Clone wiki

SCons / BugParty / IrcLog2010-01-05

16:49:01 * Jason_at_Intel (n=chatzill@ has joined #scons 16:56:31 * You are no longer marked as being away 16:56:47 <GregNoel> loonycyborg, you with us for the bug party? 16:58:14 <loonycyborg> GregNoel: What can I add to it? Besides you always have them when I should be sleeping :P 16:59:53 <GregNoel> Sleep? What's that? 17:00:23 <sgk> that's that thing other people do where they close their eyes and don't move for long periods of time 17:00:27 <sgk> or so i've heard 17:00:27 <GregNoel> And the answer to your question is that you have a better perspective on what the users are seeing than any of us here. 17:01:01 <GregNoel> s/seeing/wanting, needing, desiring, .../ 17:01:50 <sgk> no sign of bdbaddog and garyo was sounding pretty iffy... :-( 17:02:17 <GregNoel> Gary said he'd likely be late, so I'm not panicking yet. yet. 17:06:00 <Jason_at_Intel> are we waiting for Steve? 17:07:01 <sgk> i'm here 17:07:13 <Jason_at_Intel> right :-) 17:07:14 <sgk> just under a different (officially registered) nick 17:10:06 <GregNoel> loonycyborg, can you open the "current issues" spreadsheet? There's a link from the BugParty wiki page. 17:11:54 <GregNoel> And also the current issuezilla page; there's also a link from the BugParty page. I keep them in different tabs in my browser so I can flip back and forth readily. 17:13:41 <loonycyborg> GregNoel: I've opened them. 17:13:53 <GregNoel> Steven, should we start? It looks like there's close to consensus on the first few; that will give Gary a chance to arrive. 17:14:12 <sgk> sounds good, let's go 17:14:47 <GregNoel> loonycyborg, the easiest way to follow in issuezilla is to click on the "long format" button 17:14:38 <sgk> 2071: consensus 2.0 p0 sk 17:14:50 <GregNoel> done 17:16:10 <sgk> question about the license (to see if anyone has an opinion) 17:16:20 <Jason_at_Intel> what does this mean? 17:16:37 <Jason_at_Intel> is this going to cause a problem for me and Parts add on for Scons? 17:16:40 <sgk> sorry, mean release forms for code 17:17:28 <Jason_at_Intel> as in legal forms? 17:17:46 <sgk> Jason_at_Intel: to avoid possible legal hassles, we should really have some paper showing it's legal to take code 17:18:00 <GregNoel> "contributed code" 17:18:01 <sgk> contributed code 17:18:03 <sgk> right 17:18:20 <sgk> we actually have a draft of an assignment that I used for a bit years ago 17:18:28 <sgk> something I scraped together from other examples 17:18:53 <sgk> probably full of IANAL holes 17:18:23 <Jason_at_Intel> As I understand it legally you just need a .lic file shipped with the license 17:18:27 <Jason_at_Intel> like i have in Parts 17:19:01 <Jason_at_Intel> Or so I was told by Intel Lawyers 17:19:19 <sgk> Jason_at_Intel: that's fine for you distributing Parts 17:19:32 <sgk> the question is if the SCons Foundation were to incorporate Parts into SCons 17:19:44 <sgk> if we don't have paper from you and Intel saying it's okay 17:20:04 <sgk> then legally we'd be open to getting sued for improperly redistributing your IP 17:20:03 <Jason_at_Intel> I see 17:20:17 <Jason_at_Intel> course the point of Parts it to be added to SCons 17:20:41 <Jason_at_Intel> right... you have that OK 17:20:55 <Jason_at_Intel> I thought gave you an e-mail orginally with all that 17:21:17 <sgk> believe so, but we haven't been regular about getting this from contributors 17:21:33 <Jason_at_Intel> Everything in Parts is for Scons to take and use as they like 17:21:34 <sgk> current agreement basically says you have to give code to scons 17:21:39 <sgk> which makes some