Session Start: Mon Aug 18 13:45:28 2008 Session Ident: #twiki_release [13:45] * Now talking in #twiki_release [13:45] * Topic is 'http://twiki.org/cgi-bin/view/Codev/GeorgetownReleaseMeeting2008x08x04' [13:45] * Set by Lavr on Mon Aug 04 14:55:38 [13:45] * Lavr changes topic to 'http://twiki.org/cgi-bin/view/Codev/GeorgetownReleaseMeeting2008x08x18' [14:14] * eset has joined #twiki_release [14:18] * MartinCleaver has joined #twiki_release [14:18] Please ping me when the meeting starts [14:32] * MartinSeibert has joined #twiki_release [14:33] * iradel has joined #twiki_release [14:43] * CDot has joined #twiki_release [14:50] * AndreU has quit IRC [14:59] Hi all! [14:59] Hi there. :) [14:59] Hi all [14:59] Hi there [15:00] howdy [15:00] TWO-Servers are in "sprinting mode" once more. :-) [15:00] twiki.org hanging as usual. I have the feeling the persistant perl thing installed on twiki.org makes things worse. It may speed up on average but once overloaded it makes things worse [15:01] It is always some speeddy_backend processes have have hung for minutes that put twiki.org on its knees [15:02] * AndreU has joined #twiki_release [15:02] hi crawford, eset (?), iradel (?), kenneth, martin c, martin s, oliver, rafael, sven (here?), markus [15:02] hi andre [15:02] very high server load again [15:03] let me check... [15:03] hey [15:03] hi. just observing today. [15:03] Hi Peter [15:03] Hi iradel [15:04] two speedy backends had over 200mb [15:05] eset & iradel, could you quickly introduce yourself, and also state your twiki.org wikiname? [15:05] * gwash has joined #twiki_release [15:05] hi gwash [15:05] hi gwash [15:05] hi, adam hyde from flossmanuals.net AdamHyde [15:05] peterthoeny: hi :) [15:05] Hi eset! [15:06] gwash: welcome here, could you please introduce yourself and state your twiki.org wikiname [15:06] ah, hi adam, long time no see! [15:06] twiki.org is loading for over 3 minutes now. :-( [15:06] :) [15:06] will be in berlin too [15:06] MichaelRubin from Chicago. Professionally I'm a videogame developer and am doing my best to avoid learning Perl. Just installed TWiki for an organization of elderly people that are trying to put together a website and they loved my pitch. [15:06] ( == they loved TWiki ) [15:06] well, my name is M Rawash, and i just discovered/installed TWiki yesterday, and i don't have a twiki.org account =) [15:07] gwash: Get one. :-) [15:07] MartinSeibert: sure, will do [15:07] :) [15:07] ... as soon as loading times will calm down eventually ... :-) [15:07] welcome michael g and m rawash [15:08] I have started on the minutes [15:08] Peter: When will we see results on that issue? That is a pretty bad picture especially for new folks. [15:08] thanks [15:08] i also use twiki as a PIM at the moment [15:08] ok, i put one offending ip address on the deny list [15:08] i'm a student [15:09] gwash: I use TWiki personally also. Very good for keeping information and documents together. :-) [15:10] Someone broke the lock on the minutes page. Merge hell again [15:10] Lavr: you are right. IME mod_perl is a much better solution at high loads. [15:10] MartinSeibert: yea, i specially like the drawing java-applet =)) [15:10] it's what won me over, when i was torn between twiki and dukowiki [15:11] time check: +11 min [15:11] shall we start [15:11] ? [15:11] sure.. [15:11] sure [15:11] who is facilitating? who is taking notes? [15:11] As said I am already on the notes [15:12] i can faciliate [15:12] thanks kenneth [15:12] Thanks Peter and Kenneth. [15:12] proposed agenda:# 1. Feature requests for Georgetown Release [15:12] anything to add? [15:13] First item is http://twiki.org/cgi-bin/view/Codev/ModPerlStartupScript [15:13] ---++ 1. Feature requests for Georgetown Release [15:13] (server load is now down to 1 again) [15:13] * Lynnwood has joined #twiki_release [15:13] I would like people to review the last version one more time and if there are no worries then we should add this script to the tools dir already from 4.2.3 [15:14] no brainer. But the script, and the ApacheConfigGenerator, need docco improvement [15:14] several newbies have been caught out by it recently [15:15] (mod_perl, I mean) [15:15] Ah. Yes the mod_perl part requires a little more docu I agree [15:15] I can produce that. But I need some developers to review the last version of the script itself. [15:15] yes, agree to ship this with 4.2.3 [15:16] +1 to ship it on 4.2.3 [15:16] +1 to ship it on 4.2.3 [15:16] +1 from me, too. [15:17] Joshua's script is much busier than the one I use, and I'm not sure what most of it is doing. but with doc, +1 from me. [15:17] CDot this is why I want it reviewed one more time. If it is full of stuff that does no good it make be better to slim it down [15:18] I understand it now. Looks OK. [15:18] one potential gotcha: $binurlbase and $binbase need to be set [15:18] that should be documented well [15:19] ?? [15:19] my $binurlbase = '/twiki/bin'; # must be set