Skip to content
William Deegan edited this page Jan 14, 2016 · 2 revisions
16:41:57  *      garyo-home (n=[[email protected]](mailto:[email protected])) has joined #scons 
16:50:31  *      stevenknight (n=[[email protected]](mailto:[email protected])) has joined #scons 
16:51:41  <garyo-home>   Hi Steven; how's things? 
16:54:35  <stevenknight> hey gary -- too much going on, as usual, but okay 
16:54:36  <stevenknight> you? 
16:54:41  <garyo-home>   about the same. 
16:57:04  *      stevenknight tries to catch up on the spreadsheet 
16:58:49  *      garyo-home is doing the same 
17:01:11  *      [GregNoel](GregNoel) is no longer marked as being away 
17:01:06  <[GregNoel](GregNoel)>     Looks like there are at least three of us tonight... 
17:01:54  <[GregNoel](GregNoel)>     As I said in my email, I can only stay a half-hour, so we should get started. 
17:02:16  <garyo-home>   ok, fine w/ me.  I think someone is coming later too. 
17:02:31  *      bdbaddog (n=[[email protected]](mailto:[email protected])) has joined #scons 
17:02:33  <garyo-home>   2426 is the first... 
17:02:42  <garyo-home>   Hi Bill! 
17:03:07  <bdbaddog>     Hi 
17:03:14  <garyo-home>   Looking at 2426. 
17:03:45  <garyo-home>   I don't think tool*chain* redesign will help this issue particularly, I vote to put something reasonable in for 3.x. 
17:03:51  <[GregNoel](GregNoel)>     I still think it's invalid, and if we want an issue to make it configurable, we should add a new one. 
17:04:24  <garyo-home>   I'd be OK with that, but it'll be pretty similar to this one. 
17:04:28  <[GregNoel](GregNoel)>     but I'll go for 3.x with a change of subject 
17:04:34  <bdbaddog>     3.x 
17:04:39  <garyo-home>   ok w/ me. 
17:04:58  <[GregNoel](GregNoel)>     done, unless Steven has something 
17:05:13  <[GregNoel](GregNoel)>     (He's the other "invalid" vote) 
17:05:03  <stevenknight> 2426 is invalid 
17:05:10  <stevenknight> he doesn't specify CPPPATH 
17:05:38  <stevenknight> he'd have to add /usr/include to CPPPATH to find that <set> in preference to the current dir 
17:05:56  <stevenknight> we can't know in advance what system directories a given compiler will search on its own 
17:05:40  <[GregNoel](GregNoel)>     Er, in that case, I'm back to invalid 
17:05:42  <bdbaddog>     invalid, open a new bug to make configurable 
17:05:50  <garyo-home>   steven: I take your meaning, but still it ought to be configurable.  (Maybe Greg's right, should be a new issue.) 
17:06:08  <stevenknight> configurable how?  you can configure it right now in CPPPATH 
17:06:30  <stevenknight> CPPPATH=['/usr/include/directory_containing_set'] would make his configuration work 
17:06:33  <garyo-home>   A search for a <> header should *never* match one in the current dir. 
17:06:36  <bdbaddog>     whether it looks in . first or last. 
17:06:54  <garyo-home>   (gcc and msvc don't look in . at all for <>) 
17:07:29  <garyo-home>   CPP_SCANNER_LOOK_IN_DOT_FOR_SYSINCLUDES 
17:07:29  <stevenknight> okay, got it -- agree, new issue for configuring that behavior 
17:07:45  <[GregNoel](GregNoel)>     done 
17:07:53  <garyo-home>   ok, 2427 
17:08:24  <garyo-home>   Unfortunately this hack is what we have for now, I think we need to doc it. 
17:08:45  <bdbaddog>     doc +1 
17:09:04  <stevenknight> agree, doc 
17:09:04  <[GregNoel](GregNoel)>     maybe doc with a note that it will disappear? 
17:09:20  <garyo-home>   ... when a better mechanism is implemented.  Sure. 
17:09:44  <garyo-home>   The main thing wrong with it is it's global, and we really need a per-File thing. 
