IrcLog2008 08 11

19:01:05 <GregNoel> Hey, it's time. We three seem to be here; anyone else for the bug party? 19:01:11 So... is it just the 3 of us tonight? 19:01:52 hmm, no sign yet of Brandon or Bill 19:03:16 anyway, shall we dive in to the current issues? 19:03:20 Yes please. 19:03:23 <GregNoel> I'm ready 19:03:40 <GregNoel> 2124, no Brandon... 19:03:58 2124: defer 19:03:58 Race conditions: Steven, do they go away w/o a virus scanner? 19:04:08 not sure, haven't had a chance to check 19:04:12 ok. 19:04:21 leave 2124 w/Brandon 19:04:34 I may end up diving in to solve my own problems 19:04:35 <GregNoel> 2124, defer but if Brandon doesn't show up, we make it research 19:04:45 * bdbaddog (n=bdeegan@adsl-71-131-3-114.dsl.sntc01.pacbell.net) has joined #scons 19:04:52 <GregNoel> Hey, Bill 19:04:55 Hi Bill. 19:04:59 but given how much i have going, don't want my name on it to prevent others from making progress 19:05:07 <GregNoel> concur 19:05:08 Good evening! 19:05:10 hey bill 19:05:17 We're just getting going. 19:05:22 1500: 19:05:35 wontfix. 19:05:36 <GregNoel> wontfix, steven 19:05:38 i'll WONTFIX like i said i would 19:05:58 ok. 2161 19:06:00 2161 19:06:14 I think this should be low pri. 19:06:20 i'm agnostic, my 1.x p3 is partly real, partly just to re-evaluate it later 19:06:30 agree re: low pri 19:06:40 <GregNoel> I'd push it straight to future 19:06:45 Someone who wants it & understands it needs to drive it. 19:06:45 bill, you object to 1.x p4? 19:06:53 or future? i could go there 19:07:07 nope that's fine. 19:07:16 ok. 19:07:19 <GregNoel> what pri? 19:07:31 p4? 19:07:34 <GregNoel> I'd suggest p2 or p3 19:07:35 p4 19:07:37 p4 19:07:42 <GregNoel> p4 it is. 19:07:54 <GregNoel> 2159 19:08:00 2159: greg, 1.0.1 19:08:17 <GregNoel> consensus 19:08:31 done 19:08:33 2160: 19:09:02 <GregNoel> Who's an AIX guru? 19:09:14 not me, sorry. Ask on the list? 19:09:23 not me. 19:09:27 <GregNoel> and not me 19:09:57 <GregNoel> Asking on the list might yield nothing; what do we do if noone shows up? 19:10:14 ask original poster for a patch, I guess 19:10:22 that sounds good 19:10:26 <GregNoel> Er, that would be me. 19:10:35 then it sounds doubly good! 19:10:36 oops. 19:10:44 :-) 19:11:00 Then I guess we guess or else do nothing. 19:11:08 <GregNoel> I just noticed the problem when I was looking at GetBuildPath for our last party 19:11:30 <GregNoel> and wondering what the implications of deprecating it would be 19:11:44 just a guess from looking at the code, it looks like it should be an argument of '.' 19:12:10 give it to me, i think i can figure out something semi-reasonable to do 19:12:22 <GregNoel> OK, low priority? 19:12:25 most likely involving getting rid of the GetBuildPath() call in favor of Dir() 19:12:27 yes, p4 19:12:43 <GregNoel> or even 2.0? 19:12:49 OK, but without an actual AIX user it won't be much good. p4 1.x or 2.x 19:12:56 right 19:12:58 <GregNoel> 2.x p2 19:13:03 done 19:13:12 2162: consensus 1.0.1 p3 gregnoel 19:13:25 <GregNoel> done 19:13:31 2163: 1.0.x, p?, who? 19:13:38 <GregNoel> (my google connection just crashed?) 19:13:43 mine too! 19:13:53 ditto... 19:14:17 gmail and google apps for domains were out for a bit mid day today. 