corporate lawyers nervous 17:21:59 <Jason_at_Intel> sure.. I understand 17:22:19 <sgk> our lawyer (last I talked) said since we're MIT [license] we could get by with just having contributors license their code to SCons under the same MIT terms we use 17:22:29 <sgk> so contributors still retain ownership 17:22:20 <Jason_at_Intel> if "we" needed to clarify anything here let me know 17:22:46 <Jason_at_Intel> sort of why we released under MIT 17:22:57 <Jason_at_Intel> normally Intel would have wanted a BSD 17:23:12 <GregNoel> to proceed, 2509, 1.3 p1 Gary +doc 17:23:22 <sgk> 2509: done 17:23:25 <Jason_at_Intel> K 17:23:33 <GregNoel> 2518 17:23:54 <sgk> you okay with the other consensus? 17:23:55 <GregNoel> I'll make it a dup of 2536 17:25:11 <sgk> 2518: done 17:25:17 <GregNoel> 2521 17:25:36 <GregNoel> I'll go with the flow 17:25:43 <sgk> 2521: ditto, research p2 bdbaddog 17:25:48 <GregNoel> done 17:25:58 <GregNoel> 2522 17:26:08 <sgk> 2.x p4 okay w/you? 17:26:34 <GregNoel> Already too much in 2.x; 3.x? 17:26:42 <sgk> fine with me 17:26:48 <GregNoel> 3.x p3? 17:27:00 <sgk> +1 17:27:02 <GregNoel> done 17:27:24 <sgk> 2523: 2.x p3 +symlink +sconf_revamp 17:27:28 <GregNoel> 2523 2.x p3 17:27:36 <GregNoel> er, sure, we agree 17:28:27 <GregNoel> 2531, this came up in the mailing list today 17:28:52 <GregNoel> I'll go with the flow; future is OK. 17:27:56 <Jason_at_Intel> is the auto config going to get redone in 2.x 17:28:32 <Jason_at_Intel> or better yet are the symlink nodes going to be handed by SCons? 17:29:27 <GregNoel> Already in train. 17:28:56 <Jason_at_Intel> moving on 17:29:08 <Jason_at_Intel> you would need to redo the taskmaster 17:29:33 <Jason_at_Intel> and the task queue 17:29:46 <sgk> 2531: future p2 17:29:50 <GregNoel> done 17:29:54 <Jason_at_Intel> agreed 17:30:26 <sgk> Jason_at_Intel: all of it needs attention 17:30:36 <sgk> I'm hoping to get guidance from you all on what's top priority for my time 17:30:44 <sgk> instead of ending up all over the map like I usually do... 17:30:12 <GregNoel> 2532 17:30:39 <GregNoel> Steven, you want it? 17:30:56 <sgk> 2532: 2.x p2 sk 17:31:06 <GregNoel> I don't like it that far out, but I'm willing to try. 17:31:17 <sgk> you'd like it sooner than 2.x? 17:31:28 <Jason_at_Intel> is this a regression? (2532) 17:31:44 <GregNoel> No, I don't like something with your name on it that far out, for precisely the reasons you just said. 17:31:57 <sgk> right 17:32:11 * sgk goes to re-read the issue... 17:32:39 <GregNoel> Maybe assign it to Gary for research and recommendations. 17:34:12 <GregNoel> Steven, we lose you? 17:34:26 <sgk> no, was off taking a look at the code 17:34:55 <GregNoel> Maybe assign it to Gary for research and recommendations. 17:35:07 <Jason_at_Intel> so 2532.. research it? 17:35:16 <GregNoel> Gary for research and recommendations? 17:35:46 <sgk> no, give it to me, i think I just figured out a pretty simple fix 17:35:55 <GregNoel> OK, done 17:36:09 <GregNoel> what milestone and priority? 17:36:11 <sgk> we should handle that like we do the other allowable exceptions in substitution 17:36:15 <sgk> 2.1 p2 17:36:18 <GregNoel> done 17:36:35 <GregNoel> 2533, should be 2.1 p3 garyo 17:36:49 <Jason_at_Intel> 2533... this seem to be a bug in how the win32 installer is made... ... user has to elevate it to run it 17:37:10 <Jason_at_Intel> or the installer has to be made to get elevation by the system 17:37:32 <sgk> 2.1 p3 garyo 17:37:32 <sgk> done 17:37:37 <GregNoel> done 17:38:11 <GregNoel> 2534 17:38:30 <sgk> i'm okay with doc+test p3 17:38:43 <GregNoel> milestone? 