17:09:57  <garyo-home>   But anyway that's a different point. 
17:10:12  <stevenknight> thought it was per-environment, so it can be configured 
17:10:25  <garyo-home>   Sorry, right it is per-env, but per-File is better. 
17:10:27  <stevenknight> but I agree w/Greg's point about an Archive() being better in the long term 
17:11:03  <garyo-home>   I'm not 100% sure about how that would work but am willing to go with it for now. 
17:11:16  <stevenknight> doc it 
17:11:24  <[GregNoel](GregNoel)>     is that a consensus? 
17:11:27  <garyo-home>   +1 
17:11:27  <stevenknight> but should we mention it disappearing if we don't know what the replacement will be? 
17:11:34  <stevenknight> that would bug me as a user 
17:11:46  <bdbaddog>     I'd say doc it, once we have a plan to replace, then add that to doc. 
17:11:52  <stevenknight> +1 
17:11:56  <garyo-home>   Or deprecate it the usual way. 
17:12:08  <garyo-home>   doc it for now anyway. 
17:12:20  <[GregNoel](GregNoel)>     done 
17:12:41  <stevenknight> 2428:  consensus 3.x  p4 ? 
17:13:15  <garyo-home>   2428 consensus ok w/ me. 
17:13:17  <bdbaddog>     2428 +1 consensus 
17:13:25  <[GregNoel](GregNoel)>     done 
17:13:29  <[GregNoel](GregNoel)>     2429 
17:14:05  <garyo-home>   I think it's a real bug. 
17:14:12  <bdbaddog>     ditto. 
17:14:17  <stevenknight> agree 
17:14:47  <[GregNoel](GregNoel)>     The OE is an internal object, but its effects are visible, so it's a bug. 
17:14:34  <garyo-home>   2.x p2? 
17:14:45  <bdbaddog>     2.x p2 +1 
17:14:56  <[GregNoel](GregNoel)>     agree 
17:14:57  <garyo-home>   agreed. 
17:14:58  <stevenknight> 2.x p2 
17:15:09  <[GregNoel](GregNoel)>     who? 
17:15:22  <stevenknight> i have a prototype of a really different substitution mechanism that looks faster 
17:15:37  <[GregNoel](GregNoel)>     Sounds like a volunteer to me. 
17:15:37  <garyo-home>   But it may not even be subst related? 
17:15:53  <garyo-home>   steven, go for it. 
17:16:33  <[GregNoel](GregNoel)>     Bug is because call is applied to Env, not OE. 
17:16:40  <garyo-home>   Put a note in that I'll do it if Steven doesn't get to it. 
17:17:03  <[GregNoel](GregNoel)>     OK, I'll add you to the issue. 
17:17:14  <garyo-home>   +1 
17:18:00  <garyo-home>   done? 
17:18:20  <[GregNoel](GregNoel)>     yes, done 
17:18:28  <stevenknight> (sorry, afk for a bit) 
17:18:56  <stevenknight> the prototype would basically replace [OverrideEnvironment](OverrideEnvironment) 
17:19:15  <stevenknight> so there wouldn't be any distinction between "real" and "override" 
17:19:18  <stevenknight> they're just all stackable dicts 
17:19:34  <stevenknight> it takes the technique of string.Template and extends it for our purposes 
17:19:43  <garyo-home>   steven: that sounds great.  I'll help test it :-) 
17:19:49  <[GregNoel](GregNoel)>     as will I 
17:19:58  <stevenknight> the problem I'm running into is that subst_list() basically has really dumb and ill-defined semantics 
17:20:20  <garyo-home>   steven: 110% agreement there.  We've been through a few oddities with it. 
17:20:09  <stevenknight> i should write up a discussion for the ML 
17:20:13  <[GregNoel](GregNoel)>     yes 
17:20:11  <stevenknight> anyway, back to the issues 
17:18:13  <[GregNoel](GregNoel)>     2430, 2431, consensus 
17:18:18  <garyo-home>   agreed. 
17:18:54  <[GregNoel](GregNoel)>     2432, 2433, consensus 
17:19:18  <[GregNoel](GregNoel)>     2434, closed 
17:20:31  <garyo-home>   I'm fine thru 2434. 
