inside a td? [14:08] SvenDowideit: waiter, andother coffee please? [14:09] MichaelDaum: I have to investigate it ... [14:09] PeterThoeny: please take this offline and lets move on in the interest of time :-) [14:09] PeterThoeny: ---++ Item1332 The new web colour background of the top bar is not a welcome feature to all of us [14:09] MichaelDaum: Sven, Pam doesn't react? [14:09] SvenDowideit: look sdone to me :) [14:09] SvenDowideit: Pam's cleverly still asleep [14:09] MichaelDaum: yep the webcolor in the header was a nice try ... no more [14:09] PeterThoeny: i think we do not need o discuss [14:10] PeterThoeny: is part of arthurs action item: Review if WEBBGCOLOR can be shown better in Plugins.PatternSkin [14:10] Lavr_: Will reverted 1332 and we already had an action on an alternative [14:10] wbniv: Item1332 is already reverted, and has been changed [14:10] PeterThoeny: ok [14:10] wbniv: changed to the enhancement action item for arthur [14:10] PeterThoeny: ---++ Item1025 Flooded with warn messages from misc plugins (with fix diff attached) [14:10] PeterThoeny: done [14:10] PeterThoeny: sorryi am on the wrong list [14:11] CDot: I thought I closed that [14:11] PeterThoeny: ---++ Item664 TWiki.org Merge: TWiki Contributor should list area of contribution [14:11] SteffenPoulsen: there's a bit of doc in emptyplugin that needs a fix, this is minor - I'll commit that asap [14:11] wbniv: Item1336 Create new DefaultSkin [14:11] PeterThoeny: we talked about at length just before [14:11] CDot: SteffenPoulsen: coolio, thanks [14:11] wbniv: PeterThoeny: you skipped two [14:11] wbniv: and [14:11] wbniv: Item1328 RSS feed validation error [14:11] PeterThoeny: action item for steffen noted [14:11] SvenDowideit: 1328 is me - i need to investigate [14:11] wbniv: ok, cool [14:11] wbniv: Item1336 [14:12] PeterThoeny: the next few ones are wysiwyg [14:12] PeterThoeny: can we skip them? [14:12] SvenDowideit: one mo [14:12] PeterThoeny: or do you want to discuss some crawford? [14:12] CDot: only Engine bugs block the release [14:12] Lavr_: Yes Skip wysiwyg IMHO [14:13] SvenDowideit: can these be moved to not engine? [14:13] * CDot doesn't understand Item1336 [14:13] SvenDowideit: me neither [14:13] * wbniv raises hand, too [14:13] PeterThoeny: ok, [14:13] PeterThoeny: ---++ Item1336 Create new DefaultSkin [14:13] CDot: yeah, people keep reporting extension bugs without flipping the Extension marker [14:14] SvenDowideit: cool - i'll move them when i see them too :) [14:14] PeterThoeny: cairo had a default skin that was labelled classickin [14:14] CDot: ta [14:14] SteffenPoulsen: I'll move the WYSIWYG stuff from engine to Extension, my fault they're wrongly indexed in the first place - we can skip them for now [14:14] PeterThoeny: s/classickin/classikskin/ [14:14] MichaelDaum: hehe [14:14] CDot: we know what yo umean [14:14] * MartinCleaver smiles [14:14] Lavr_: Peter are you also having a drink? [14:15] PeterThoeny: no, too early [14:15] * MartinCleaver is on the Jameson [14:15] PeterThoeny: just chinese tea [14:15] SamHasler: I thought the idea of promoting Classic was that it should still the default skin if no other is selected, but that for template mantainance it was easer to have it as a full skin [14:15] SvenDowideit: me too [14:15] SamHasler: is that a fair summing up of what's happened? [14:16] wbniv: i'd prefer to have an even simpler and prettier skin, but that doesn't exist yet [14:16] SvenDowideit: y, something as ugly as the kwiki one :) [14:16] PeterThoeny: well, the problem is that we have now view.tmpl and view.classic.tmpl, edit.tmpl and edit.classic.tmpl, etc [14:16] PeterThoeny: that is we have now two sets to maintain [14:16] SvenDowideit: oh [14:17] CDot: no, we have two separate skins [14:17] ArthurClemens: No, we will do this for edin [14:17] MichaelDaum: yep [14:17] CDot: we have Classic skin, maintained for compatibility [14:17] PeterThoeny: one is the default template, activated if you select a ?skin=that_does_not_exist [14:17] ArthurClemens: edit.tmpl etc will be basic templates [14:17] PeterThoeny: and a real classicskin [14:17] SamHasler: yes, but if someone goes to ?skin=notaskin they still see classic [14:17] CDot: and we have the default templates, which are the fall back position [14:17] PeterThoeny: which gives us more work then we need [14:17] CDot: SamHasler: no, they see the default templates [14:17] SamHasler: oh, that's bad [14:17] CDot: which *happen* to be identical to classic at the moment [14:18] CDot: until we start changing them [14:18] PeterThoeny: the reason why i really think we should have just onbe default skin that is equal to the classic look [14:18] SamHasler: so why do it? [14:18] CDot: because we want to change the default skin [14:18] SvenDowideit: y [14:18] CDot: but we don;t want to break Classic skin [14:18] ArthurClemens: a better template mechanism [14:18] MichaelDaum: top prio should be to get pattern work [14:18] SvenDowideit: thus we are only double handling for the next week [14:18] SvenDowideit: until dakar is done [14:18] CDot: zactly [14:18] SvenDowideit: then the deafult ones are ready for edin* [14:19] SvenDowideit: and the contribs arch is done [14:19] CDot: I have default ones ready to roll [14:19] CDot: but not until edinburgh [14:19] SvenDowideit: with css tags i hope :) [14:19] ArthurClemens: Of course PatternSkin also relies on the basic templates still [14:19] CDot: yes [14:19] CDot: yes [14:19] SamHasler: so does anything need doing with Item1336 for Dakar? [14:19] PeterThoeny: so i do not understand why we needed to break the skin into two skins that are _almost_ the same [14:19] CDot: no [14:19] PeterThoeny: that is confusing to users [14:19] CDot: because they are only the same *for now* [14:19] CDot: the defaulkt skin will be changing [14:20] MichaelDaum: make pattern default for dakar [14:20] PeterThoeny: it makes sense no going forward [14:20] PeterThoeny: but why for dakar? [14:20] ArthurClemens: time [14:20] CDot: because I needed to make sure I could separate out, and independently test, classic skin [14:20] SvenDowideit: cos consistency is the bugbear of hobgoblins [14:20] SamHasler: remember it's default as in "fallback" not as in what you normaly see [14:20] CDot: right [14:21] PeterThoeny: right now it is really confusing for user because the default template is not even accessible [14:21] CDot: how do you work that out? [14:21] CDot: it's not confusing at all [14:21] CDot: it was confusing *before* [14:21] SvenDowideit: but PeterThoeny has a good point [14:21] SvenDowideit: customising and maintaining skins is already horridly confucing for users [14:21] CDot: when you referred in the docs to classic skin [14:21] SamHasler: PeterThoeny, my summary above as incorrect [14:21] SvenDowideit: the more files in the templates dir, the harder it is [14:21] CDot: but there were not *.classic.tmpl files [14:21] MichaelDaum: what is the minimal change needed to clear things for dakar....make pattern the fallback aka default [14:21] PeterThoeny: so one of two thinsg need to happen: (1) roll back classicskin into defaulttemplate, or (2) document the default template as a DefaultSkin [14:22] CDot: now it is clear; classic skin -> *.classic.tmpl [14:22] CDot: it is *not* a skin; it is the default condition (i.e. skinless) [14:22] CDot: a skin is something you put *over* the default templates [14:22] CDot: that is clear to *everyone*, I thought [14:23] PeterThoeny: crawford: exactly, and needs to be made accessible and documented a such [14:23] SvenDowideit: i didn't say that :) [14:23] PeterThoeny: make accessible via twikiskinbrowser [14:23] CDot: why make it accessible? [14:23] CDot: *it* *is* *not* *a* *skin* [14:23] PeterThoeny: so that people can discover it and see how it looks like! [14:23] SvenDowideit: mmm [14:23] ArthurClemens: So when is it confusing? When a user tries to edit a template in /templates and nothing happens. [14:23] CDot: so don;t confuse people by trying to make it into one [14:24] SvenDowideit: ArthurClemens, y [14:24] CDot: we don;t *want* people to discover it! We want them to use a *skin*, not the default templates [14:24] SvenDowideit: the fact that we have files that want to keep read only mixed with files that we talk about modifying is a bummer [14:24] CDot: the default templates are the fallback, o-mi-god consition [14:25] SvenDowideit: CDot, what you say is correct, but [14:25] ArthurClemens: So is there docs for editing templates that says to create a *.skinname.tmpl file [14:25] ArthurClemens: (i believe so) [14:25] PeterThoeny: if a user does a Set SKIN = foobar, (s)he will see the default template [14:25] SvenDowideit: we should also change the skin modification docs to say the same thing for pattern [14:25] SvenDowideit: ie if you want to edit a skin (like pattern) [14:25] CDot: PeterThoeny: right; because there *is* no foobar skin [14:25] SvenDowideit: create a new skin(by copying the files), and edit that [14:25] SamHasler: is the default unskined template eventually going to be less complex than Classic then? [14:26] SvenDowideit: SamHasler, y [14:26] CDot: a lot less complex [14:26] wbniv: y, but please let's keep the new skin discussions for Edinburgh [14:26] * CDot has it down to about 20% of the size [14:26] PeterThoeny: ok, what is the minimum we can do to clarify this? [14:26] SvenDowideit: no style, all substance.. [14:26] CDot: :-) [14:27] wbniv: i think MichaelDaum answered "make PatternSkin the default" [14:27] ArthurClemens: 1. documentation [14:27] CDot: I thought it *was* pretty clear in TWikiSkins [14:27] * CDot checks [14:27] ArthurClemens: I think so too [14:27] SvenDowideit: click [14:27] ArthurClemens: 2. an alert You are seeing the default templates [14:27] PeterThoeny: it is not discoverable, not accessible and not viewable [14:27] wbniv: SvenDowideit: btw, the quote is "a _foolish_ consistency is the hobgoblin of little mind" --- that extra word makes a big difference! [14:27] wbniv: s/mind/minds/ [14:28] SvenDowideit: dammit, i was avoiding calling you little :) [14:28] PeterThoeny: thanks sven [14:28] CDot: from TWikiSkins: "Skin files are located in the twiki/templates directory and are named with the syntax: ..tmpl" [14:28] PeterThoeny: can we agree on one of: (1) roll back classicskin into defaulttemplate, or (2) document the default template as a DefaultSkin? [14:28] SvenDowideit: generally, i'm the one that is accused of being retentive about consistency [14:29] CDot: PeterThoeny: NO! [14:29] wbniv: if you don't want to have two