18:04:02 <GregoryNoel> It's part of getting symlinks right. 18:04:16 sure, 1.x, p4, you 18:04:21 <GregoryNoel> done 18:04:44 1707: consensus 2.x p4 18:04:53 <GregoryNoel> done, or future? 18:05:26 hmm, i'm torn 18:05:38 part of me says future because no one seems to have asked for it 18:05:47 but maybe 2.x because there's already code 18:06:06 <GregoryNoel> Yeah, but infected 18:07:12 <GregoryNoel> Let's leave it at 2.x p4 and revisit later 18:07:24 okay 18:07:31 <GregoryNoel> 1708, I'll go with Ken to look at it. 18:08:00 1708: okay 18:08:15 I may need to take it back if he doesn't pop up again 18:08:26 but we should at least see if he can take it 18:08:27 <GregoryNoel> If he doesn't like it, he can kick it elsewhere. 18:08:32 yeah 18:08:38 <GregoryNoel> I'll put that in the note. 18:08:43 okay 18:09:14 <GregoryNoel> 1711, quite a mix 18:09:18 1711: yeah 18:09:30 when in doubt, shade to the earlier target 18:09:37 <GregoryNoel> Huh? 18:09:46 <GregoryNoel> Oh, I see. 18:09:55 i tend to go with the earlier/earliest milestone 18:10:26 i'd rather make sure it gets considered and reprioritize to later if necessary 18:10:27 <GregoryNoel> Let's make it 1.x then and give it to Gary, since he's not here 18:10:48 <GregoryNoel> p3? 18:10:53 ah, good idea -- he's done subst stuff 18:10:54 yes, p3 18:10:58 <GregoryNoel> done 18:11:17 1712: 2.x, p3 18:11:27 perhaps Benoit if we want to assign it 18:11:32 he's good at things like this 18:12:00 <GregoryNoel> I'd want measurements. I don't think scanners are that slow. 18:12:19 good point, they're probably not 18:12:26 Aye. Not clear where the tradeoff would be as to whether it'd be worth it or not 18:12:37 Most of mine, it wouldn't be worth spawning 18:12:46 actually, (off topic) i have an optimization i'm thinking of that I'd like to discuss with you some time 18:13:02 let's get through bugs first though 18:13:19 <GregoryNoel> In fact, I think a small rewrite so that scanners overlap with the previous command would cure it. 18:13:36 <GregoryNoel> I do that in TaskmasterNG 18:13:42 nice 18:13:43 oh, very cool 18:13:48 simple and effective 18:14:14 1714: 1.x, p3 18:14:15 <GregoryNoel> Is that the optimization? 18:14:59 no, it's basically trying to make searching CPPPATH O(1) instead of O(n) 18:15:06 <GregoryNoel> 1714, agreed, but spin off JAR to another issue 18:15:16 1714: agreed 18:15:26 1.x, p3, garyo 18:15:38 could also go to Russel or whoever gets to be Java guru 18:15:53 <GregoryNoel> done; I'll note that 18:16:13 <GregoryNoel> OT: yes, they should be hashed better. 18:16:16 good 18:16:43 OT: actually, even beyond that, the search is attached to the wrong object 18:16:51 <GregoryNoel> 1717, you, VS 18:17:07 <GregoryNoel> OT: yes, I've noticed that 18:17:08 1717: yes 18:17:15 <GregoryNoel> done 18:17:41 1722: it's Bill's, let's WONTFIX it... :-) 18:17:52 <GregoryNoel> 1720, has Bill left? 18:18:05 <GregoryNoel> oops, 1722 18:18:10 yeah he's gone 18:18:19 so he gets what he deserves... :-) 18:18:27 <GregoryNoel> OK, WONTFIX unless he provides a test case 18:18:37 done 18:19:00 1723: can this be part of the toolchain stuff you and Gary have on the backburner? 18:19:10 <GregoryNoel> yes 18:19:15 oh, yeah, your comment even says that... 18:19:21 <GregoryNoel> yup 18:19:32 future, you? 18:19:36 <GregoryNoel> done 18:19:51 1730: 1.x, p3, Rob? 18:20:33 <GregoryNoel> Uh, maybe not Rob 18:21:11 <GregoryNoel> Oops, I was thinking of another issue; yes, Rob. 18:21:19 okay 18:21:27 <GregoryNoel> It's a little out of his area, but he can work with you. 18:21:56 okay 18:22:21 1735: research, Rob? 18:22:30 <GregoryNoel> 1735, what if it's still a bug? Where to put it? 18:22:56 I'm agnostic -- 1.x p3? 18:23:22 <GregoryNoel> works; I'll tell him to contact me if he needs to 18:23:27 done 18:23:39 1716: research, me, VisualStudio 18:23:57 <GregoryNoel> done; quit for the evening? 18:24:02 yeah, i have to run 18:24:06 real quick re: CPPPATH 18:24:06 * off-topic discussion between stevenknight and GregoryNoel 18:30:36 okay, really gotta run 18:30:39 thanks! 18:30:39 <GregoryNoel> When shall we all meet again? 18:30:39 <GregoryNoel> In thunder, lightning, or in rain? 18:30:39 <GregoryNoel> Where the place, ... same time next week? 18:30:52 oh, damn, that's right 18:30:57 yes, default, same time and place 18:31:01 <GregoryNoel> done; cu 18:31:05 l8r 18:31:06 * stevenknight has quit ("Leaving") 18:31:07 see you 18:31:09 * jrandall (n=[email protected]) has left #scons