by the user my $binbase = '/var/www/html/twiki/bin'; # must be set by the user [15:19] Ah. Yes that is what a config generator can do much better than a distributed script. [15:20] Lavr: why not build it in =configure=? [15:20] But couldn't a good default work in 90% of the time. [15:20] configure - yes that is also a good idea [15:21] it would be nice to have a one-touch enable/disable mod_perl [15:21] It would be nice if you did not have to hack bin/LocalLib.cfg to do an installation also. [15:22] please make sure there is a standard place for this script. [15:22] This file hacking to install is really bad [15:22] The only obvious place for the script would be in the tools dir [15:22] agreed [15:22] Why do we have two cfg files to hack, anyway? [15:22] we don't [15:22] can't it be one to rule them all? [15:23] there is no need to hack any .cfg file [15:23] Isn't there LocalSite.cfg and LocalLib.cfg? [15:23] yes; neither needs hacking [15:23] ok [15:23] the defaults should "just work" (or used to, at least) [15:23] LocalSite.cfg is handled by configure but you still need to hack LocalLib.cfg [15:23] no you don;t [15:24] the default "just works" as long as you don't have any exotic path to include [15:24] Why it is there then? It has ONE line you need to edit when you install. [15:24] i have to admit that is a bit confusing [15:24] probably a discussion for later [15:24] it does 'just work' without hacking [15:24] The default in LocalLib.cfg does NOT work [15:24] take this into Codev; not a feature request [15:24] but some options are hidden in the configure interface [15:25] time check: +25 min [15:25] Lavr: it worked here, after running configure [15:25] I haven't hacked LocalLib.cfg since my first 4.0 install [15:25] i was talking about manual hacking [15:25] I have a topic for the end of this session: "How to fund a developer for TWiki.org-development?" Peter: Could we put this on the agenda please? [15:25] Why don't we remove that file then? And why is it still in the install manual that you have to hack it? [15:25] let's discuss this in Codev [15:26] OK. I will create a Codev topic about LocalLib [15:26] next [15:27] Lavr: ah, i was confused between LocalLib and LocalSite, LocalLib does need hacking.. [15:27] Next is Database store - in general [15:27] kenneth: shall we talk about the toc first? less time than data store [15:27] gwash yes I also experience that nothing works without changing LocalLib [15:28] Actually I do not want to spend a lot of time on Database storage [15:28] ok, then first datastore [15:28] http://twiki.org/cgi-bin/view/Codev/DatabaseStore [15:28] Lavr: i think it's a bad idea, text storage is part of what makes TWiki attractive to people who don't wanna bother with LAMP [15:28] What I want is to kick off some Codev topics the next days to address this subject [15:28] gwash: we are not talking about *replacing* the text store, but of offering a DB store as an alternative. [15:28] ah good! [15:29] gwash - we are many that do not want to let go of the .txt files and the database store many of us consider does not replace the txt storage [15:29] CDot: ah, np then :) [15:29] we would like to be able to select the store impl on a web-by-web basis [15:29] a database store or a database cache is needed for some large twiki deployments [15:29] there are some with over 400k pages, over 1000 webs and over 10k users [15:29] Actually we are not at all alligned what we mean with Database store [15:29] Which is why we need some Codev activities started on this [15:29] no. Start your Codev topics. :-) [15:30] I will make a page similar to the Governance topic overview where we can collect what we have and start new [15:30] sure [15:30] are we talking about a "DatabaseStore" for the current topic format? [15:31] maybe [15:31] I doubt it, though [15:31] The current storage is very mature. I, for one, would have used the DatabaseStorage if it had been an option @ installation... but is there a database expert who has implemented it? What are my guarantees before using DatabaseStorage for production? [15:31] that's one of the key points to clarify. [15:31] none of my implementations uses the existing storage format [15:31] but there may be good reasons to use it. Dunno yet. [15:32] I think a combo can make miracles. [15:32] by topic format I mean META and TEXT, where text is one single entity. [15:32] I would prefer a database storage from a management perspective. I simply sounds mor professional. I guess that most managers will think this way. [15:32] Soronthar: I know what you meant [15:33] more [15:33] just for techies, the main problem is how to handle the fact that TWiki has what is essentially a dynamic schema [15:33] that is not a problem, actually [15:33] we have several approaches, from recomputing the schema on the fly, to using schemaless DBs [15:33] as long as the database support "alter table add column" [15:33] there are always tradeoffs [15:34] The key goal for me is to provide TWiki with a fast way to look up access rights, form data etc without having to crawl through 10000s of text files. [15:34] key points as i see: performance (page load time, search, backlinks...), compatibility (such as easily inherit twiki content from an acquired company, or from 5 year backup for an audit), simple architecture (code maintenance) [15:34] lots to discuss. In Codev. :-) [15:34] Dynamic schema sounds terrible. How can you optimize indexing that way? [15:35] iradel: it depends on what you mean by dynamic schema [15:35] * StephaneLenclu1 has joined #twiki_release [15:35] iradel: good question; and one with no good answers.... yet [15:35] and most important: complete audit trail, also for meta data [15:35] there should be no uses of "alter table columns" etc. [15:35] check out what I wrote on http://twiki.org/cgi-bin/view/Codev/ExtractingTopicDataToADatabase [15:35] Any kind of DatabaseStorage should ship with "the" .sql file you apply and that's your schema. [15:35] btw, will there be a way to switch to DB without jeopardizing data? [15:35] At least it is clear that we have different expectations, different ideas, different requirements. [15:35] hi stephane, thanks for attending the meeting [15:35] np [15:35] iradel: we can discuss this more on codev, but in my experience one "single" table for all forms just don't scale [15:36] But we all know the current data only in text files will not keeo flying. That is a good start [15:36] Hi all! [15:36] Soronthar: You can use a level of indirection... [15:36] Hi Stephane! [15:36] the product my company produces is based on dynamically created schemas [15:36] Soronthar: modern DB's are good at joins....that's what it's for. No-one in production uses dynamic schema.... [15:36] and we tried that approach before... [15:36] what do you mean by dynamic schema? [15:36] It is also a clear goal for me that this gets defined in small chunks that many people can address each their share of [15:36] iradel: I think that we agree, but are talking about different things :) [15:37] Lavr: very good idea from my view. [15:37] I would like to suggest that this technical discsussion can continue on #twiki-dev, so we can continue with this meeting....? [15:37] sure [15:37] We know the action. Give me 60 seconds to update the minutes [15:37] One last note: will this be supported on mysql, postres, oracle, sql server, etc.? people use different things... [15:38] anything supported by DBI, at minimun [15:38] hopefully all, if we use DBI. But that's a valid question. [15:38] someone will try to use ms access... [15:38] iradel: shoot them now [15:38] CDot: agreed. [15:38] we can't help that poor soul.... [15:38] http://twiki.org/cgi-bin/view/Codev/TocFailsForIdenticalHeadingNames [15:38] I am against MS-Access-support! :-) [15:38] iradel: definitely not oracle, mysql sounds right [15:39] iradel: /join #twiki-dev [15:39] MartinSeibert: +1 :P [15:39] there IS a DBD driver for MSAccess! [15:39] old problem, TOC? [15:39] if it can be fixed, then I vote for it, sure [15:39] Actually a lot of teams are using MSAccess (even instead of SQLServer, quite frightening) ;) [15:39] it's been bugging me for years [15:40] markus: you created a patch for the toc issue [15:40] TocFailsForIdenticalHeadingNames - not a feature proposal to decide on but Markus need feedback on proposed fix I guess [15:40] thanks! [15:40] Yes, exactly. [15:40] are teh same anchors created as before if there are no identical headings? [15:40] Yes that was a requirement we identified [15:40] Yes, no change. Only clashing anchors will be renamed. [15:41] cool [15:41] Cool. [15:41] sounds good [15:41] feedback on what item do you need? [15:41] I like that. Had the problem already. [15:41] Sounds good to me too [15:41] Seemed to "low" to report. Cool, that this has been solved now! :-) [15:41] Need numbered anchor name for that [15:42] I couldn't think of any sideeffects from the change I made to Render.pm, but maybe someone could play with that. [15:42] Martin we already has 3 or 4 bug reports about this. A very welcome fix to get. [15:42] The main reason I wanted others to take a look was because it's quite some time the topics were written. [15:42] Anything special for us to look for in the fix Markus? [15:42] check it in, we can fix bugs if there are any :-) [15:42] Nice that this is getting fixed :) [15:43] Yes, after I filed the fourth redundant bug report, I thought, I should start implementing it ;) [15:43] Ok. [15:43] We have so many broken TOC on our corporate wiki [15:44] ok, next [15:44] I have used ---++ Status as a work around for years because of this bug [15:44] markus, just reading the code quickly that makes sens to me. The only note of caution I would have is that some plugins may depend on consistent anchor name generation [15:44] Markus I guess this means - check in the fix on both trunk and TWikiRelease04x02 