19:14:25 not sure what's going on today, i heard gmail had problems that caused a stir 19:14:32 <GregNoel> Ah, it's back... 19:14:36 ditto 19:14:39 2163: since none of us can test it it doesn't matter who puts in the patch. I can do it. 19:14:45 garyo++ 19:14:50 <GregNoel> done 19:15:05 But I'm not the borland guy :-/ 19:15:11 just saying. 19:15:29 2164: same thing. I'll take that too. 19:15:37 2164: garyo, thanks 19:15:46 <GregNoel> good 19:16:12 <GregNoel> 2165 19:16:13 2165: leave w/me, or combine with the other race condition, or? 19:16:22 2165: Steven, can you merge that into the other race cond ticket? 19:16:26 sure 19:16:48 ok. 2166: 1.0.1 greg p2? 19:17:05 done 19:17:07 <GregNoel> done 19:17:24 2167: i'd like 1.0.x 19:17:34 probably 1.0.1 19:17:44 transition to subprocess and kill our home-brew stuff asap 19:17:40 <GregNoel> it's not a regression 19:17:49 sigh -- good point 19:18:00 <GregNoel> I agree that it's important, but I'd make it 1.x p2 19:18:05 would you be ok just integrating this patch or should we do the spawn -> subprocess switch all at once? 19:18:05 1.1 then? 19:18:12 yes, 1.x p2. 19:18:15 sure 1.1. 19:18:18 sounds good. 19:18:21 done 19:18:30 i'm okay with just the patch -- incremental steps are fine 19:18:31 <GregNoel> I'm willing to convert to subprocess incrementally 19:18:37 ok, good. 19:18:46 <GregNoel> who? 19:18:46 1.x p2, who? 19:18:49 Then I'll do it. 19:18:52 okay 19:18:53 <GregNoel> done 19:19:15 2168: anytime, p2, stevenknight 19:19:20 <GregNoel> done 19:19:31 <GregNoel> but not until 1.0 is out... 19:19:35 right 19:19:31 2169: 1.0.x, p3, garyo 19:19:37 2169: I'll take it. 19:19:44 <GregNoel> done 19:19:57 2170: 1.0.x, p2, stevenknight 19:20:06 <GregNoel> done 19:20:21 2171: 1.0.x, p3, stevenknight 19:20:44 <GregNoel> ok, and depends on 2170 19:20:50 right 19:20:58 2172: let bill and greg arm-wrestle for it... 19:21:05 anytime p2 19:21:31 <GregNoel> I'll do it 19:21:38 google's down again 19:21:54 <GregNoel> I'm still up, although I had a flash a few minutes ago 19:22:02 i'm still up too. 19:22:08 ok, that's the current sheet anyway. 19:22:17 <GregNoel> Oops, I just went down 19:22:24 Can I ask quickly about schedule? 1.0 is soon, then what? 19:22:38 <GregNoel> 1.0.1 19:22:52 three-four weeks later, i should think 19:23:02 <GregNoel> two weeks to a checkpoint, then push it out a week later 19:23:18 OK, so we start work on those tickets as soon as 1.0 is out and see how far we get. 19:23:38 Greg: I like that schedule. I'll be out for a week in mid-Sept though for work. 19:23:46 <GregNoel> I figure we clear about one issue a day between us, so no more than 15-20 issues in 1.0.1 19:24:00 Probably right. 19:24:31 I can do more in a burst but then real life intrudes for a while. 19:24:46 <GregNoel> Should we talk about the Python floor while waiting for Google to return? 19:24:46 agreed 19:25:10 My google is back. But I think 2.3 is OK. 19:25:26 garyo, if i can actually push out 1.0 tonight or tomorrow, do you have cycles to help w/re-branching this week? 19:25:30 We can backport fixes to 1.0 for old python users if they scream. 19:25:54 Steven: not really, but I'll make some. It's important. 