17:39:13 <GregNoel> And is node_class=None the right solution? 17:40:09 <sgk> looking... 17:41:44 <sgk> sheesh, who designed this API? 17:42:05 <GregNoel> Er, that would be you? 17:42:11 <sgk> yep... :-) 17:42:34 <Jason_at_Intel> SEP for cleaned on API are needed 17:42:39 <Jason_at_Intel> ;-) 17:42:46 <loonycyborg> Scanner api indeed seems kinda weird. 17:42:54 <sgk> I'm confused because the default is actually Entry, not File 17:43:04 <sgk> which normally means that returning a Dir should be okay 17:43:27 <GregNoel> Yeah, I agree... So why's it fail? 17:43:49 <sgk> +1 re: a SEP to clean up APIs 17:44:07 <sgk> ah 17:44:19 <GregNoel> Who would write the SEP? Or should there be more than one? 17:44:28 <sgk> more than one 17:44:47 <sgk> one per area of cleanup, probably 17:45:11 <sgk> okay, i'm taking it back, i think 2534 needs research 17:45:28 <loonycyborg> I'd prefer if a scanner always was a function taking node, returning list of nodes. 17:46:11 <sgk> loonycyborg: that would be a step in a more sane direction, but it's a little more complicated 17:46:28 <sgk> because a scanner is really conceptually attached to an edge in a DAG, not a node 17:46:41 <GregNoel> ... and there are some other considerations 17:47:31 <GregNoel> but an API that ran a function with an upstream node would go a long way toward solving a number of problems. 17:47:18 <sgk> re: 2534, need to figure out where the File lookup is coming from 17:47:49 <GregNoel> OK, who? not you? 17:48:51 <sgk> me... not me... me... not me... 17:49:12 <sgk> i dunno 17:47:50 <loonycyborg> Probably api is like that due to taking specifics of scanning c/c++ files in account, e.g. search path etc. 17:48:24 <GregNoel> API would need an Environment, but that's a secondary consideration. 17:48:24 <loonycyborg> But you could just make a canned scanner for that case.. 17:49:05 <GregNoel> loonycyborg, caching becomes a problem. 17:49:22 <GregNoel> sgk, decision, or bypass? 17:49:42 <Jason_at_Intel> research seem to be best 17:49:44 <sgk> let's defer until next week 17:49:55 <sgk> it'd be sane to have someone else research 17:49:57 <GregNoel> done 17:50:10 <GregNoel> 2535 17:50:26 <Jason_at_Intel> I have this working in Parts 17:50:32 <Jason_at_Intel> you can take my code for this 17:50:44 <sgk> 2535: 1.3 p1 garyo 17:50:55 <GregNoel> 2535, I don't have a clue 17:50:56 <Jason_at_Intel> but gary is best guy for this 17:51:08 <sgk> Jason_at_Intel: could you update the issue with that info re: code in parts, so he'll see it when he looks? 17:51:25 <Jason_at_Intel> sure... 17:51:24 <GregNoel> I'll resist 1.3 17:51:55 <GregNoel> We should be cutting it next week, unless there's a regression in the checkpoint. 17:52:33 <sgk> re: 1.3, is that on a separate branch or is it on trunk still? 17:53:00 * sgk is worried that he's potentially messing up 1.3 with recent checkins... 17:53:18 <GregNoel> I think you did a rebase recently; that should be the candidate 17:53:36 <GregNoel> If it's not, it should be twiddled so that it is 17:54:41 <GregNoel> Were you the one to release the checkpoint? If so, which branch did you use? 17:55:05 <sgk> bdbaddog did 17:55:13 <sgk> that's right, we have the checkpoint branch for that... duh 17:55:48 <sgk> so 2.1 p1 garyo? 17:55:51 <sgk> for 2535? 17:56:38 <GregNoel> p1 or p2? I don't think it's p1-urgent 17:57:09 <sgk> good point, p2 17:57:18 <sgk> can be escalated if it starts burning anyone 17:57:49 <GregNoel> done 17:58:12 <GregNoel> Should we go on to your research issues? 