branches so we can start testing it [15:44] and there is a risk of breaking them if you change the anchor format [15:44] (doesn;t look like you have, just thought I would mention it) [15:44] these are all three feature requestes listed in the minutes topic. any other items? [15:45] I only change the anchor names if they are redundant, no. [15:45] that's what I thought [15:45] MartinCleaver has an item that he wanted to discuss but not put on the agenda [15:45] making VarVARTopicCapable [15:45] * MartinCleaver digs for the URL [15:46] http://www.twiki.org/cgi-bin/view/Codev/MakingVarVARTopicCapable [15:47] VAR should be capable of returning variable values from the topic-level too, by a natural extension of the syntax. It provides an implementation and contrasts this simple extension against the Query Search mechanism introduced in 4.2.0 . [15:47] adds: # Syntax 2: %VAR{"NAME" web="Web" topic="Topic"}% [15:47] # Example: To get %ALLOWTOPICVIEW% of the Main.%USERSTOPIC% topic write %VAR{"ALLOWTOPICVIEW" web="Main" topic="%USERSTOPIC%"}%, [15:48] spec is sound [15:48] Not sure I had "objected" to it per se [15:48] I would love this feature. Why did you raise concern against it yourself? You have a doubt? [15:48] looks fine, with one caveat; please don't use TWiki::Func from core modules. Extensions and plugins only. [15:49] No, I think its all quite reasonable, actually [15:49] oh [15:49] how would you do it? [15:49] The "effective permissions for the current page" trick is really cool, BTW [15:50] time check: +50 min [15:50] i have a hard stop in 10 min [15:50] * MartinCleaver really wishes someone would enable subscribe plugin for TWiki.org [15:50] Ok. So we pass this? [15:50] accept? [15:50] yes. You seem to have everyone for. So Accepted [15:50] +1 from me. [15:50] i would say the accepted [15:50] +1 [15:50] +1 [15:50] +1 [15:51] good! [15:51] Next! [15:51] * MartinSeibert wishes someone would update TWiki.org to the current version. :-) [15:51] any other feature item to discuss? [15:51] Yes, I'd love to see that, too. [15:51] Not today. [15:51] ---++ 2. Brainstorming [15:51] without subscribe we lose the periphery [15:51] No more I guess. Peter why not give a short status on the t.o update? [15:51] anyhow [15:51] I have a topic for the end of this session: "How to fund a developer for TWiki.org-development?" Peter: Could we put this on the agenda please? [15:51] carry on [15:52] i'd like to give a quick update on the new servers [15:52] Peter: Yes, is there an update on the way? [15:52] we finally received the x86 box [15:52] we racked it over the weekend [15:53] now need to configure os, apache and install twiki [15:53] i will ask sven if he can help [15:53] OS is Centos? [15:53] it shipped with solaris [15:53] would that be a good opportunity to switch to 4.2.x "on the fly"? [15:53] Will you update to Twiki 4.2.2 or is that too much effort? [15:53] 8gb ram [15:53] uberall: dito. :-) [15:54] yes, that is the idea, to install 4.2.2 [15:54] Can we please drop that Solaris/sun stuff and go for mainstream Linux we all know? [15:54] 8 GB RAM is cool. :-) [15:54] in fact, i already have 4.2.0 installed on the t5220 [15:54] peterthoeny: if you install stock 4.2.2, fix the rogue

:) [15:54] ok [15:55] * MartinSeibert would love to see a fast and up to date twiki.org. The world deserves to see it! [15:55] tonight we will do a firmware update on the nas [15:55] We have been showing off old stuff for too long now. [15:56] ok, that is all on twiki.org server for now [15:57] martin s: go ahead on the funding question [15:57] Okay. [15:57] let's brainstomr on this a bit [15:57] Looking forward for increased performance on twiki.org [15:57] I would like to have a kind of virtual box for donations. [15:57] I would put in 500 Euros each time, TWO does something cool for //SEIBERT/MEDIA. [15:58] There have been several things, that would have been worth it. [15:58] Just to name some: [15:58] maybe some kind of "I'd love to see this certain feature and would spend x$ once it's ready", too? [15:58] 1. Bring in cool new projects for TWiki-implementations for //SEIBERT/MEDIA. [15:58] 2. Raise customer awareness. [15:59] 3. Generate leads like SIEMENS and other companies for //SEIBERT/MEDIA. [15:59] 4. Create a very cool intranet element for our own organization. [15:59] i think it would be good to have some kind of formalized direct compensation to developers, tied to a particular feature [16:00] what sort of projects? [16:00] The reality is: There is no other way for //SEIBERT/MEDIA for giving back time of a lame CEO how can't program a line at all. Bad deal for TWO. :-) [16:00] i would stay away from a pool of money that needs to be distributed for work already done (tax implications, who gets how much, ...) [16:00] I am not fishing for compliments. :-) [16:01] It would be cool, if we could give back some development resources from our programmers. But I cannot at the