Session Start: Mon Aug 14 22:01:28 2006 Session Ident: #twiki_edinburgh [22:01] * Now talking in #twiki_edinburgh [22:01] we are in a restaurant on the shore [22:01] very nice [22:01] place [22:01] I've been there in 2003 [22:01] yes [22:01] but lot of light [22:01] Hello everyone [22:01] i have a hard time seeing the screen [22:02] and the wifi is low [22:02] so i might drop off [22:02] 'evening lavr [22:02] peter: sounds like it takes a fighter to stay online, keep hanging in there :-) [22:03] btw: feel free to visit some of the pictures from our wedding, they are at http://www.ibensverden.dk/fotos/20060805?page=1 [22:03] he, i like to do what i like to do :-) [22:04] ok, i will check it out later [22:04] i am here with friends from germany [22:04] nice :-) .. ok, we should not waste too much time then [22:06] i am half way "listening" [22:06] please go ahead whenever ready [22:06] someone should take the lead on moderating [22:06] any priorities we could take a stab at with the headcount we have available? [22:06] and taking minutes [22:06] I started minutes. [22:06] great [22:06] http://twiki.org/cgi-bin/view/Codev/EdinburghReleaseMeeting2006x08x14 [22:07] And I updated the agenda just before the meeting. Cleaning out already decided stuff and add new. [22:07] thanks lavr [22:07] proposed agenda: [22:07] # 1. Review Previous Action Items [22:07] # 2. Process to accept change proposals [22:07] # 3. Review proposed items [22:07] # 4. Profiling framework for TWiki [22:07] I changed 4 and 5 to [22:07] 4 Reply to NuwikiRip [22:07] 5. Profiling framework for TWiki [22:08] i will be full "ears" on # 2 [22:08] I can't remember the first review item, perhaps I was not there [22:08] Doc Team: Discuss if we also want to reduce content as proposed by Sven and supported by many in community??? [22:09] Which content? [22:09] distro or to? [22:09] Guess that was an input to doc team. Persoanlly I do not understand this general statement. [22:09] t.o. [22:09] http://twiki.org/cgi-bin/view/Codev/EdinburghReleaseMeeting2006x07x18 has "Sven suggested to also eventually remove old topics from twiki.org." [22:10] phone [22:10] i think this should be discussed by doc team [22:10] we have not met after that meeting [22:10] ok [22:11] let us move that point to next time [22:12] action item for doc group: schedule next meeting [22:12] next would be: All: Add to "Wishlist" BenchmarkFramework - requirements and wishes - very few wishes added. Need more input! [22:13] I have no input to that one [22:13] i think we need input on this [22:13] and we should put some effort into profiling [22:13] I reviewed the inputs tonight and nothing to discuss without CDot and Harald [22:13] there is a good start [22:14] * Soronthar has joined #twiki_edinburgh [22:14] Hi Raf [22:14] harald has a very good start [22:14] looks promisong [22:14] ..ing [22:15] hi rafael [22:15] Hi all [22:15] Hi Peter, Kenneth [22:15] i think that action item needs to be there for next time [22:16] I think we should try and let Harald push this since he has fresh eyes on everything and none of the wishes are non-supportive at all. [22:16] * OliverKrueger has joined #twiki_edinburgh [22:16] agreed, yep, action item for next time, and let's do what we can to encourage Harald on his efforts [22:16] the problem is that there is not enough feedback on haralds proposal / prototype [22:17] i am sure he can create something very useful [22:17] with more feedback it will be even better [22:18] are we talking about Profiling framework for TWiki? [22:18] yep, http://twiki.org/cgi-bin/view/Codev/BenchmarkFramework [22:18] ok [22:19] Harald has some ideas and I think we should support him. And expect him to work at a pace of a person that has a fulltime job other than TWiki. [22:19] It is good with fresh input and new ideas. [22:20] we can ask him if there is something that can be split out, so that he gets help on implementation [22:20] Hi Oliver. Almost missed your entrance. [22:21] Hi Lavr. Sorry 4b'ing l8. [22:25] I think that's a good idea .. ok, let's leave it there and put it on again for next agenda? [22:25] Yes. [22:25] Next would be: "Meredith: Create TWikiFn? branch of TWiki4 with no other changes." [22:25] I did some benchmarking on it and found that.... [22:25] A virgin TWiki is marginally slower with TWikiFn. [22:26] A Twiki with 550 users and some hundreds of topic as marginally faster with TWikiFn. [22:26] So in total. I see no performance issue. [22:26] great [22:26] But I see an implementation issue. Given the last statement from Meredith. Ie. the implementation may depend on others taking over. [22:27] meredith said that she will be leaving the twiki project [22:27] no surprise [22:27] do we have enough resources to finish this project? [22:27] what is the work? [22:27] I am ignorant to what is missing to have it finished [22:27] there is doc