17:20:44  <[GregNoel](GregNoel)>     2441, needs priority 
17:20:54  <garyo-home>   p3? 
17:21:01  <bdbaddog>     +1 p3 
17:21:05  <stevenknight> p3 
17:21:06  <[GregNoel](GregNoel)>     works for me 
17:21:24  <garyo-home>   great 
17:21:39  <stevenknight> 2435:  since I just attached my name to [OverrideEnvironments](OverrideEnvironments)... 
17:22:26  <stevenknight> 2.x p3 stevenknight  
17:22:28  <garyo-home>   agreed, this one's related.  It can get arbitrarily complex, but this proposal is pretty reasonable.  Would it fit with stacked dicts? 
17:22:46  <[GregNoel](GregNoel)>     The global names are available, and I looked at how hard the implementation would be once (should also work for env.Clone()) and it didn't look that bad. 
17:22:47  *      stevenknight goes to look at the original issue... 
17:23:43  <stevenknight> yes, i think stackable environments takes care of this 
17:23:49  <stevenknight> or most of what people want from it, anyway 
17:23:58  <[GregNoel](GregNoel)>     This is newenv = Environment(CPPFLAGS = Append('whatever')) 
17:24:22  <garyo-home>   right; the override env has to append to the original env. 
17:24:47  <garyo-home>   anyway, Steven will look at it, let's move on. 
17:25:00  <stevenknight> i don't think that specific syntax is viable, but the concept is the same 
17:24:56  <[GregNoel](GregNoel)>     done 
17:25:18  <stevenknight> moving on... 
17:25:30  <garyo-home>   2436: I'll take it 
17:25:43  <stevenknight> garyo-home++ 
17:25:48  <bdbaddog>     Gary+1 
17:25:49  <[GregNoel](GregNoel)>     (Hmmm...  I think my spreadsheet just crashed.) 
17:26:17  <garyo-home>   my gdocs still shows you viewing... 
17:26:28  <bdbaddog>     ditto. 
17:26:46  <stevenknight> 2437:  consensus 2.1 p3 ludwig 
17:26:57  <garyo-home>   agreed 
17:27:16  <stevenknight> 2438:  2.1 p3 who? 
17:27:23  <stevenknight> could kick it back to Jason for the test case 
17:27:31  <stevenknight> but still needs a comitter 
17:27:41  <garyo-home>   I'll commit it and work w/ him to get the testcase. 
17:27:49  <bdbaddog>     +1 gary 
17:28:01  <stevenknight> thnx 
17:28:54  <[GregNoel](GregNoel)>     2438, look at SQEC to see if it gives you any ideas 
17:29:28  <garyo-home>   2438 wouldn't be needed w/ SQEC I agree, but in the near term... 
17:31:02  <stevenknight> SQEC? 
17:31:25  <garyo-home>   "[SubstQuoteEscapeCache](SubstQuoteEscapeCache)" 
17:31:29  <stevenknight> ah 
17:28:36  <[GregNoel](GregNoel)>     (Google spreadsheets lost my login, but I'm back...) 
17:28:34  <stevenknight> 2439:  2.1 p3 
17:28:47  <stevenknight> who? 
17:29:47  <garyo-home>   someone want to integrate 2439? 
17:30:03  <bdbaddog>     I'll take it. 
17:30:10  <garyo-home>   excellent 
17:30:22  <[GregNoel](GregNoel)>     ok, works for me 
17:30:49  <[GregNoel](GregNoel)>     2440, 2442, consensus 
17:30:50  <garyo-home>   Greg, before you have to go, want to talk about 1.3? 
17:31:06  <garyo-home>   (agree w/ 2440, 2442) 
17:31:45  <[GregNoel](GregNoel)>     garyo-home, I'll leave my session running; I'll read it later 
17:31:59  <garyo-home>   ok, sounds good. 
17:32:16  <[GregNoel](GregNoel)>     2443 
17:32:17  <garyo-home>   2443's next. 