19:25:54 <GregNoel> And the Olympics will conflict for the next two weeks, so fixing issues will be slower; my TiVo is recording 20 hours a day 19:26:08 what's in 2.3 that's really crucial / helpful for internal coding? 19:26:15 I'm olympics agnostic, so it doesn't affect me at all. 19:26:26 ditto (re olympics) 19:26:42 <GregNoel> The built-in @ stuff on functions 19:26:52 decorators? 19:26:57 <GregNoel> yes 19:27:24 <GregNoel> It was a royal pain trying to backport tarfile to 2.2, since it uses a lot of decorators 19:27:35 <GregNoel> and I never did get all of the unused stuff to work 19:27:56 i see 19:28:33 absent data to the contrary, not sure if tarfile itself is all that compelling, though 19:28:34 * garyo-home is looking at release notes 19:29:45 <GregNoel> I was going to research it before the meeting, but when did 2.2 and 2.3 come out? I think 2.2 is 2003, so that's five years ago. 19:29:55 sounds about right 19:30:16 oh, no 19:30:25 2.2 was 21 December 2001 19:30:35 2.3 was 29 July 2003 19:30:56 <GregNoel> OS X is the system fielding the oldest Python I'm aware of, and it had 2.3 in 10.4; 10.5 has 2.5 19:31:22 <GregNoel> so 2.3 is what I was recalling; that's already a bit old. 19:32:07 sets, generators, and a bunch of module changes are what I see in the relnotes. 19:32:07 how about we poll the ML? 19:32:08 RH3 has 2.2.3 19:32:33 <GregNoel> I think 2.3 is very reasonable, particularly if we don't release 2.0 until 2009; 2.3 will already be almost six years old 19:32:47 @bdbaddog: Really? We actually use that. But we may have upgraded the python on it. 19:32:59 <GregNoel> will RedHat be updated before 2009? 19:33:13 GregNoel: doesn't matter 19:33:16 Greg: RH3 is already ancient. 19:33:34 the issue isn't what the cutting edge has, it's how much of the old stuff still exists in enterprise development shops and on user sites 19:33:51 we have some time before deciding, so i'd suggest we poll the ML 19:34:02 RH is up to 8 or 9 now. 4 is considered minimal standard by our customers (it has at least 2.3 I'm sure) 19:34:04 and also follow-up with a reporting module in SCons that lets people send data back to us 19:34:06 if they choose 19:34:09 rh4 is 2.3.4 19:34:34 RH5 is 2.4.3 19:34:53 RH4 is the mainstream in EDA now. 19:34:59 with some on RH5 19:35:24 Our customers are visual effects people, RH4 is mainstream there. Little RH3 left. 19:35:35 garyo-home: little or none? 19:36:08 little. Still some diehards. But shouldn't use that as a guide, they're not going to run scons. 19:36:41 Polling the ML seems like a good idea. 19:37:02 right 19:37:11 Can we start into the 2006h1 spreadsheet? 19:37:12 should that be a TASK issue so it doesn't get lost? 19:37:12 yes. For me and all the clients I've ever had building new python,perl, whatever for use by build system hasn't ever been an issue, though sometimes it had to be staged. 19:37:18 <GregNoel> I'd suggest setting the warning floor to 2.3 and see if anyone makes noise 19:37:24 i'll open it 19:37:30 Sounds good. 19:37:32 stevennight: good. 19:37:34 GregNoel: good idea 19:37:35 I'm off for a bit back in a while. 19:37:59 <GregNoel> Google is back; I have 2006H1 up 19:38:17 me too. 1307 is the first one. 19:38:44 <GregNoel> only 20 issues are still live 19:38:53 1307: anytime, p2, Ludwig? 