copies, use svn link to "copy" one to the other [14:29] SvenDowideit: 3) make default look different to classic [14:29] wbniv: but major skin work is for Edinburgh [14:29] PeterThoeny: ok, lets get a sense on where we stand, everyone: (1) or (2) or (3) something else? [14:29] MichaelDaum: we only need to aggree on some interim solution til the base tempaltes get refactored for Edin. [14:30] SvenDowideit: remove the colours and other non-essentials [14:30] CDot: SvenDowideit: I avoided (3) because of the risk for Dakar. i really didn't expect to be having this discussion. [14:30] SamHasler: 3 [14:30] SvenDowideit: i don't mean the full thing [14:30] SvenDowideit: just cosmetic [14:30] SvenDowideit: so its obvious to the user that its not the same thing [14:30] ArthurClemens: 3, an alert or message will do [14:30] CDot: sure; re-engineer the default templates to remove the action bars [14:30] wbniv: why do they care if they never get to the default skin? [14:30] CDot: wbniv: good question! [14:30] MichaelDaum: why's there a need to do _anything at all_ [14:31] SvenDowideit: only if they edit the wrong file.. [14:31] ArthurClemens: for ?skin=foobar [14:31] CDot: MichaelDaum: another good question! [14:31] wbniv: SvenDowideit: ah, so it's a developer issue only? [14:31] SvenDowideit: otherwise i agree do nothing [14:31] SvenDowideit: wbniv, i think so [14:32] PeterThoeny: so you do not mind that people cannot discover the default template and see how it looks like? [14:32] MichaelDaum: yes [14:32] CDot: no [14:32] SvenDowideit: correct [14:32] CDot: erm, clarification "no I do not mind" [14:32] PeterThoeny: the whole point of the skin browser is to discover all possible looks twiki has [14:32] MichaelDaum: if they want classic they have to install it [14:32] CDot: ArthurClemens: from TWikiSkins: "Although work is underway at TWiki:Codev.CssClassNames, the regular templates files currently do not use style sheets." [14:32] SvenDowideit: y, and default is not a skin, so its not discoverable [14:33] SamHasler: the "default" skin isn't a "look" though [14:33] CDot: Classic is installed by default [14:33] SvenDowideit: snap! [14:33] MichaelDaum: ah oh [14:33] ArthurClemens: "regular template files" - that are the default template files [14:33] CDot: Classic and Pattern skins are both installed by default. So is PlainSkin (one template file!) [14:34] ArthurClemens: but the doc should be updated [14:34] PeterThoeny: i see we do not reach a consensus on this [14:34] CDot: ArthurClemens: to most people, the "regular template files" means PatternSkin [14:34] MichaelDaum: go for minimal changes to get dakar out [14:34] SvenDowideit: y - do nothing :) [14:34] MichaelDaum: yep [14:34] CDot: do nothing [14:34] wbniv: MichaelDaum: yes, and frankly, i'm still not sure what the problem is... [14:34] MichaelDaum: me neither [14:35] SvenDowideit: sounds like consesnsus to me [14:35] wbniv: because default isn't findable? users don't care [14:35] PeterThoeny: let me ask differently then: do you mind if i add the DefaultTemplate so that it is discoverable in the skin browser? [14:35] wbniv: perhaps better to not say anything, as well, since they will radically change in Edinburgh ? [14:35] CDot: Yes, I do [14:35] SamHasler: default doesn't need to be findable [14:35] SvenDowideit: yes, because its wrong [14:35] wbniv: it should be added for Edinburgh, tho, right? [14:36] SvenDowideit: you know what, i'm starting to wonder about the whole thing. [14:36] SvenDowideit: maybe there should be no default. [14:36] PeterThoeny: so no consesnus, lets moveon to the next item [14:36] CDot: PeterThoeny: It sounds like we have consensus among everyone except you. Are there any other voices supporting Peter's position? SteffenPoulsen? Lavr_? ArthurClemens? [14:36] SvenDowideit: if you ask for a non-existant skin, it should tell you to bugger off [14:36] CDot: SvenDowideit: please don't cloud the issue [14:36] SvenDowideit: ok [14:36] Lavr_: No oppion. ONly uses pattern [14:36] wbniv: PeterThoeny: hold on peter, i think we have something close to a concensus [14:37] ArthurClemens: But when we do have the new default templates, it needs to be accessible, no? [14:37] CDot: ArthurClemens: no [14:37] ArthurClemens: or is it developer only? [14:37] CDot: we may choose to ship a skin closely based on the default templates [14:37] CDot: but when we do, it will be as a *skin* [14:37] SteffenPoulsen: we need to move on, if we wan't to reach the end of the item list - this is a can of worms, clearly. In my opinion skins have always been complicated to newcomers, and PatternSkin is an extreme help already. This is intellectual/developer to me. [14:37] SamHasler: default will become a mimimal styling that's enough to view a topic. Minimal enough that it doesn't require maintainance. It won't be something that users actually want to use. [14:37] MichaelDaum: the template infrastructure is a different issue [14:37] PeterThoeny: my personal preference is (1) roll back classicskin into defaulttemplate [14:37] wbniv: ah, ok, so default is a skin *library* [14:37] wbniv: ? [14:37] PeterThoeny: to avoiud confusion [14:38] ArthurClemens: ok: the default templates are base templates, to be used by skins [14:38] MichaelDaum: yes [14:38] CDot: PeterThoeny: you are a committee of one. You are ignoring the voices you are hearing. Why? [14:38] PeterThoeny: but i can settle for (2) document the default template as a DefaultSkin [14:38] SvenDowideit: can we move to seperate out the template sets into directories? [14:38] wbniv: therefore, not for users, therefore not in TWikiSkinBrowser, etc. [14:38] wbniv: let's move on, nothing to see here... [14:39] SvenDowideit: y [14:39] ArthurClemens: I object to have it labeled as a skin then [14:39] wbniv: ok, fair point [14:39] ArthurClemens: DefaultSkin will make it more confusing [14:39] wbniv: call it a SkinContrib [14:39] wbniv: SkinLibrary [14:39] wbniv: SkinBaseTemplates [14:39] wbniv: or something else [14:39] ArthurClemens: DefaultSkin is a skin, so PatternSkin or ClassicSkin [14:39] MichaelDaum: base templates [14:39] SamHasler: CDot, am I right in thinking that the main drive for this is so that default is minimal enough that it doesn't require maintainance? [14:40] CDot: don;t call it anything; everyone seems to be agreed to do nothing, except Peter [14:40] CDot: SamHasler: nothing is "maintenance free", but yes, that's kinda the goal [14:40] wbniv: SamHasler: i think it's about refactoring common functionality so that creating skins is easier, too [14:40] CDot: I wanted to make a sllear separation between "Skinned" and "Unskinned" [14:40] SamHasler: that's what I meant, I just wanted to hammer home the point [14:40] CDot: that way people can understand the concept of a Skin better [14:41] CDot: cos right now it's really confusing [14:41] SvenDowideit: y, and causes ppl not to re-use common TMLP:DEFs [14:41] CDot: right [14:41] PeterThoeny: ok, lets move on [14:41] SvenDowideit: as consensus == broad unanimity [14:41] wbniv: it's helping break down skins into the multiple parts of functionality that they actuall are [14:41] SvenDowideit: i'd say we have it. [14:41] PeterThoeny: no consensus for me [14:41] wbniv: hold on [14:42] SvenDowideit: consensus is not everyone [14:42] wbniv: we *do* have concensus [14:42] wbniv: you just don't agree [14:42] SvenDowideit: most agree, so we know what we will do [14:42] wbniv: but i'm willing to move to the next item [14:42] PeterThoeny: ok, lets move on [14:42] PeterThoeny: ---++ Item1328 RSS feed validation error [14:42] SvenDowideit: me [14:42] wbniv: that's on sven's list, next [14:42] SvenDowideit: will be done within the week [14:42] PeterThoeny: ok [14:42] wbniv: (probably this weekend, he said) [14:42] CDot: i.e. today [14:43] PeterThoeny: ---++ Item1322 TWiki.org doc merge: TWikiHistory needs Dakar history [14:43] PeterThoeny: we talke before [14:43] PeterThoeny: action item for me [14:43] SvenDowideit: addtoagenda: new beta this weekend? [14:43] wbniv: SvenDowideit: wait until the end [14:43] wbniv: we're not there yet [14:43] PeterThoeny: souns good to me! [14:43] * CDot doesn't have time to build one, so someone else will have to do it [14:43] SvenDowideit: net [14:43] SvenDowideit: next [14:43] wbniv: monday would be better for me, but let's have that discussion at the end [14:44] PeterThoeny: ---++ Item1345 $username in LINKTOOLTIPINFO expands to TWiki::User=HASH style [14:44] PeterThoeny: what is that? [14:44] CDot: unverified (I haven't investigated yet) [14:44] SteffenPoulsen: testcase in sandbox, move your mouse :-) [14:44] PeterThoeny: ---++ Item663 TWiki.org Merge [14:44] SvenDowideit: it'll be a lack of de-Ref? [14:44] PeterThoeny: we talked before [14:45] PeterThoeny: ---++ Item1338 %SEARCH across multiple lines causes infinite loop [14:45] CDot: SteffenPoulsen: on 1338, does it infinite loop with tags other than %SEARCH? did you try? [14:45] CDot: also not investigated [14:45] SteffenPoulsen: oh, nope, tags like? [14:45] CDot: %INCLUDE{" [14:45] CDot: "}% [14:45] PeterThoeny: i had an uneasy feeling when the syntax was relaxed to cover multiple lines [14:45] SteffenPoulsen: good point, I'll poke around a bit [14:45] CDot: it shouldn't matter, but we need to check [14:45] SvenDowideit: funny :) [14:46] CDot: the line-by-line parsing is still there in places, and it's a right PITA [14:46] PeterThoeny: multiple lines means that any s///go needs to be now s///gos [14:47] CDot: the " on it's own may be causing an issue. [14:47] CDot: PeterThoeny: correct [14:47] PeterThoeny: and that needs to be checked for every substitution [14:47] CDot: it has been, as far as I am able. That's what the dozens of testcases are for, among other things. [14:47] PeterThoeny: so, any parameter passed to any variable potentially can have newlines that potentially break something [14:48] PeterThoeny: as it did in cairo with one security alert [14:48] PeterThoeny: very prone to bugs if not tested thoroughly [14:48] CDot: If you find a case, then please report it and we can fix it. [14:48] CDot: Otheriwse you are speculating. [14:49] SvenDowideit: next [14:49] PeterThoeny: one safe thing that web revert the variable over multiple lines for dakar and do it right for edin* [14:49] SvenDowideit: no, there are too many things that rely on it, and its more readable .. [14:50] PeterThoeny: oops once more, "one safe thing is to revert the "variable over multiple lines" feature in dakar and do it right for edin* [14:50] CDot: Absolutely not. That would