17:32:39  <garyo-home>   Steven: what about the line I list as suspect? 
17:33:03  <stevenknight> 2443:  sounds exactly right 
17:33:25  <stevenknight> i thought sure we had/have some tests of aliases with actions 
17:33:44  <stevenknight> either i'm hallucinating or those take a different code path 
17:33:53  <bdbaddog>     I"m looking at the path, and suspect maybe he's got a locally modified scons? 
17:34:10  <bdbaddog>     /home/Checkouts/Bazaar/SCons_trunk/... 
17:34:05  <garyo-home>   Well, this is a pretty nice testcase in the ticket. 
17:34:18  <stevenknight> greg confirmed the failure 
17:34:30  <bdbaddog>     ah..true. 
17:34:32  <bdbaddog>     donno. 
17:34:53  <garyo-home>   There's no way that line 699 in Action.py can work. 
17:34:54  <bdbaddog>     is this a 1.3 type issue? or 2.x? 
17:35:14  <garyo-home>   Good q.  What's the 1.3 schedule? Frozen? 
17:35:47  *      garyo-home hears nothing... great silence... 
17:35:50  <bdbaddog>     my understanding was. One more checkpoint wait 2 weeks if nothings seriously broken then 1.3 
17:36:04  <bdbaddog>     then charge forward to 2.0 
17:36:19  <stevenknight> uhh.... 
17:36:23  <stevenknight> that line looks fine, actually, 
17:36:27  <garyo-home>   That works for me; if so, then this can get squeezed into 1.3. 
17:36:30  <stevenknight> it's calling the Environment.subst_list() method 
17:36:36  <stevenknight> not Subst.scons_subst_list() 
17:36:45  <stevenknight> Environment.subst_list() does take an executor= keyword argument 
17:36:47  <garyo-home>   Right, but that eventually calls scons_subst_list. 
17:37:11  <garyo-home>   Ah, the env's subst_list should strip it out? 
17:37:19  <stevenknight> right, but it doesn't try to pass executor= to it 
17:37:22  <stevenknight> so far as i can see 
17:37:25  <[GregNoel](GregNoel)>     Taking too long; defer until next time 
17:37:31  <stevenknight> [GregNoel](GregNoel)++ 
17:37:41  <garyo-home>   hmm, ok. 
17:38:03  <[GregNoel](GregNoel)>     I propose to stop here and go on to 1.3 discussion. 
17:38:05  <bdbaddog>     put research bill?  
17:38:22  <garyo-home>   ok w/ me! 
17:38:25  <stevenknight> 2443 research bill ok by  me 
17:38:38  <bdbaddog>     o.k. on to 1.3 
17:39:02  <garyo-home>   Bill, are you still OK making the checkpoint? 
17:39:05  <[GregNoel](GregNoel)>     ARGV, got to go; cul 
17:39:12  <garyo-home>   ok bye 
17:39:21  <stevenknight> later 
17:39:31  <bdbaddog>     Later Greg! 
17:39:37  <garyo-home>   I've done one before, I can help if needed. 
17:39:49  <stevenknight> if it would help, i could open up the system that I use for cutting the releases 
17:39:53  <bdbaddog>     yes. Just taking a bit to get the changes together and coherent. the other parst are easy. 
17:39:55  <stevenknight> it's a VM 
17:40:04  <bdbaddog>     ahh. 
17:40:10  <bdbaddog>     how big's the footprint? 
17:40:24  <bdbaddog>     I can bring you a usb hardrive.. 
17:40:33  <stevenknight> i was going to let you ssh in 
17:40:38  <bdbaddog>     oh. o.k. 
17:40:53  <stevenknight> but the creation of the image is also automated 
17:41:22  <garyo-home>   I have a small VM (ubuntu) that can build a release, w/ doc tools etc. if that helps? 
17:41:47  <bdbaddog>     I'm not too worried about that part. It's just been tough getting a block of time to get the text part together. 
17:42:09  <stevenknight> that's usually been the most time-consuming part for me, too 
17:42:29  <bdbaddog>     I think we should start to enforce/encourage update Changes.txt with each checkin. 