19:39:03 ok w/ me 19:39:14 <GregNoel> done, although GSoC is almost over 19:39:28 * stevenknight nods 19:39:33 1308: FIXED 19:39:37 Hopefully we can rope him in as a regular contributor. 19:40:00 that'd be cool 19:40:03 1310: WONTFIX 19:40:07 <GregNoel> yes 19:40:12 yes. 19:40:35 1318: WONTFIX? 19:40:39 actually, doc the solution 19:40:40 agreed. 19:40:45 yes, w/ doc. 19:40:48 <GregNoel> done 19:41:09 <GregNoel> 1320 19:41:10 1320: WORKSFORME 19:41:17 agreed. 19:41:17 <GregNoel> done 19:41:36 1323: anytime 19:41:36 <GregNoel> 1323, was just on the mailing list... 19:41:50 oh, yeah 19:42:03 <GregNoel> wait a sec; let me look 19:42:40 <GregNoel> axel_0001@gmx.de Axel W. 19:42:54 shall we assign it to him and see if that gets him more involved? 19:43:06 Interesting idea. 19:43:18 <GregNoel> I'm for it; he must have a Tigris ID 19:43:20 it's a future thing anyway 19:43:24 If he drops it we're no worse off than today. 19:43:28 right 19:43:29 <GregNoel> concur 19:43:33 greg, could you follow up w/him? 19:43:37 <GregNoel> yes 19:43:47 <GregNoel> so Axel anytime? 19:43:50 done 19:43:58 1328: I can do that one I think. 19:43:59 1328: 1.x p4 garyo 19:44:04 <GregNoel> done 19:44:19 1331: FIXED 19:44:36 1332: 1.x, p2, who? 19:44:56 I think you have to do that one, Steven. 19:45:12 <GregNoel> I don't know the code at all, so I agree 19:46:06 <GregNoel> hello? 19:46:19 silence... 19:46:38 Maybe Steven's looking at the code? 19:46:47 sorry, other distraction 19:46:55 <GregNoel> it happens 19:47:04 agree i don't have to do 1332 19:47:11 sorry, have to do 19:47:19 <GregNoel> done 19:47:24 1332: 1.x, p2, sk 19:47:46 1344: 1.x, p3, mati 19:47:51 Mati's looking into 1334, not sure of his status on it. 19:47:56 sorry 1344 19:48:09 <GregNoel> anytime, mati? 19:48:21 <GregNoel> it's GSoC 19:48:28 Yes. 19:48:29 i'm okay w/anytime 19:48:40 <GregNoel> 1345 19:48:57 1345: 2.x, p3, defer who 19:49:07 good idea. 19:49:30 <GregNoel> ok, I barely remember it. 19:49:52 1346: anytime, mati 19:49:54 no, I meant good idea not to pick someone now :-) 19:50:04 :-) 19:50:08 yes. I need to catch up w/ him since my vacation. 19:50:16 <GregNoel> yes 19:50:43 1350: 1.x, p4, stevenknight 19:50:50 <GregNoel> done 19:50:53 agreed. 19:51:14 1351: 2.x, p4, gregnoel? 19:51:22 <GregNoel> done 19:51:23 1351: as soon as we can use tarfile() 19:51:28 yes 19:51:53 1365: FIXED 19:52:18 yes, and 1377: wontfix. 19:52:31 1377: consensus WONTFIX 19:52:39 <GregNoel> done 19:52:25 I'm off for the evening. Steven I'll see u at baypiggies 19:52:38 cool, thanks bill 19:52:41 bye Bill. 19:52:45 <GregNoel> later 19:53:08 1382, no good easy general solution. 19:53:13 agreed 19:53:47 do we even detect it and handle it gracefully? 19:53:59 Maybe .sconsign could be chowned to real uid on linux/mac though. 19:54:13 <GregNoel> That would help most of the issues 19:54:24 <GregNoel> If you can find them. 19:54:21 I could do that. 19:54:42 1382: 1.x, p3, garyo? 19:54:42 <GregNoel> Even better would be to remove .sconsign before writing it. 19:55:00 Can't remove it if owned by root unless perms are loose. 19:55:19 Anyway, could probably do a bit better than today. 