work, plugins web setup work, test cases and other stuff pending [22:28] hmm [22:28] ah yes, that is right [22:28] Implementing it without a major follow up in making small plugins into Fns removes the whole point in doing it. [22:29] TWikiFns will be listed alongside plugins as a "new kind"? [22:29] (in the Plugins web) [22:29] i think we need to see some commitment for taking this into 4.1 [22:29] The idea was to make lightweight plugins that were easy to make and fast [22:29] can be anybody, including meredith [22:30] meredith has stated that it was important to her to see this going in, but yes, someone will have to demonstrate the momentum for the last part of the road [22:30] I am trying to find the "almost goodbye" note. [22:31] it's in the quiet leavers topic [22:31] http://twiki.org/cgi-bin/view/Codev/BewareOfTheQuietLeavers [22:32] Thanks [22:33] I propose an action item: Write a checklist of "things that need to happen for this to be approved" in http://twiki.org/cgi-bin/view/Codev/TWikiFns [22:33] So in principle she says that she is leaving #twiki -not TWiki. [22:33] Lavr: yes. but she has stated several times that her ties to TWiki end as soon as she finish her current project [22:34] not much we can do about it now as I see it, the branch is running and will need to remain stable for a while merging in fixes from 04x00 and getting the remaining work done [22:34] She had everything ready. I even tested it. So the code it there for the framework. But TWikiFn without commitment to develop it further is not worth anything. [22:34] everuthing ready == code and some doc [22:34] What does "without commitment to develop it further" means? [22:35] That is, besides having the code, several tags, documentation and tests in place? [22:35] packaging of the fns needs to be designed and decided [22:35] Yes. But the whole work in really implementing it is waiting ahead and it much more work. Ie. setting up twiki.org. Converting simple plugins to TWikiFns etc etc [22:35] testcases need to be written [22:36] ok. So what is needed is the commitment for the "non-development" part. [22:36] it takes a very committed person to follow up on this for the next 6 months [22:36] i posted a note in some topics in the plugisn web where meredith posted some preliminary stuff as "tags" [22:36] i did not get any feedback [22:36] It is also development. Turning small plugins into TWikiFns is real work. [22:36] Real code work I mean [22:36] so i am concerned and wondering if there is commitment [22:37] again [22:37] please, somebody write a checklist of what is missing. [22:37] I doubt there is a commitment right now. So unless someone walks in and says - I will drive this - we need to park it for now. [22:37] agreed [22:38] thsi someone can be anyone in the community, incl meredith [22:38] I would not even know how to make a complete checklist. We need a driver. And if Meredith says - I will do it - I will support that it gets continued. [22:38] not to say we do not find it useful [22:39] No. I have been supportive of the TWikiFn idea from the start and it is sad if it is wasted effort. [22:40] I do want to support this, but cannot be the responsible person [22:40] you know? There won't be any commitment outside this channel if there is not a clear view of what needs to be done [22:40] But if something happens in Merediths life that makes other things more important right now then she has my greatest respect for that. She may come back in 2 months when the summer is over and the days get dark and cold. [22:41] Note - Not many have done anything on the project the past month! [22:41] there will be a lot of work in the design / packaging / doc / t.o. doc areas until it's good and ready, but these issues will only start acceleration once it is in the main branch and people start using it [22:41] i have about 20 min left [22:42] Action. Ask Meredith if she is committed to drive it further - and when. [22:42] *sight* [22:42] can we forget that Meredith exist, and try to push this feature further without her? [22:43] not at this time [22:43] yes, if someone takes the commitment [22:43] Soronthar: It's really just a question of putting in man hours, anybody willing can do it [22:43] and how can we get someone outside this room to take commitment? [22:43] she hasn't said she will stop completely [22:44] My point is: several things can be done without her, like documenting or defining packaging or whatever else is needed. [22:44] Let us ask her first. And if she throws the towel, the task is open for the community to take over [22:44] I'm clear tha we need docs and packaging. The rest is not clear to me [22:44] sometimes, to get commitment the scope of the work must be defined first. [22:45] can we move on to next item? [22:45] action item for soranthar, taking a stab at defining scope? [22:45] we need the input from meredith [22:45] ok [22:46] ok for the action item for me. [22:46] thanks [22:46] I'll put the checklist in http://twiki.org/cgi-bin/view/Codev/TWikiFns in the following hours [22:46] great [22:46] good! [22:46] PeterThoeny: Any agenda item we need to definetely touch before you leave? [22:47] Yes. The NuwikiRip [22:47] i'd like to cover "Process to accept change proposals" [22:47] ok, nuwikirip first [22:47] short one i think [22:48] the guy has agreed to remove the icons and css [22:48] th core team members have seen the reply from gnu legal [22:49] the gnu person basically said that twiki is a collection of gpled work [22:49] the collection is gpl [22:49] and its part are gpl [22:49] Yes. The company can probably regard the skin as independant work and release their modified version as GPL. [22:49] ArthurClemens: yep, saw much were updated at their site already [22:49] the stylesheets and icons could have another license, do not need gpl [22:50] But it is clear that IF they modify our style sheets they must release their skin as GPL. Unless we dual license. [22:50] I see no reason to dual license. There is absolutely no money in it! [22:50] so, in other words, from gnu perspective, the copyrigh holder of the part can decide what to so (different license etc) [22:50] but this is from a programming perspective, not from a design pov [22:50] (the idea to use a different license) [22:51] i think this is really up to arthur [22:51] It it important to understand that design and program in the eye of GPL is the same thing. [22:51] and arthur stated his position [22:51] I have released it under GPL [22:51] _and_ the company is retreating [22:51] so there is not much left to say! [22:51] Yes. And you can decide to add additional license but not revoke the GPL: [22:51] (we have to see them finish it) [22:52] We still need to give them an official reply. [22:52] i suggest this: [22:52] - arthur sends a proposed reply to the twiki-core [22:52] - core team members give feedback [22:52] - arthur posts reply on website [22:52] I did step 1 [22:52] this is basically a polite "no thank you reply" [22:53] I propose that the answer contains the statement that we confirm that the skin is released under GPL... [22:53] so I will await the feedback and then put the comment in the topic [22:53] the company was kind and concerned enough to publikly post on twiki.org [22:53] and that we do not wish to release it under any other commercial license. [22:53] so we should handle it politley [22:54] I agree [22:54] ok, move on [22:54] ? [22:54] i have only a few min left [22:54] OK. [22:55] i think everyone read http://twiki.org/cgi-bin/view/Codev/TWikiRelease04x01Process [22:55] we got some feedback but not from many [22:56] i am wondering if there are any concerns or showstoppers for the proposal [22:56] I am happy to see that CDot supports the proposal provided that we can find names for the essential roles. [22:57] yes, it is very good to have support from crawford [22:57] i really appreciate this since it means for him to step back from his proposal [22:58] Rafael - any comment from you? This is a dramatic change from previous process towards more empowerment to both developers (accept by default) and customers (anyone can ask for a proposal to be discussed in release meetings). [22:58] any other feedback from active members? [22:58] yes, rafael, i aws just going to ask if you have any feedback [22:58] gimme a sec to refresh my memory [22:59] I would of course just really like to start effectuating it .. but if we missed the obvious it's better handled up front [22:59] I think that the only think that "worries" me is the Veto [23:00] I mean, a Veto right is really needed, sometimes [23:00] but needs to be implemented in a way that is not seen as "dictatorship" [23:00] Yes. Let me try and explain. In the past the veto has defacto been Founder (Peter) veto only. [23:00] Now it changes. So veto takes TWO core team votes. [23:01] The idea is that we want to protect against someone gathering 10 of his friends for a release meeting to vote a bad proposal through. [23:01] can a Veto be contested/objected? How? [23:02] let me explain [23:02] bad == as measured with the twiki mission [23:02] following the principle of "good faith" [23:02] two core members think that something must be Vetoed, and then issue the veto [23:03] The developer should be given a chance to "defend" it's possition one more time. [23:03] i think this is a good point [23:03] and then the Core members can decide if there are merits to retire the veto [23:03] a veto does not mean an absolute "no" [23:03] There is nothing in the proposal that prevents someone to put forward a proposal an infinite number of times. [23:03] so, the veto cannot be "overruled", but can be contested. [23:03] it can mean going "back to the drawing board" [23:04] A veto will be given because