17:42:37  <bdbaddog>     and the release message. 
17:42:50  <bdbaddog>     though svn would be fine too. 
17:43:01  <bdbaddog>     and then pushing the button is easy. 
17:42:35  <garyo-home>   Want to write it as a google doc w/ irc? 
17:42:45  <garyo-home>   +1 on both of those! 
17:43:22  <bdbaddog>     I'll try and get it done tonight. 
17:43:39  <garyo-home>   OK, if you want review just let me know. 
17:44:15  <stevenknight> agree on CHANGES.txt 
17:44:19  <bdbaddog>     sure. I'll send out text to release mail list for review. And then how do we post it to all the correct places. 
17:44:32  <garyo-home>   That, for me, was time consuming. 
17:44:45  <stevenknight> yes 
17:44:46  <bdbaddog>     Changes.txt and release notice. 
17:44:50  <garyo-home>   Tigris, sf, website... 
17:45:15  <stevenknight> first, we should give you appropriate privileges on those sites 
17:45:24  <stevenknight> and then second, there's gotta be a way to automate doing those 
17:45:13  <bdbaddog>     so the changes and release are since 1.2.x or since last checkpoint? 
17:45:36  <stevenknight> last checkpoint 
17:45:53  <garyo-home>   (but the 1.3 changes will be from 1.2) 
17:45:59  <bdbaddog>     yes. 
17:45:57  <stevenknight> originally i started trying to adjust CHANGES.txt so it would be since last release (e.g. 1.2.x) 
17:46:01  <stevenknight> but that got too confusing 
17:46:25  <stevenknight> seemed easier to grok that all of the accumulated checkpoints since the last 1.2.x line in CHANGES.txt 
17:46:31  <stevenknight> were part of 1.3.x 
17:46:26  <garyo-home>   If we have people update it on commit, won't it have to be since last release? 
17:46:30  <bdbaddog>     Could have Changes.release.txt and Changes.Checkpoint.txt or something like that. 
17:47:02  <stevenknight> ?  not following 
17:47:17  <garyo-home>   Maybe on release we could just remove the checkpoint lines, leaving only the changes? 
17:47:21  <bdbaddog>     so 3 files. Changes.txt which is running change list. 
17:47:44  <bdbaddog>     hmm. never mind..  
17:47:50  <bdbaddog>     o.k. I like Gary's idea. 
17:48:02  <stevenknight> could do that 
17:48:02  <bdbaddog>     since the checkpoints are discardable. 
17:48:14  <garyo-home>   right. 
17:48:24  <stevenknight> but I think some people do treat the checkpoints as releases 
17:48:45  <stevenknight> is there actual harm in preserving the info? 
17:48:53  <garyo-home>   it's just visual noise. 
17:49:08  <garyo-home>   Maybe we indent those or something. 
17:49:19  <bdbaddog>     O.k. also, let's checkin the announcment file, which get's wiped clean at each real release? 
17:49:45  <bdbaddog>     And for checkpoints, let just refer people to the changes.txt ? 
17:50:16  <garyo-home>   +1 on checking in the announcement file for sure. 
17:50:33  <stevenknight> dunno, doesn't seem worth extra effort to remove and reorganize 
17:50:42  <stevenknight> +1 to checking in announcement 
17:50:49  <stevenknight> yeah 
17:51:17  <bdbaddog>     O.k. I"ll check in a Blank. 
17:51:30  <garyo-home>   release-announcement.txt?  RELEASE.txt? 
17:51:53  <bdbaddog>     Announcement.txt ? 
17:52:08  <garyo-home>   works for me 
17:52:46  <stevenknight> announcement.txt (your choice capitlization) 
17:53:11  <garyo-home>   So for changes.txt we'll leave the checkpoints in for now (maybe indent or something)? 
17:53:46  <bdbaddog>     Yes. I guess we can just leave what's there now. And when we go 2.0 move Changes.txt to Changes-1.txt 
17:53:53  <bdbaddog>     In 2.0 indent checkpoints? 
17:54:31  <garyo-home>   Sure, we can iron out the details when we get there. 