break far more than it fixes. [14:50] PeterThoeny: agreed, it is much more readbale [14:50] CDot: You are speculating on the cause of 1338, ithout any facts to substantiate your accusations. [14:50] SvenDowideit: it would be better to revert to cairo code than to revert it (its pretty fundamental) [14:51] CDot: Please wait for an analysis before slinging mud. [14:51] wbniv: can we please move to the next item as SteffenPoulsen is investigating more [14:51] SvenDowideit: next [14:51] PeterThoeny: slinging mud? [14:51] SvenDowideit: next [14:51] PeterThoeny: lets move on [14:52] PeterThoeny: i see that it on urgent and required for engine :-) [14:52] PeterThoeny: next agenda item [14:52] PeterThoeny: ---++ Other items that need to happen for release [14:52] PeterThoeny: anything? [14:52] ArthurClemens: A new twiki.org home [14:53] SvenDowideit: twiki.org home? [14:53] PeterThoeny: arthur, can you elaborate [14:53] ArthurClemens: the page [14:53] SvenDowideit: or a new develop.twiki.org home [14:53] SvenDowideit: ah [14:53] ArthurClemens: you see, the marketing machine for Dakar [14:53] CDot: one item I would like to open again, sorry about this, is the release naming [14:53] SvenDowideit: y [14:54] CDot: I am finding that the dates are doing my head in [14:54] PeterThoeny: oh gosh [14:54] SamHasler: me too [14:54] CDot: I would strongly prefer to move to a more friendly scheme [14:54] SvenDowideit: i have never dealt successfully with the dates either [14:54] SamHasler: (about dates doing my head in) [14:54] CDot: I thought I could handle it, but I'm finding it very difficult [14:54] ArthurClemens: I still feel for TWiki4.0 [14:54] SamHasler: I even found the names of the new TWiki webs meaningless [14:55] wbniv: it also makes the refresher packing naming _much_ more simplified !! [14:55] SvenDowideit: real release number would help for debain and other packaging systems - simply adopting the normal x.x.x.x standard would be nice [14:55] SamHasler: especially since there can be re-release that make the web names suspect [14:55] SvenDowideit: M.m.p.b [14:56] SvenDowideit: Major.minor.patch.svnnumber [14:56] MichaelDaum: three are enuf [14:56] SvenDowideit: where the last is ever increasing, and only the first 3 for public [14:56] CDot: whatever; there are two questions that have banjaxed me in the last few days: [14:56] SvenDowideit: the last is very very handy [14:56] MichaelDaum: hm ok [14:57] MichaelDaum: as long as we get that scheme at all ... [14:57] Lavr_: I do not fancy the date release names much I must admit. We see many people on #twiki that are confused about the mix of code names and mysterious date names [14:57] CDot: 1) What is the release packages for the last major release? Every time I am asked, I have to go and hunt, which is very inefficient [14:57] SvenDowideit: again, there is consesnsus, just peter disagreeing [14:57] SvenDowideit: ? [14:57] CDot: "Cairo" or "3.0" is much easier to remeber [14:58] CDot: and, second, "Was there a release between XXX and XXX" [14:58] MichaelDaum: most confusing was 4 Sep [14:58] SvenDowideit: y, and when there's a security fix, we bump the patch level [14:58] CDot: if the numbers are sequentioal. it's easy [14:58] CDot: but if you use dates, it takes ages to find out [14:59] Lavr_: Calling Dakar 4.0 also makes it clear that 4.1, 4.2, 4.3 is an enhancement release. And 4.0.1 is a small bug fix release. [14:59] CDot: and you can never be sure that some other bugger didn;t do a build in between [14:59] wbniv: especially when release dates are "lied" about [14:59] MichaelDaum: most people think that it is the time of the release which is not the case [14:59] SamHasler: and if we continue using codenames it's easy to convert then to release numbers in most alphabets [14:59] wbniv: PeterThoeny's gone quiet [14:59] SvenDowideit: OK, can we turn this around? [14:59] SvenDowideit: who has points agains M.m.p.b [14:59] PeterThoeny: just reading what you have to say... [14:59] MichaelDaum: we should document when to increase major, minor and patch [15:00] SvenDowideit: MichaelDaum, mostly automated in the buld sys [15:00] MichaelDaum: how does that work then [15:01] SamHasler: what's the difference between minor and patch? [15:01] SvenDowideit: create new release branch ups the Major [15:01] Lavr_: If I download TWiki the 14 of January and in March there is a security update called TWiki02Jan2006 (which was in reality released sometime in February) a normal user will have no clue that it is important to update to the 02Jan. [15:01] wbniv: y, it would be added to the build system if/when we use those numbers; i don't think there's really support in there for it now .... [15:01] SvenDowideit: please STOP [15:01] PeterThoeny: fwiw, (Link: http://twiki.org/cgi-bin/view/Codev/TWikiReleaseNamingConvention)http://twiki.org/cgi-bin/view/Codev/TWikiReleaseNamingConvention [15:01] ArthurClemens: We can still use Cairo, Dakar and Edinburough for consistensy: TWiki4.0 (Dakar) [15:01] SvenDowideit: the details are simple, and not important now [15:01] * CDot can't see what's wrong with TWikiDakarBeta5..TWikiDakar...TWikiDakarPatch1...TWikiDakarPatch2 etc [15:02] wbniv: well, can we agree to all comment on that topic and revisit this discussion next week ? [15:02] CDot: maybe inconsistency with Debian etc? [15:02] SvenDowideit: the only thing remaining is arguments against the consensus proposal [15:02] wbniv: (and, because i would like to have weekly release meetings until the release) [15:02] SvenDowideit: mmm, can't we release next week [15:02] SvenDowideit: ? [15:03] wbniv: depends on the status of the bugs... [15:03] PeterThoeny: ok, here is a deal: I am willing to review the TWikiReleaseNamingConvention without a date if it fullfills the needs i outlined in (Link: http://twiki.org/cgi-bin/view/Codev/TWikiReleaseNamingConvention)http://twiki.org/cgi-bin/view/Codev/TWikiReleaseNamingConvention [15:03] MichaelDaum: did we discuss the twiki.org switch already ... sorry to ask [15:03] wbniv: PeterThoeny: more than fair. we should comment and talk about it next week ? [15:03] SvenDowideit: so simply that its what the users want is not enough? [15:03] MichaelDaum: I'd expect more bugs as a result of this [15:04] PeterThoeny: we spent now 3 hours, i think we should move on [15:04] SvenDowideit: y - ok [15:04] wbniv: agreed [15:04] SvenDowideit: next meeting it is [15:04] CDot: PeterThoeny: I'm OK with that. [15:04] PeterThoeny: and discuss proposals in Codev.TWikiReleaseNamingConvention [15:04] PeterThoeny: good [15:05] PeterThoeny: i have one "other" item [15:05] PeterThoeny: upgrade of twiki.org [15:05] wbniv: floor's open for new items now, right? i have (up to) 3 additional points [15:05] PeterThoeny: that should happen at the release or shortly after [15:05] ArthurClemens: I notice that (Link: http://twiki.org/cgi-bin/view/Codev/TWikiReleases)http://twiki.org/cgi-bin/view/Codev/TWikiReleases first lists the alphas, then betas then stable. Should this not be the way around? [15:05] SvenDowideit: mm, can't we go before, to get more testing? [15:05] MichaelDaum: no, it shoud happen NOW [15:05] PeterThoeny: i think we can take this offline and discuss in #twiki [15:06] PeterThoeny: arthue: i do not have a strong opinion on that, change it if you like [15:06] PeterThoeny: s/arthue/arthur/ [15:06] ArthurClemens: k [15:07] PeterThoeny: will? [15:07] wbniv: 1. can we have weekly meetings until the release? [15:07] SvenDowideit: but on a different day? [15:07] MichaelDaum: yea, weekends are for my family in general [15:08] wbniv: well, first let's agree that we want weekly meetings and THEN we can schedule them [15:08] SvenDowideit: y to weekly, but prefferably during week [15:08] PeterThoeny: sounds good to mee, both [15:08] MichaelDaum: mondays [15:08] SvenDowideit: or are we excluding people who work for a living? [15:08] Lavr_: Docs are ready in one week. Bug rate is falling and moving towards plugins. Sounds to me like release is possible no more than two weeks from now. Provided we still commit a bugfix release a month or so later. [15:09] wbniv: any day is good for me, really; you guys decide [15:09] CDot: I'm easy, though bear in mind that we are already 7 days past the most pessimistic release date from our last meeting. [15:09] MichaelDaum: MMM --- monday morning meetings [15:09] SvenDowideit: y [15:09] PeterThoeny: hehe [15:09] PeterThoeny: with mmm - monday morning meeting minutes [15:09] SvenDowideit: next. [15:09] PeterThoeny: mmmm [15:10] CDot: Lavr_: you are our conscience [15:10] wbniv: 2. what should i be working on? i have very limited connection to the internet and i'm not sure i can do much more than fixing plugins and random doc fixes [15:10] wbniv: that i've been working on lately [15:10] CDot: wbniv: sounds good to me [15:10] SvenDowideit: testing & docco? [15:10] SvenDowideit: grin [15:10] MichaelDaum: m3c --- monday morning meeting consensus [15:10] wbniv: ok, last item [15:10] PeterThoeny: so next one monday in a week? [15:10] CDot: note; we still haven't agreed how to release plugins (one for next meeting) [15:11] wbniv: cdot's gonna slap me ;-) [15:11] * CDot readies a wet kipper [15:11] SvenDowideit: hehehe [15:11] wbniv: i want to split the MANIFEST so that the TWikiKernel has its own MANIFEST file [15:11] CDot: SLAP! [15:11] wbniv: and the distro (ie, a specific TWikiFor !includes TWikiKernel/MANIFEST, or whatever) [15:11] SvenDowideit: move this one to codev [15:11] SvenDowideit: next. [15:11] wbniv: arg, fine [15:11] wbniv: i've already done it [15:11] SvenDowideit: :) [15:11] wbniv: i just wanted to see if it's ok [15:12] wbniv: ok, back to plugin releases then [15:12] PeterThoeny: oh, we forgot one item: review action items, see "old" at end of (Link: http://twiki.org/cgi-bin/view/Codev/DakarReleaseMeeting2006x01x07)http://twiki.org/cgi-bin/view/Codev/DakarReleaseMeeting2006x01x07 [15:12] CDot: no, agendum for next week [15:12] wbniv: CDot: ah, ok, nm [15:12] wbniv: PeterThoeny: ok, continue [15:12] PeterThoeny: we talked abot peter's [15:12] PeterThoeny: arthur, on classic skin? [15:13] PeterThoeny: "Update classic skin to include a login and logout button" [15:13] wbniv: PeterThoeny: i thought you said those items were all resolved [15:13] wbniv: (with i18n deferred) [15:13] PeterThoeny: is arthur gone? [15:13] ArthurClemens: No, I was doing header art [15:14] wbniv: he just checked in a new TWikiSkins [15:14] PeterThoeny: oh, great, multitaksin :-) [15:14] MichaelDaum: s/ksin/skin/ [15:14] ArthurClemens: Ah, login and logout button. I forgot about these [15:14] SamHasler: *groan* [15:14] PeterThoeny: on action items for "all" [15:15] PeterThoeny: any one on "Help find tech writer"? [15:15] PeterThoeny: i did not action on this one [15:15] PeterThoeny: quiet, so no action [15:15] SvenDowideit: i'm afraid that to attract someone, we need to fix up twiki.org first [15:16] PeterThoeny: ok, move on [15:16] PeterThoeny: ---++ agenda item: Discuss tentative release date [15:16] wbniv: Docs are ready in one week. Bug rate is falling and moving towards plugins. Sounds to me like release is possible no more than two weeks from now. Provided we still commit a bugfix release a month or so later. [15:17] CDot: FYI, the %SEARCH problem is nothing to do with the parser. It works fine with newlines in values. The problem is somewhere else. [15:17] SteffenPoulsen: still too early to set release date imho, will be easier next meeting - personally I'm just thrilled there's visible progress now, as opposed to last meeting where everything seemed to have more or less halted (for no apparent reason .. perhaps except holiday season :-)) [15:18] wbniv: y, i think it was mostly holidays [15:18] PeterThoeny: i have once concern related to release date [15:18] * PeterThoeny thnks on the list of labelled again as mud slider [15:18] PeterThoeny: s/list/risk/ [15:19] PeterThoeny: the rate of bugs arriving still concern me [15:19] * MichaelDaum needs to go to bed ... good night all [15:19] SvenDowideit: night :) [15:19] PeterThoeny: twiki has a reputation of being rock solid [15:19] SvenDowideit: i agree, the bug rate is too low to release yet [15:19] SvenDowideit: funny! [15:19] SamHasler: does it? [15:19] PeterThoeny: night michael, thanks for hanging out [15:19] SvenDowideit: i thought it had a rep of being a steaming pile of security flaw [15:19] SteffenPoulsen: night md [15:19] *** MichaelDaum is now known as MichaelDaum_. [15:20] SvenDowideit: (i suspect my opinion is very low compared to most - i'm a depressive) [15:20] PeterThoeny: sven: cairo had only two major issues for over one year, plus (i think) three security issues (that were not that dnagerous for behind firewall use) [15:20] CDot: erm, according to standard SW engineering practice, the bug rate is irrelevant. It is the bug *turnaround* rate that is important. [15:20] SvenDowideit: the number of ppl that reported a compromise to me was huge [15:20] wbniv: PeterThoeny: i suspect that some of that has to do with how most installations don't push twiki very hard --- we've certainly fixed a lot of bugs in the last year [15:21] wbniv: but, please continue with your statement [15:21] SvenDowideit: and most lost servers [15:21] PeterThoeny: so, my concern is if we release now and get 20 bug reports within a month it will ding the reputation of twiki [15:21] CDot: only if they stay open for a year, like most of the Cairo bug reports..... [15:21] wbniv: but that's part of why we have to have refresher packges [15:21] CDot: right; release often [15:21] SvenDowideit: compared to the ding we get from not releaseing often enough [15:22] PeterThoeny: i did some statistics [15:22] SvenDowideit: 100 bugs in the next month would be good. [15:22] wbniv: and, btw, i suspect more than 20 bugs reports in the first month [15:22] SvenDowideit: if we have a fix release for them [15:22] * wbniv listening [15:22] SamHasler: yeah, I'd be surprised if it wasn't more than 20 [15:22] PeterThoeny: new bug items created by month: Date Total 2005/08 5 2005/09 10 2005/10 50 2005/11 243 2005/12 201 2006/01 61 [15:23] wbniv: is this from the Bugs web? [15:23] SvenDowideit: peter, any chance of going back further than that? [15:23] CDot: it has to be from the Support web [15:23] SvenDowideit: sept oct were major depression of contributions [15:23] SvenDowideit: die to conference [15:23] SvenDowideit: and that we promised a release and did not deliver [15:23] CDot: but those are Support questions, surely, not bug reports? [15:23] PeterThoeny: little trivia from early days of twiki, (Link: http://twiki.org/cgi-bin/view/Codev/ReadmeFirst)http://twiki.org/cgi-bin/view/Codev/ReadmeFirst 's mission has " Keep the quality as high as it is, e.g. few or no bugs for production releases and deployable in many environments." [15:24] wbniv: well, no bugs is a fantasy [15:24] wbniv: and, of course, we want few bugs [15:24] Lavr_: The raise in November was when the first beta was released. That is what got me activated. I started testing Dakar at beta 3 [15:24] SvenDowideit: yes, but I for one do not agree that the previous releases had fewer bugs [15:24] wbniv: SvenDowideit: actually, i don't think we know that either way [15:24] PeterThoeny: the stats are from (Link: http://develop.twiki.org/~develop/cgi-bin/search/Bugs/WebSearchAdvanced?search=META.*ExtensionName.*value.)http://develop.twiki.org/~develop/cgi-bin/search/Bugs/WebSearchAdvanced?search=META.*ExtensionName.*value.