of a specific reason and of a proposal is improved it has a good chance next time. [23:04] I would expect that "vetoed" means "vetoed for this release" [23:04] i many cases, yes [23:05] i can see also a veto that is a "no" [23:05] for example, if someone wants a 'myspace" like feature in twiki, which is against the twiki mission [23:06] in otehr words, a veto can mean, "not for twiki core, but feel free to realize as extension" [23:06] Someone proposes - let us completely change the TWiki plugin API. To hell with all 220 plugins. [23:06] I guess that would give 4-5 veto votes. [23:06] the purpose of the veto right is to keep the project focused [23:08] the veto right has a string attached, it must be based on the mission; that is a core team member cannot veto because "i do not like it" [23:08] yes, both kind of vetos exists, hopefully most vetos will be of the "need more doc work" / "need more thought" / "need this specific improvement" type - not the "simply not acceptable" kind [23:08] So, veto is "feature rejected for some reason" [23:08] ? [23:08] yep [23:09] the reason needs to be given [23:09] of course [23:09] a veto (if/when it occures) automatically results in a dialog [23:09] it should. [23:09] thsi dialog may result in a new solution that is better then one or the other [23:09] That's why I want to make clear that a veto can be contested (but not overruled) [23:10] that makes sense, yes [23:10] And a core team member that still vote "no" without saying "veto" and accept a majority vote. [23:10] exactly [23:10] the veto should not be abused [23:10] ok, i need to sign off now [23:11] any other things before i go? [23:11] I expect that a developer that received a veto comes back with an improved proposal. E.g. a new API command instead of a change of existing which can then be accepted. [23:11] if not, i will sign off [23:12] please continue the discussion [23:12] OK. Have fun Peter. [23:12] i am looking forward reading the logs [23:12] ciao Peter [23:12] PeterThoeny: see you later, have fun :-) [23:12] san francisco cable car, china town etc :-) [23:12] ttyl [23:13] cya [23:13] The major reason why I was against Crawfords proposal was that it was a 70 hours per week job. Not realistic. The new roles can be devided on several people and rotate as required. And the roles are more a maintainer role than a power role. [23:14] Looking at new proposals and make sure that if someone is against - it is added to the agenda of a release meeting is not a one person power role. [23:15] I think we can agree that overall this might just work, but we should add a few more words on the veto role? [23:16] veto action, that is [23:16] yes [23:16] Yes. We can try and summarise what was discussed today on the proposal. [23:16] And refer to the minutes of todays discussion. [23:16] action item for you, Lavr? [23:17] OK. I can do that. [23:17] excellent - thanks [23:17] ok, then .. let's backtrack to the agenda? [23:17] we broke off at: "All: Try Configure script in DEVELOP and provide feedback for Crawford." [23:18] I can only say I am very impressed with this feature .. it has some issues with network timeouts when t.o. is slow, but it "just works" as I see it [23:18] I looked briefly at it. It looks good as such. But I am concerned if Plugins can only be installed this way. [23:19] AFAIK, they can be installed manually the same as today. [23:19] I know I cannot run build contrib type installer scripts behind our Motorola firewall. [23:19] yep, no changes there, unzip, enable in configure still works great .. this is just added usability / feature [23:20] (for the ones more comfortable with their browser than with a CLI) [23:20] Yes, currently. But I just want to make sure it stays that way. [23:20] But in general it is a great thing to make plugin installation easier. [23:21] Configure was a dammed good start. But there are obvious things that can be improved. [23:21] The initial LocalLib.cfg step could be avoided. [23:21] The many errors you see first time could be hidden until the first LocalSite.cfg is created. [23:22] they should! [23:22] ah, yep, that is an issue that still lives in Bugs [23:22] I am also not sure what "Warning:Unsaved changes will be lost!" means [23:23] all changes of the whole form? [23:24] Actually the two issues I just mentioned could be implented easily. Configure can know that LocalLib.cfg is in the same directory as itself so it could be a simple form field that the user fills out and submits. [23:25] And only showing ONE of the sections until LocalLib.cfg is present is also quite simple. [23:25] makes sense [23:26] should be reported / re-opened in Bugs .. but this is still feedback regarding the "old" configure - any showstoppers encountered on the new functionality? [23:26] * OliverKrueger has quit IRC (" HydraIRC -> http://www.hydrairc.com <- The dawn of a new IRC