moment. [16:01] one recent example is the expected webdav work [16:01] there are a few people willing to fund the work [16:01] Peter: You end up getting nothing to share this way. [16:01] Peter: wrong! [16:02] I think I wrote a proposal in Codev that may help, let me see.... [16:02] I believe, that there are loads of other companies, that feel a duty to give back. [16:02] How to leverage this potential? [16:02] Even if it does not exist, we should try! [16:02] I also wrote a proposal on that. [16:03] found it : http://twiki.org/cgi-bin/view/Codev/AllowDevelopersToCollectDonations [16:04] i think it would help if we formalize the work requests [16:04] Mine is here: http://twiki.org/cgi-bin/view/Codev/TWikiFoundation#Funding [16:04] a per-developer donation sounds good imo [16:04] Peter: I do not want to care who gets the money. [16:04] yes, a per-developer donation is good [16:04] I just want to unload the duty we have. [16:05] Yes. Direct contact to developer will work. Foundation means someone decides who gets = war! [16:05] peterthoeny: that's what is discussed on http://twiki.org/cgi-bin/view/Codev/CodersForHireTalk [16:05] agree with kenneth [16:05] i need to sign off now [16:05] please continue this discussion [16:05] (I don't know whether this is the proper forum, but) I would suggest that an html/css/js/etc. validation is adopted as part of the pre-release process. Two releases went out where the pages (blatantly) did not validate. [16:05] thanks all! [16:05] g'nght, Peter [16:05] What a pity. [16:05] Bye Peter. [16:06] bye, Peter. [16:06] Ciao! [16:06] I cannot make myself plainer, can I? [16:06] just one other thing, why does twiki ship with most of its files without write access? [16:06] Martin did you read up the CodersForHireTalk? [16:06] You will just have to stick with me ... :-) [16:06] MartinSeibert: would you like to borrow a hammer? [16:06] is there some reason behind that? [16:07] I did look at it, yes [16:07] cdot: What effect will it have? :-O [16:07] MartinSeibert: stress relief ;-) [16:07] Please, don't hit the new server (even if it's running under Solaris)! :p [16:08] If a donator puts forward a "I want to give xxx Euro for this feature" then it is up to someone to take the task and do it and collect the money. [16:08] cdot: I save the money and remain with my will ... [16:08] I'm afk having dinner. Just wanted to say that I have good hope to be in Berlin on the 4th... but have nothing organized yet. [16:08] Lavr: Did you ever donate for wikipedia? [16:08] I did. [16:08] Did you ask, what the would use the money for? [16:08] I did not? [16:08] Did you donate for Firefox 3? [16:08] No but I donated for others but always to a developer. [16:08] I did. [16:08] well, i find the process of raising money for work quite exhausting [16:09] Did you know, what they would use it for? I do not. [16:09] for example, i recently put several hours of work into a quote for people who said they wanted a webdav solution [16:09] since then... silence [16:09] It's as easy as allowing it. [16:09] so, I guess I wasted a couple of hours. [16:09] I would not mind that a foundation would exist under different circumstances. But I cannot see WHO would distribute your money for you without becomming a shooting target [16:09] very unfortunate. [16:09] cdot: Where is it? [16:10] TWiki:Plugins.WebDavPluginRequirements [16:10] I have rather money distributed unevenly than no money to distribute at all. [16:10] Don't you see, that you exchange no money with your attitude. [16:10] http://www.twiki.org/cgi-bin/view/Plugins/WebDAVPluginDevRequirements [16:10] How can this be the better solution? [16:10] * MartinCleaver just commented: I'd be surprised (but delighted) if the TWiki community has the spare manpower needed to create a replacement for elance.com ; I'd suggest we pick one of those, but to try to work out how good TWiki devs would not lose out to established elance devs who have comparably little or no TWiki experience. [16:11] does sun donate twiki.org's server, or did someone actually decided to run solaris? [16:11] That would make the TWikiFoundation efectively a "company" [16:11] with 'employees' [16:11] gwash: donated [16:11] sun donated. [16:11] CDot: ah, that explains alot :P [16:12] Firefox and Wikipedia have companies behind that receive and distribute the money. [16:12] right. [16:12] yes, but they buy contracts, which is fine [16:12] Lavr: Please explain again what your fear is. [16:12] I would really like to profoundly understand the threads. [16:13] We are maybe 15 active - really active people developing and being active on the TWiki project... [16:13] So 2-4 of us would sit managing a foundation that gives money to the others. [16:14] Do you think it is that much of an effort to have 2-4 people pondering about that? [16:14] RIght now we cannot write a press release without having a small world war 3. How do you think the mood will be when someone