17:54:49  <stevenknight> yeah 
17:54:51  <garyo-home>   (I'd be OK w/ deleting the older checkpoints too, just keep 1 release back or so) 
17:55:23  <bdbaddog>     Can we breach a 2.0 topic? 
17:55:24  <bdbaddog>     ;) 
17:55:27  <stevenknight> but i personally wouldn't invest a lot of time on it, it doesn't seem like anyone's really complaining 
17:55:38  <bdbaddog>     ok. 
17:55:42  <garyo-home>   agreed. 
17:55:50  <garyo-home>   sure, 2.0? 
17:55:56  <stevenknight> to really clean it up, you not only have to delete the checkpoint lines, but you have to merge the individual contributor sections 
17:56:06  <garyo-home>   (good point) 
17:56:00  <stevenknight> 2.0 
17:56:30  <bdbaddog>     :) My normal python question. Since time has marched on and we drew the line in the sand a while back, can we more to a newer version for 2.0 than python 2.2? 
17:57:25  <garyo-home>   what features would we gain by going to, say, 2.3? 
17:57:51  *      stevenknight will go with the collective wisdom 
17:57:53  <bdbaddog>     2.5 gets us subprocess right? 
17:57:53  <stevenknight> that said 
17:58:23  <stevenknight> 2.3 did seem only marginally better than 2.2 
17:58:27  <stevenknight> 2.4 starts to get significant 
17:58:29  <stevenknight> iirc 
17:58:51  <garyo-home>   We already have a bunch of compat stuff; I think it would have to be a language feature. 
17:58:53  <stevenknight> i don't think modules (e.g. subprocess) are a compelling reason to prefer one over the other 
17:58:59  <stevenknight> because we can handle them in compat 
17:59:18  <stevenknight> agree w/gary, language features are stronger determinants 
17:59:33  <bdbaddog>     2.5 gets' with. 
17:59:41  <garyo-home>   What about unicode? Anything important? 
18:00:08  <stevenknight> i'd have a hard time going with 2.5; google internal standard is still 2.4 
18:00:20  <garyo-home>   Bill: do you think we could really jump all the way to 2.5 though?  We'll lose all the IRIX people for sure, and some older Linuxes too. 
18:00:47  <bdbaddog>     does python 2.5 not build on irix? 
18:01:02  <bdbaddog>     2.4 gets us generators. 
18:01:15  <garyo-home>   Last I knew the latest nekochan build was 2.3. 
18:01:26  <bdbaddog>     do you not build from sources? 
18:02:11  <garyo-home>   I take it back, there's a 2.5.2 there now. 
18:02:38  <garyo-home>   (It's not what *I* do, it's what my *users* do. :-/) 
18:02:50  <bdbaddog>     ahh. users=customers? 
18:02:54  <garyo-home>   yep. 
18:02:59  <bdbaddog>     they build from sources? 
18:03:04  <stevenknight> 2.3 gets generators 
18:03:15  <garyo-home>   of course they won't run scons.  I'm just using them as an example of "typical IRIX users" 
18:03:32  <garyo-home>   generators are very useful. 
18:04:00  <stevenknight> 2.4 has decorators, which are kind of nifty but basically syntactic sugar for something you can code by hand 
18:04:03  <garyo-home>   ... but you can import generators from future in 2.2 (I think) 
18:04:12  <bdbaddog>     I've never been in an environment where I couldn't build a new version of scripting language for use by build system. 
18:04:38  <bdbaddog>     true on decorators, but anything which makes the code easier to read will be a win.. 
18:04:48  <stevenknight> true 
18:04:56  <garyo-home>   One good thing is, once we have Lukas's all-in-one Windows installer, we won't even require python on a windows box. 
18:04:59  <bdbaddog>     I'd be up for saying 2.5, pushing the checkpoitn with it and 1.3 
18:05:06  <bdbaddog>     and if the world freaks out, we backtrack. 
18:05:14  <bdbaddog>     we'lll have some time before 2.0's out. 