era") [23:27] I need to try again. I had other stability issues last time because DEVELOP branch has been - and still is shakey. [23:27] ok, makes sense .. [23:27] I will do a test as well [23:27] But as a general statement, there is no need to tell CDot anything else than - go ahead - and here is our additional wishlist. [23:28] agreed on that [23:29] ok, that covers "Review Previous Action Items" [23:30] And we covered 2 [23:30] any point in going through "Review proposed items", with the proposing parts not in situ? [23:30] * PeterThoeny has quit IRC (Read error: 110 (Connection timed out)) [23:32] PluginHandlerForContentMove has not been covered. [23:32] But that was Peters [23:32] AutoIncTopicNameOnSave - same [23:32] TemplatePathIsCounterintuitive - Martin is not here to defend it. Arthur - you have any comments? [23:33] I don't [23:34] (looking at the topic) [23:35] well, Peters (contentmove/autoinc) are ~no-brainers to me and in the new setting I think they would just run straight through [23:36] did you try out http://twiki.org/cgi-bin/view/Codev/PreInstallSmartEditAddOn, Lavr? [23:36] Plugin Handlers has always been controversial - adding new has always been at least challenged. [23:37] AutoIncTopicNameOnSave needs a little more debate after the vacation before acceptance. But probably will end up with acceptance in some form. It is a matter of which spec. [23:37] SmartEditAddOn. Yes. I installed the very first version. And I am already addicted. It is a great work. [23:37] PreInstallSmartEditAddOn: I would really prefer more support for Safari [23:37] I still need to test performance where I run it via my ADSL. [23:38] It is not finished work. But the concept is right. [23:38] ArthurClemens: Makes sense .. any assessment on what fails / what kind of work is needed to get it right with Safari? [23:39] Styles does not work, and neither does link insertion [23:40] Maybe the developer need a virtual machine with a Mac OS and Safari to test it. Is such available? [23:40] ok [23:40] also icon, search [23:40] select whole line [23:40] auto fit [23:40] Lavr: Unfortunately not, but the idea is cool :-/ [23:41] ArthurClemens: OK, sounds like its not just a minor thing [23:41] Since so much does not work - it must be something basic that is broken. [23:41] available: Mac Pro? [23:42] I tried to look at the js, but there are many files [23:42] I have not owned a Max since I had a powerbook 100 15 years ago. If the guy cannot run Safari he has no chance to get it to work. [23:42] s/Max/Mac [23:42] but it is at Colas' office [23:43] they surely have a mac lying around [23:43] Yes. But do they have Macs? [23:44] I surely could not find one in Motorola Copenhagen in year 2006. Sun, HP, Windows, Linux sure. Mac? No. [23:44] I've asked for Safari support in the plugin topic [23:45] How is Safari licensed? [23:45] Proprietary or open source? [23:45] WebKit is open source [23:45] WebKit? [23:46] http://nightly.webkit.org [23:46] the framework that OS X (and Safari) use [23:47] Also Shiira is based on it: http://hmdt-web.net/shiira/en [23:47] I ask because if it is OSS and runs on BSD then maybe a VM machine could be used as test client. [23:47] no chance [23:48] needs Mac OS X [23:48] we need somebody with a mac os x machine to setup a VNC connection for the developer, but that's not that likely to happen / nor a very fast way of working [23:49] I can have a look, maybe I can find something in the code [23:49] ok, time is running out for me also [23:50] I get errors on Styles for instance: Value undefined (result of expression input.createTextRange) is not object. [23:50] http://twbw-test.luddeni.net/pub/TWiki/SmartEditAddOn/twikismartEngine.js [23:51] anything left to do now? [23:51] This Addon has a strong potential to become standard part of TWiki so it is worth the effort. But it has to proof itself strong enough and must support all major browsers reasonably. [23:52] hmmm, http://www.getswift.org/ is not useful for Safari comparison or? [23:52] MainTWikiPreferencesOverridePluginSettings - is still in discussion on COdev so nothing to decide yet [23:53] interesting, getswift [23:54] We are through the agenda of things that we can discuss with the people present. [23:54] let's call it the day then [23:54] I will run along then, I'm flat out for tonight [23:54] so am I [23:55] nice meeting, hope we will see start to see some more developer action again these coming weeks :-) [23:55] see you later [23:55] Me too. Nice meeting. For the record I am on vacation in France next week and next again. So if there is a meeting in two weeks I will not be there. [23:56] I will upload the log tomorrow. too tired now. [23:56] Bye all [23:57] have a nice holiday! [23:57] bye [23:57] * ArthurClemens has quit IRC ("Leaving") Session Time: Tue Aug 15 00:00:00 2006 [00:06] * Disconnected Session Close: Tue Aug 15 00:06:47 2006