gives money to one and not another? [16:14] depends how it is done [16:14] Who is there to program? [16:14] if there is a transparent process, then no problem [16:14] cdot. [16:14] Sven [16:14] Michael Daum [16:14] if it is nepotistic, or behind closed doors, then.... [16:14] and a couple of others. [16:15] They have work that competes with their TWO-activity. [16:15] cdot: transparent process +1 [16:16] * eset has quit IRC (Read error: 113 (No route to host)) [16:16] This is my concept on the distribution: [16:16] Distrubution by Voting on Funds [16:16] As it is the will of the community not to let few individuals decide on the distribution of funds, it should be a task for all TWiki-members. At first every Community-Member should be able to vote for or against a proposed project. This is how funds should be distributed: [16:16] * Every project is measured in TWiki-funding-points. Each point is worth 25,00 USD. [16:16] * The measure of TWiki-funding-points is defined by the shortest amount of time, that a community member evaluates for the completion a project. Example: "I think this boring work would be fulfilled be MaxMiller? in 2 hours. It should be worth 2 points or 50,00 USD." [16:16] * Projects can be granted by the community before or after their completion. But each proposed project has to be proposed within 1 month. [16:16] * A project proposal can be raised by the project members themselves or by others. [16:16] * A project is not funded until the community had 7 days to vote on it. [16:16] * A project is not funded if 3 community members object on funding it. There is no need to argue or reasoning about objections. [16:16] * A project is not funded if not at least 5 community members, other than the one proposing it, approve it and vote for it. [16:16] * A project cannot be funded, if the PayPal? -account does not have enough balance. In that case all approved projects are funded in chronological order as money "arrives". [16:16] * After 6 months projects, that could not be funded expire. [16:16] * Everybody can vote on a project until it is approved by the above mentioned rules and can be funded from the available TWiki-funds. [16:16] * Projects are not funded, if the project members object to accept the money. Then the money goes back to the fund. [16:17] sidenote: Is there anyone here who can give me a Codev tarball to work on a script for CleaningUpCodevRevisited? [16:18] Lavr: But you might have read that already. So distribution is a thread. I agree. [16:18] What else? [16:18] MartinSeibert: how about distributing donations per-line of code? :P [16:18] I like the sound of that ;-) [16:18] gwash: there are more projects in t.o that code projects. [16:18] I had not read that and this is not the foundation proposal I originally opposed with 2-4 people deciding who should be funded. [16:19] ie, documentation, release management, etc. [16:19] Lavr: so you like it? :-O [16:20] one point is that funded foundation projects should be the exception, not the rule. This is still primarily an open source, volunteer, project [16:20] cdot: That is only a question of how many money the fund has. [16:20] and the danger is that once you start funding stuff, all people will want their work funded. [16:20] If we had 200K. Why not pay everybody for lame work? [16:21] CDot. yes the minute noone will do anything without money the project is not a place I want to be part of anymore. [16:21] If we have 300 Euros. No one will care about it either. :-) [16:21] Hm. [16:21] All I'm saying is that we must be careful not to build up expectations [16:21] stating the obvious, really. [16:21] I will be happy to see a project I helped to grow fly away faster than I can see. :-) [16:22] besides [16:22] cdot. Agreed. [16:22] There will not be loads of money. [16:22] Yes. funding should be for the BIG things which do not get done unless someone can work full time on it for a while. [16:22] those that would require funding to shift focus are those that are earning money from TWiki. [16:22] But there is the objection rule. [16:22] * A project is not funded if 3 community members object on funding it. There is no need to argue or reasoning about objections. [16:23] TWiki can't be the only project with this problem [16:23] Imagine being one of those that objects to a funding. I would end up not on a T-shirt this time but as a tattoo on someones butt. [16:23] Soronthar: At the moment people like cdot and Sven do the coding and I get the lead. Is that fair? Hardly ... :-( [16:24] I know, that I am not the only one to get requests, but still the status quo is not paradise! [16:25] Lavr: You will have to explain yourself, which you can perfectly today also. :-) [16:25] And it's three tattoos on the butt. :-) [16:26] Lavr: Can you describe a situation a funding would be okay for you? [16:26] Lavr: Let's assume we have fair money distribution and little organization efforts. Would a foundation