\%22\%22&scope=text&order=created®ex=on&limit=all&format=|[[$topic]]|$createdate|$rev%20-%20$date| [15:24] SvenDowideit: just that we were able to ignore them better [15:24] PeterThoeny: with manual counting of month [15:24] SvenDowideit: PeterThoeny, thanks :) [15:24] CDot: but those stats are irrelevant to the beta, surely? [15:24] CDot: oh, ok, sorry [15:25] CDot: ok, that makes sense, bugs against Dakar [15:25] CDot: I thought you were ana;ysing Cairo :-( [15:25] wbniv: one flaw: "bug" == "action item", so not all things listed are BUGS [15:25] PeterThoeny: and for what it is worth, here are the stats of bug items i filed: Date Peter 2005/10 54 2005/11 29 2005/12 27 2006/01 19 [15:25] SvenDowideit: y, they are not related to anything other than changing the process [15:25] SteffenPoulsen: I say let's just enjoy the current progress / turnaround rate and do our best fixing for a yet a while .. next meeting things will be clearer [15:26] Lavr_: When looking at bug reports it is important to look at what the bugs are raised against. There are many plugin and doc related bugs now. [15:26] PeterThoeny: my point is, the rate of new bugs is not declining at the moment, which is what i am concerned [15:26] wbniv: i think peter accounted for that in his search [15:26] SteffenPoulsen: I have tested with most of our "delicate" content from our site now, and my personal bug-rate is dramatically falling already - even with WYSIWYG ;-) [15:26] SvenDowideit: I'm happy to listen to Lavr_ 's gut feel - he's reliable for that [15:26] wbniv: not declining against the kernel or against everything ? [15:26] PeterThoeny: we can release if we commit to release a fix after one month as kenneth suggsted [15:27] wbniv: because i've created lots of new bug reports to log against the plugins i've been fixing [15:27] SvenDowideit: I think we should start getting used to monthly releases [15:27] wbniv: maybe every other month [15:27] wbniv: but whatever [15:27] wbniv: yes, more frequently [15:27] wbniv: more like, whenever there's a need [15:27] wbniv: eg, a bunch of bug fixes [15:27] CDot: ok, the hang in %SEARCH is due to the attribute parsing code inherited from Cairo, which isn't robust to newlines in parameter values. [15:27] wbniv: or some refinement that wants to get published [15:28] PeterThoeny: more frequent releases are good, but do not forget that many admins do not want to be bothered with frequent bugfix updates [15:28] PeterThoeny: so many _stable_ releases, yes [15:28] SvenDowideit: y, most admins don't care about most bugs [15:28] SteffenPoulsen: CDot: Sounds like it's more or less fixed already :-) [15:28] SvenDowideit: only the ones that affect their users [15:29] CDot: SteffenPoulsen: I'm on the case ;-) [15:29] Lavr_: The bug fix releases must be made so you can "patch" a live server. A full reinstall is a pain. But is we make a download option where you only download the "diff" since last release and with no preference topics then it will be much easier to upgrade. [15:29] PeterThoeny: ok, looks like we should defer the decision to monday in a week [15:30] PeterThoeny: (decision on release date) [15:30] SvenDowideit: Lavr_, y, that too [15:30] PeterThoeny: 3.5 hours on meeting now [15:30] wbniv: Lavr_: and now there's a TWikiExtensionInstallerContrib [15:30] SvenDowideit: 1.5 to go [15:30] wbniv: heh [15:30] wbniv: no, i think we're done, right ? [15:30] PeterThoeny: i think so too :-) [15:31] Lavr_: We should end now. It has been a very good meeting. [15:31] wbniv: yup [15:31] SvenDowideit: ok [15:31] PeterThoeny: i willspend some time with my kids now, i neglected them too much in the last few days :-( [15:31] wbniv: hang up, no, you hang up first..., no you... :) [15:31] SamHasler: before everyone leaves, a bit of personal news. I've finally got a job [15:31] SvenDowideit: you what? [15:31] ArthurClemens: great! [15:31] SvenDowideit: thats no fair [15:31] wbniv: congrats [15:32] SvenDowideit: i don't [15:32] PeterThoeny: hey sam, congrats! [15:32] SamHasler: It's working from home but it is full time. [15:32] wbniv: sounds great [15:32] SvenDowideit: well done indeed ! [15:32] PeterThoeny: ____ __.---""---.__ ____ /####\/ \/####\ ( /\ ) ( /\ ) \____/ \____/ __/ \__ .-" . . "-. | | \.._ _../ | | \ \ \."-.__________.-"./ / / \ \ "--.________.--" / / ___\ \_ _/ /___ ./ ))))) ((((( \. \ / \ \_ _/ / \ \____/""-.____.-""\____/ / \ \ / / \. .| |. ./ ." / | \ / | \ ". ." / | \ / | \ ". /.-./.--.|.--.\ /.--.|.--.\.-.| [15:32] Lavr_: Congrats Sam [15:32] PeterThoeny: it was a good meeting [15:32] SteffenPoulsen: SamHasler: GRATZ! :-) PeterThoeny: Great idea, I'm sure they won't turn you down :-) .. oh no, Peter is about to kill himself :-) [15:33] PeterThoeny: thanks everyone for hanging in there for such a long time [15:33] PeterThoeny: especially in european and australian time zones [15:33] CDot: SamHasler: what doing? [15:33] Lavr_: Good night. [15:33] ArthurClemens: ok, bye everyone! [15:33] wbniv: l8r all [15:33] *** wbniv has left #twiki_dakar. [15:33] SvenDowideit: caio [15:33] *** ArthurClemens has signed off IRC ("Leaving"). [15:33] PeterThoeny: ttyl [15:34] *** CDot has left #twiki_dakar. [15:34] *** Lavr_ has left #twiki_dakar. [15:34] SamHasler: CDot, developing this (Link: http://www.probeops.co.uk/)http://www.probeops.co.uk/ [15:34] SamHasler: drat, he's gone [15:34] SteffenPoulsen: - looks like you learned the lines/minute spam-lesson after all, Peter :-) - thanks for now, cya around [15:36] *** SteffenPoulsen has left #twiki_dakar. [15:41] *** MartinCleaver has signed off IRC (Excess Flood).