19:55:30 I'll look into it. 19:55:32 <GregNoel> directory perms control delete; can remove file owned by anyone 19:55:42 if you own the dir 19:56:00 <GregNoel> you have to have write on the dir; you need that anyway to build 19:56:01 or rather, if dir's perms permit it 19:56:23 true if the .sconsign is in the same directory 19:56:31 which of course it usually is... 19:56:43 greg: not true that dir perms control delete, I just tried it on ubuntu. 19:56:59 <GregNoel> I'll bet money; check my .sig 19:57:13 jars@jars-desktop:/tmp$ sudo touch foo 19:57:15 [sudo] password for jars: 19:57:17 jars@jars-desktop:/tmp$ sudo chown root.root foo 19:57:19 jars@jars-desktop:/tmp$ sudo chmod 444 foo 19:57:20 jars@jars-desktop:/tmp$ rm foo 19:57:22 rm: remove write-protected regular empty file foo'? y 19:57:23 <garyo-home> rm: cannot remove foo': Operation not permitted 19:58:08 <GregNoel> /tmp is special; try it on another directory 19:58:24 right, without sticky bit set 19:58:56 wow, you're right. I bow to your l33t sk1llz. 19:59:18 <GregNoel> 133t? 19:59:27 leet == elite 19:59:28 =elite 19:59:59 <GregNoel> oh. thanks for the compliment. I think... 19:59:50 h4x0r speak 20:00:14 <GregNoel> better translate that, too, for my simple mind 20:00:25 hacker 20:00:39 (garyo and I sound so with it for guys as long in the tooth as we actually are... :-)) 20:00:46 <GregNoel> more likely cracker... 20:00:55 mmm, I just read Megatokyo :-) 20:01:22 <GregNoel> Man, you guys are making me feel ancient! 20:01:16 anyway, 1382: 1.x, p3, garyo? 20:01:25 anyway, 1 or 2 more tonight? Yes, I'll take 1382. 20:01:42 1385: 1.x, p3, garyo 20:02:01 <GregNoel> I'll go for that 20:02:02 Yeah, that one will take a bit of work but worth it. 20:02:07 1388: uhh... not sure 20:02:13 not me not me not me 20:02:17 please please please 20:02:24 make the unicode stop hurting 20:02:26 Maybe more recent python will help? 20:02:34 <GregNoel> not to my knowledge. 20:02:46 help a little, but someone on the ML explained that this is a really really thorny problem 20:02:53 <GregNoel> stevenknight, do you know the sound a weasel makes as it dies? 20:03:00 pop? 20:03:15 <GregNoel> (squeeky voice) not me not me not me 20:04:06 <GregNoel> I guess you have to be there 20:04:15 ah, got it -- took a little while 20:04:36 1388: 1.x p3 to reevaluate 20:04:46 ok w/ me. 20:04:49 <GregNoel> anyway, i18n is a huge job 20:04:57 right 20:04:59 <GregNoel> maybe even later than that 20:05:06 i keep hoping some hero will show up... 20:05:13 <GregNoel> somebody needs to research it, a hero 20:05:23 <GregNoel> ah, that's what you said! 20:05:23 1388: 1.x, p3, hero 20:05:33 don't look in my direction 20:05:38 <GregNoel> ok, I'll figure something out 20:05:43 hey, i kind of like the idea of a nameless "hero" that we can assign things to... 20:05:57 <GregNoel> except that they wouldn't get done... 20:06:10 oh, details... :-) 20:06:18 ok guys, I'm getting punchy. Time for me to go to bed. 20:06:25 sounds good 20:06:25 <GregNoel> let's defer it until next time and see if we can kill the rest 20:06:30 agreed 20:06:33 same time next week? 20:06:42 OK, same time next week is ok for me. 20:06:50 <GregNoel> ok 20:07:05 all right -- good night, all 20:07:07 <GregNoel> pick up here with i18n? 20:07:10 yes 20:07:16 <