then be okay for you? [16:26] Yes. I can go back in time. The work Crawford did on Wysiwyg for 4.2.0 was partly fonded. And It was a huge task and a lot of work. [16:27] It would have been good if ALL that work had been funded and more of us could have contributed [16:27] * CDot can't argue with that [16:28] In future I see some chunks of the DB storage being funded work where people like Sven and Crawford could dedicate a full month or two for the work. [16:28] Hm ... [16:28] Who is interested in funding a database ... [16:28] It is difficult for ONE small company to fund it all alone but if we can coordinate the funding then it may be possible [16:28] ? [16:29] The database storage is not about DB. It is about scaleability. [16:29] Lavr: Correct. Just let them give back, for what they took. [16:29] They got the cool TWiki-software? [16:29] MartinSeibert: I have a couple of clients who might pitch in. But they have already given a lot back [16:29] They are not waiting for a database. They just have gotten a solution for their problem. [16:29] the challenge is to tap into the reservoir of people who *don;t* pay [16:30] exactly. [16:30] cos right now, they have no easy way to make donations [16:30] They just dont care about the future. [16:30] and their individual sums are too small to make a difference [16:30] I don't see, what Wikipedia might do with my money. Their software is cool. The servers are fast. [16:30] again, Wikipedia has employees. [16:31] When you donate to Wikipedia you have a pretty good idea what you donate to. [16:31] Same with TWiki. [16:31] I would donate for the future development of TWiki. [16:31] If I was to let Motorola pay for development I would want to know what I got for it. I would not fund something I do not need. [16:31] To keep it up and running. [16:31] To keep my beloved intranet wiki. [16:32] http://wikimediafoundation.org/wiki/Donate [16:32] Lavr: You should calculate the benefits Motorola has from TWiki. That should be enough to argue why a donation is in Motorola's interest. [16:33] Just see in your server logs and stats and you will know, what I am talking of. [16:33] Motorola already donates a lot of Kenneth's time, which also has a high value [16:34] Yes I think Motorola has "paid" our part as "work" yes but I could still see myself arguing to my manager to donate money for a particular feature we really want and which requires a lot of work [16:34] Sure. I did not want to push anybody. [16:34] I think the problems are twofold: [16:34] 1. it is hard to donate right now [16:34] 1. +1 [16:34] 2. there is no clear picture of what you are donating to [16:34] 1 impossible is the word [16:34] 2 +1 [16:35] (no roadmap, or pie chart like wikipedia has) [16:35] and obviously no support for decision-makers, to help them see the benefits of donating [16:36] The biggest problem is in my opinion, that we do not start on 1. [16:36] If we did, the rest would evolve. [16:36] Yes. That is what I tried to address with my follow up on CodersForHireTalk [16:36] If we do it plain and transparent. No pressure will arise to do it fast. [16:36] We could start to simply collect the money. [16:37] And as it grows, we can think how to distribute. [16:37] When you are a small company you can just donate on paypal. In Motorola to be able to do this I would have to. [16:37] If it does not grow, no distribution is possible at all. [16:37] 1. Get the vendor registered in our purchasing system Tigers including scanned logos etc [16:37] 2. Get a written quotations [16:37] 3. Issue a purchase order on the Tigers system [16:37] Lavr: There is no need for a Motorola-donation. [16:38] 4. "Receive" the delivery on the system. [16:38] Lavr: actually, motorola has more discretionary routes than that, through community support programs. But I understand your problem. [16:38] Companies could get a receipt for their donation afterwards. That is enough for most, I believe. [16:38] I could do all this with CDot as a company but I could not just click on a donate button and do a paypal payment. [16:39] Motorola will not pay an invoice unless it is from a registered company in our system and there is a PO number to match. [16:39] Let's focus on this: CDot 1. it is hard to donate right now [16:39] Redtape beyond belief but this is reality in large corporations [16:39] Or in your words: impossible. [16:39] not impossible; difficult [16:39] Why dont we simply change that? [16:39] you have to do it directly to someone like me [16:40] No I am just saying - I need a commercial entity to work with - but I could do it and may very well do it in future. [16:40] Or Sven [16:40] Or Michael [16:40] Or Peter [16:40] Or Lavr [16:40] MartinSeibert: I think the issue is the commercial entity that would be required to gather the cash [16:41] Exactly! [16:41] My assumption is: Peter fears that this entity will overtake TWiki.net. [16:41] I made the point that incorporating such