18:05:27  <stevenknight> probably 
18:05:50  <stevenknight> i'd have a lot of internal projects thought that would break 
18:06:13  <stevenknight> though 
18:06:16  <garyo-home>   I'd be pretty scared to go to 2.5 
18:06:32  <stevenknight> i can see either 2.3 or 2.4 
18:06:34  <bdbaddog>     steven - due to 2.4 internal to google? 
18:06:41  <stevenknight> yes 
18:06:41  <bdbaddog>     o.k. let's go with 2.4 
18:07:01  <bdbaddog>     If we slip another 6 months or more on 2.0, then revisit. 
18:07:07  <bdbaddog>     and/or google updates to 2.5..  
18:07:09  <bdbaddog>     ;) 
18:07:13  <stevenknight> yes  :-) 
18:07:27  <bdbaddog>     Gary - what'd be the basis of your fear? 
18:07:40  <garyo-home>   from [http://python-history.blogspot.com/2009/01/brief-timeline-of-python.html](http://python-history.blogspot.com/2009/01/brief-timeline-of-python.html), 2.4 was Nov 2004. 
18:07:43  <bdbaddog>     then again I"m the let's break some egg's kind of guy. 
18:07:55  <bdbaddog>     5 years ago almost. 
18:07:55  <stevenknight> how about we poll the ML for objections to 2.4, with 2.3 as the fallback? 
18:07:57  <garyo-home>   My fear? We lose users due to them not being able to upgrade their pythons. 
18:08:10  <garyo-home>   +1 on poll ML (again :-)) 
18:08:12  <bdbaddog>     they'll yell at us, and we can backtrack. 
18:08:39  <bdbaddog>     I think the mailing list hasn't provided any insight, and the only reall proof will be when the tool starts yelling at the users. 
18:08:42  <stevenknight> okay, how about:  release 1.3 first 
18:08:48  <bdbaddog>     :) 
18:08:49  <stevenknight> then float 2.4 on the ML 
18:09:06  <bdbaddog>     well we'd be putting the warning in 1.3 about next version 2.x minimum right? 
18:09:31  <bdbaddog>     that's why I bring it up now. 
18:09:54  <garyo-home>   hmm. 
18:10:13  <stevenknight> ah 
18:10:45  <garyo-home>   even if disablable, that's a little annoying. 
18:10:59  <bdbaddog>     don't we alreayd have that in place for 2.2? 
18:11:11  <garyo-home>   do we? 
18:11:25  <stevenknight> sorry bill, you kicked the ball in your own goal -- i'm back to preferring 2.3 ... :-) 
18:11:41  <bdbaddog>     oh dude.. ur killin me. 
18:12:06  <bdbaddog>     2.3 is 2003. 
18:12:08  <garyo-home>   (my vm is being annoying, or I'd look) 
18:12:11  <bdbaddog>     6 years aog. 
18:12:20  <stevenknight> so we turn the clock forward five years! 
18:12:27  <bdbaddog>     wheel's were square then. 
18:12:37  <stevenknight> that's almost half way! 
18:12:40  <stevenknight> :-) 
18:12:58  <garyo-home>   Steven: what changed your mind 2.4 -> 2.3?  I don't think we'd lose that many users. 
18:13:01  <bdbaddog>     I don't think anyones using 2.3 
18:13:08  <stevenknight> having to put the warning in 1.3 
18:13:23  <garyo-home>   But Bill's saying we already have a warning. 
18:13:27  <bdbaddog>     we can always patch it back in 1.3.1 if we get a lot of negative feedback. 
18:13:52  *      stevenknight breathes deeply 
18:14:01  <stevenknight> oooo... kayyyy.... 
18:14:14  <bdbaddog>     it'd be a 1 line patch and realease. if it's really bad. 
18:14:36  <stevenknight> you want to make the change in this checkpoint?  or only for 1.3 release? 
18:15:04  <bdbaddog>     hmm. 
18:15:08  <garyo-home>   If we get zero feedback from the warning, then I think we're safe.  If we get even one negative, I'll want to revisit. 
18:15:12  <bdbaddog>     if the codes already there then for checkpoint. 
18:15:21  <garyo-home>   bdbaddog: agreed. 