17:58:24 <Jason_at_Intel> added notes of basic code 17:58:32 <sgk> sure, let's just hit obvious ones for now (consensus, etc.) 17:58:42 <GregNoel> I think we should settle the lawyer issues; don't know about the rest 17:58:49 <sgk> 2130: 2.0 p0 sk 17:59:06 <GregNoel> 1910, no consensus; bypass 17:59:16 <GregNoel> 2130, yes 17:59:31 <GregNoel> (I'll really make them p1) 17:59:42 <sgk> crap, i thought I went through these 17:59:51 <sgk> obviously I didn't 17:59:48 <Jason_at_Intel> ideally this is just asking for a license to be added in the documentation 18:00:04 <GregNoel> 765, 2.x p2 garyo 18:00:05 <sgk> yeah, just need to stamp it with the appropriate creative commons license 18:00:27 <sgk> 765 done 18:00:35 <GregNoel> 2361 bypass 18:00:59 <GregNoel> 780 bypass 18:01:04 <Jason_at_Intel> I need to do a SEP for packaging 18:01:26 <GregNoel> 914, bypass reluctantly 18:01:40 <GregNoel> 1187 bypass 18:01:52 <GregNoel> 1745 bypass 18:02:12 <GregNoel> 1883 bypass (dup?) 18:04:59 <GregNoel> None of the rest have enough comments... 18:05:10 <sgk> yep, sorry about that 18:02:03 <sgk> 914: probably wontfix at this point 18:02:19 <sgk> it's been superceded by the stuff I copped from Chromium 18:02:37 <GregNoel> your choice 18:03:17 <sgk> re: reluctantly: is there specific functionality you had in mind that you wanted from 914? 18:04:12 <GregNoel> No, but the XML output in a standardized format is a good idea. 18:04:21 <GregNoel> I don't know what you added from Chromium 18:04:48 <sgk> hmm, what if we just mark it future so it doesn't fall off the radar screen? 18:04:51 <sgk> that's a little lame, but... 18:05:08 <GregNoel> Future is on the radar screen? 18:05:19 <sgk> fair point 18:05:31 <sgk> it's less off the radar screen than WONTFIX... 18:05:31 <GregNoel> bypass until next time 18:05:34 <Jason_at_Intel> I can't seem to edit the file so i have been unable to add comments 18:05:58 <sgk> Jason_at_Intel: oh, I meant to update the issue at, not in the spreadsheet 18:06:03 <sgk> if that's what you were trying to do 18:06:19 <sgk> the spreadsheet is just to try to streamline the triage process in these meetings 18:06:34 <sgk> it's not for long-term tracking of info on specific bugs 18:06:43 <Jason_at_Intel> No i added comment on the bug at tigris 18:07:01 <sgk> okay, thanks 18:08:26 <GregNoel> decision on 914? 18:09:22 <GregNoel> (we've run over and I don't know if there's anything to discuss about 1.3) 18:10:02 <Jason_at_Intel> Steve? 18:11:04 <sgk> 914: defer to next time 18:11:11 <sgk> along with rest of research 18:10:21 <Jason_at_Intel> 2347 will be fixed by taskmaster NG? 18:10:35 <Jason_at_Intel> greg? 18:12:04 * GregNoel brb 18:14:59 <GregNoel> back; 2347 not related to taskmaster; related to how symlinks should work: 'value' of symlink is string reference, but has to be worked out so read and write work. 18:15:49 <Jason_at_Intel> Greg.. thanks! 18:35:57 * GregNoel just had his wife suggest that it would be a good thing to come to dinner... 18:36:21 <sgk> GregNoel: thanks, say hello to your wife 18:36:38 <GregNoel> wilco, cul 18:36:49 * You have been marked as being away 18:37:00 <Jason_at_Intel> later greg! 18:36:33 <sgk> i should go, too -- i'm still at work and have to buy a printer on the way home 18:37:53 <Jason_at_Intel> well guess you got to go.. I should go help take care of my kids 18:45:30 <sgk> gotta get going, catch you guys later 18:45:44 <Jason_at_Intel> ok later! 18:45:51 * sgk (n=sgk@nat/google/x-rfygfhizlqsajbfq) has left #scons 18:46:03 <Jason_at_Intel> I got to go as well 18:46:07 <Jason_at_Intel> later 18:46:13 * Jason_at_Intel has quit ("ChatZilla 0.9.86 [Firefox 3.5.3/20090824101458]")