an entity is easy in europe, but there is still a lot of resistance to the idea [16:41] It might just get much bigger and raise much more money. [16:42] dunno. you will have to ask Peter that. [16:42] At the moment a lot of people think, that Twiki.net runs Twiki.org. [16:42] I will. [16:42] I understand that Peters fear is not TWiki.net related but related to WHO would sit in this commercial entity. Same fear as I have. [16:42] CDot: i think that's best idea, as far as opensource project funding goes [16:43] get a big company to "adopt" you [16:43] Lavr: What is "commercial" in that entity. [16:43] I would not want to sit in it for sure. But hell. If some of you make a foundation NOONE will or can stop you. Just do it. [16:43] It would not be commercial. [16:43] Wikipedia isn't. [16:43] I will not do anything without community support. [16:44] this is all tangled with the governance question. If there is a group deciding the roadmap for TWiki, shouldn't they also gather and disburse the funds? [16:44] And to gather that it surely takes your and peter's approval. That is a fact. [16:45] Martin. Nothing prevents you from creating a foundation registred in Germany that has the purpose to collect and distribute fonds for TWiki development. [16:46] true. But it's worthless without a "Donate Here" link on t.o [16:46] Lavr: It is just about 10 hours of work, that I will not invest before having a go from you and Peter. [16:46] For roadmap I prefer that no Boards or Councils or other elite decides that. I perfer this to be a community owned task. [16:46] so martin is right, it has to have support [16:47] Go from me and Peter. Hmm. Do not put power on my person I do not have. The only thing that makes me anything on this project is the hard work. I am nothing special on this project. [16:47] Lavr: Dont be shy. :-) [16:48] dito :) [16:48] you are an interim TCC member. That makes you special in my eyes [16:48] you have the future of the project in your hands [16:48] That is the ruke of any OSS. Work hard - have influence. Do nothing and have less influence. Fair I think. [16:48] Lavr: Just tell me that you like the idea. I will interview Peter again, afterwards. :-) [16:49] I will not mind anyone creating a foundation. _I_ just do not want to sit in its management. I do not want to become a shooting target when the foundation distribute money to someone and not to someone else. [16:50] Fair deal. [16:50] So you will not have me against you. [16:50] So it is down to one. :-D [16:51] I will copy a transcript of this discussion and send it via E-Mail to him. [16:51] I will ask him for a phone call on that issue. [16:52] There is a huge difference between the original foundation proposal and a foundation that distribute money to developers [16:52] points everybody to --> http://en.wikipedia.org/wiki/Commercial_open_source_applications [16:52] I have developers in mind. [16:52] That is what we all want. [16:52] The original proposal wanted Peter to give his TWiki trade mark and twiki.org domain to a foundation - that is a totally different game. [16:52] More cool features. Less bugs. Better usability. [16:53] Lavr: I do not care about Peter's possessions. TWiki will be named else if Peter decides so, but that does hardly affect the community. [16:54] All I want is a legitimate donation button. [16:54] +1 [16:54] It might, be that we raise 50 USD per year. [16:54] But I will also fight for that. :-) [16:55] If you do not get 50 USD I will pay it :-)) [16:55] If we have full transparency on the paypal account's balance. [16:55] at the rate it's going, might want to collect EUR [16:55] We will not need to distribute anything until we have raised 2500 USD. [16:55] And that might take a while. [16:55] And if not. Okay for me. [16:56] Then we drink it all up at a Summit :-) [16:56] Lavr: Yep. :-) [16:56] Hell, no! [16:56] $2500? What's that, 2.5 litres of beer? [16:56] We will be plain and transparent. Too serious for such jokes. :-) [16:56] ;-) [16:57] I will talk to our tax advisor. [16:57] gotta go; goodnight all, good discussion. [16:57] * CDot has left #twiki_release [16:57] Until today I thought, I could run away without having to do all this orga stuff. :-( [16:58] Yeah I also have to go. A thunderstorm is getting near so I may be without power soon. [16:58] Okay. [16:58] Thanks for the time. [16:58] Like in US all electrical power is hanging on ugly masts instead of being dug in the ground so everything goes when it is thunder here [16:59] This is by the way my last day in Montreal. I return to Denmark tomorrow and will be back in Copenhagen wednesday [17:00] * MichaelDaum has joined #twiki_release [17:04] Michael you missed the fun. [17:05] Hi Michael! [17:05] I just wrote a mail to the two of you and peter. [17:06] Gotta go now also. [17:06] Have fun Martin. Thanks for being an active meeting participant :-) [17:08] Bye. [17:10] * MartinSeibert has left #twiki_release