18:15:22  <stevenknight> warning in a checkpoint, or in a release? 
18:15:40  <garyo-home>   both (assuming it's already there now) 
18:15:40  <bdbaddog>     checkpoint if the check is already there, otherwise 1.3 
18:15:59  <stevenknight> although some people treat checkpoints as release, people that are still using 2.3 are unlikely to track checkpoints 
18:16:23  <stevenknight> so silence from the checkpoint warning has strong potential to be a false positive 
18:16:25  <garyo-home>   agreed. iit needs to be there in 1.3 anyway. 
18:16:50  <stevenknight> okay, i can go with it 
18:17:04  <stevenknight> now we just have to twist Greg's arm after he reads this... :-) 
18:17:08  <bdbaddog>     codes already there. 
18:17:12  <bdbaddog>     :) 
18:17:22  <bdbaddog>     eh.. sorry I can't hear you.. zztt zttt static on the line.. 
18:17:26  <bdbaddog>     True. 
18:17:27  <garyo-home>   ... so our existing checkpoint is already warning at 2.2? 
18:17:29  <stevenknight> you sneak, you... :-) 
18:17:34  <bdbaddog>     yes. already there. 
18:17:42  <bdbaddog>     I didn't do it. somebody else did it. 
18:17:52  <stevenknight> oh, wait -- i knew it was warning re: 2.2 
18:17:57  <garyo-home>   Right, I kind of remember that now. 
18:18:00  <stevenknight> i thought you meant you already checked in the 2.4 warning 
18:18:00  <bdbaddog>     yes warning 2.2 
18:18:11  <bdbaddog>     no.. didn't do that.. dang. wish I'd thought of that. 
18:18:13  <garyo-home>   So we just bump that warning level up a notch. 
18:18:20  <bdbaddog>     exactly. 
18:18:21  <stevenknight> right 
18:18:23  <garyo-home>   or two. 
18:18:29  <stevenknight> or .2 
18:18:30  <bdbaddog>     +.2 
18:18:36  <garyo-home>   ok, I'm on board, let's see what happens. 
18:18:49  <bdbaddog>     o.k. I just don't want the project to get stuck in the past like Plone.. 
18:18:58  <bdbaddog>     and be too worried about moving forward. 
18:19:18  <garyo-home>   right, or like not changing Makefile tab syntax because it already had 100 users. 
18:19:45  <garyo-home>   ok, so we can call it a night I think? 
18:19:52  <bdbaddog>     yes. Thanks to all! 
18:19:55  <garyo-home>   Bill, let me know if I can help w/ the checkpoint. 
18:20:10  <bdbaddog>     will do. I'll try to get the text out tonight and packages ready too. 
18:20:17  <garyo-home>   Sounds great. 
18:20:34  <garyo-home>   Thanks all. 
18:20:36  <garyo-home>   cul 
18:20:40  <bdbaddog>     if whomever can give me access to the appropriate uploads I'd need can do that and/or push the packages when done. 
18:20:59  <garyo-home>   Oh yeah, Steven, can you do that? 
18:21:40  <garyo-home>   I'll email you the website login/password, Bill. 
18:22:03  <bdbaddog>     k. thanks.  
18:22:11  <stevenknight> okay, i'll add bill to SF, tigris.org and...  what else? 
18:22:21  <stevenknight> feel like i'm missing something 
18:22:26  <stevenknight> pair.com? 
18:22:37  <garyo-home>   I think it's just those two, I'll get him the pair login/password. 
18:22:46  <stevenknight> okay, i'll take sf and tigris.org 
18:22:48  <stevenknight> many thanks guys 
18:22:52  <garyo-home>   np 
18:23:08  <garyo-home>   'night. 
18:23:16  <bdbaddog>     night! 
18:38:17  *      garyo-home has quit ("[ChatZilla](ChatZilla) 0.9.85 [Firefox 3.5.2/20090729225027]") 
19:12:38  *      stevenknight has quit (Read error: 110 (Connection timed out)) 
20:31:57  *      [GregNoel](GregNoel) has been marked as being away 

Clone this wiki locally