<div>Hey all,</div><div> </div><div>I would like to move selectable intro trunk asap, best this weekend. This would effectively mean it would become</div><div>part of the 1.8 release, so please vote for its inclusion.</div> <div> </div><div>I'm positive it would fix almost all issues here: <a href="http://dev.jqueryui.com/report/10?P=selectable">http://dev.jqueryui.com/report/10?P=selectable</a> , so please don't start</div><div>working on any of these tickets before we come to a decision on this.</div> <div> </div><div>The remaining issues are tiny:</div><div> </div><div>- add the dotted outline to ui-state-focus in TR (spoke to Maggie about this yesterday, FG agreed)</div><div>- implement single select/deselect methods (will do this tonight)</div> <div>- finish the spec (I request help with that - especially need a changelog / compatibility information)</div><div>- update existing selectable tests</div><div> </div><div>all of those, except for deselect/select methods, can easily be done after the move to trunk. It's a special case,</div> <div>because full test and spec coverage didn't exist for selectables anyway, so it can do no harm here.</div><div> </div><div>Any objections?</div><div><div> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div></div>
Hi all,<div> </div><div>see this ticket, Jörn already replied to it: <a href="http://dev.jqueryui.com/ticket/3782">http://dev.jqueryui.com/ticket/3782</a></div><div> </div><div>I remember it was in there once, but we removed it. Was there a good reason to remove it?</div> <div>I also think it's good to have it in there.<br clear="all"> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div>
Hi all,<div> </div><div>in the current Selectables, we have a two certain special states for items called 'selecting and 'deselecting'.</div><div>A couple notes about them:</div><div> </div><div>- 'selecting' happens when you are in the process of drawing a lasso and didn't release the mouse yet</div> <div>- 'deselecting' (or unselecting, don't remember) happens when you are in the process of deselecting an item via lasso and didn't release the mouse yet</div><div>- selecting comes in as ui.selecting in the callbacks and the items receive a 'ui-selecting' class name</div> <div>- once the mouse button is released, elements are being selected properly (with the class 'ui-selected' I think)</div><div> </div><div>I always found this feature weird, since it never had a real use (imho) other than styling the elements differently - another use I could see is if you</div> <div>have to make calls or things like that when you select, and it's costly to make calls one by one, so it's done after the lasso draw.</div><div> </div><div>However, other than that, I'm seeing mostly added complexity and other disadvantages - for instance, it is only applicable for the lasso - not to normal mouseselection, not to keyboard selection (maybe while holding shift, thought that would feel weird as well).</div> <div> </div><div>In the new selectables, this whole feature is gone. This has the side effect that it's not completely backwards compatible (along with some other much needed API changes), but I think that almost all previous features can quickly be rewritten since I added one additional boolean in callbacks: ui.lasso (true/false). </div> <div> </div><div>Are there any good reasons not to remove it, or rather, get it in again?</div><div> </div><div>Cheers,</div><div>Paul<br clear="all"> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div>
Hi guys,<div> </div><div>working with getScript can be major pain because there's no way to actually find out if the include failed or not. There are two tickets</div><div>related to this in the bugtracker: <a href="http://dev.jquery.com/ticket/1768">http://dev.jquery.com/ticket/1768</a> and <a href="http://dev.jquery.com/ticket/321">http://dev.jquery.com/ticket/321</a> , but more specifically,</div> <div>simply adding a onerror (doh!) callback to the script tag fixed the issue for me.</div><div> </div><div>So even regardless of the best way to solve it, I really think we need a featured solution for informing users that loading a script</div> <div>has failed.</div><div> </div><div>Any objections?</div><div> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div>
Hi everyone,<div> </div><div>as promised, I've started working on the wiki again. Today, I've worked hard to fill the Selectable page at <a href="http://wiki.jqueryui.com/Selectable">http://wiki.jqueryui.com/Selectable</a> with</div> <div>information about the refactor.</div><div> </div><div>It's important to note that the actual code is very far ahead and stable across browsers, so I'm highly suggesting to include it with 1.8, if</div><div> we can get the specs finished in time. Please have a look at my writeup so far, and also checkout the demos at the bottom, and the actual</div><div>implementation at branches/labs/selectable-labs/ui.selectable.js.</div><div> </div><div>I'm looking forward to lots of feedback, and if you have any question about the implementation, please ping me anytime.</div><div> </div><div>Thanks,</div><div>Paul<br clear="all"> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div>
Hey guys,<div> </div><div>any idea what's up with branches/labs/selectable? My new refactored selectable seems to be in selectable-labs/, so I have</div><div>no idea what the other folder is. Can I delete it?</div><div> </div><div>Also, I would like to propose to move the selectable refactor to dev/. I will try to update the spec page along, but the</div><div>refactor is pretty straight forward with some nice changes - all options of the old selectables are supported at this point.</div> <div> </div><div>Cheers,</div><div>Paul <div> -- Paul Bakaus UI Architect @ <a href="http://smart.fm">smart.fm</a> -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a> </div></div>
Hey everyone, I today found out that you can use the attribute filter to filter for keys in generic objects, which is *way* cool. Check this out: $([ { foo: 1, bar: 0 }, { foo: 0, bar: 0 } ]).not("[foo=0]") => returns [{ foo: 1, bar: 0 }] This is insanely useful if working with JSON or in general big data sets, and the only piece missing right now is better comparison methods. So I'd like to see "[foo>5]" working, for instance. What do you think?<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, might be worth investigating: <a href="http://ajaxian.com/archives/souders-dont-use-import">http://ajaxian.com/archives/souders-dont-use-import</a> I don't know about your experience, but loading up themeroller or the index.html in the generated zip always froze my browser (FF3!) for a couple seconds. Maybe that's the key? Cheers, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hi guys, as of now, it's too difficult to find the possibility to download jQuery UI in seperate files. We seriously need links to the developer package on the download builder page (and gettin started guides). Can anyone tackle this asap? Thanks, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, Trac will be not accessible for about 10 minutes for maintenance reasons. Thanks for your patience, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, we're currently collecting everything that has changed between jQuery UI 1.5.3 and jQuery UI 1.7. We will later use this information to write a comprehensive upgrade guide for all users that will be featured in the jQuery wiki and in our website. You can find the current collection here: <a href="http://wiki.jqueryui.com/Upgrade-Guide">http://wiki.jqueryui.com/Upgrade-Guide</a> I've pretty much updated everything I could with the exception of tabs. Please feel free to add anything you think would be valuable to note, and of course, information on how the tabs plugin changed. Thanks, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey team, I invested plenty of hours to manage tickets in Trac and update the changelog at <a href="http://docs.jquery.com/UI/Changelog/1.7">http://docs.jquery.com/UI/Changelog/1.7</a> . My time is over for today, but if everyone wants to do more tonight, here's what's left to do, and how to do it: The following plugins still need to be updated: - Resizable - Slider - Sortable - Tabs I have prepared a custom trac report that gives you a list of only these, and only the items you'll have to enter. Open the following epic link: <a href="http://dev.jqueryui.com/query?status=closed&group=component&component=!CSS+Framework&component=!effects.*+(individual+effect)&component=!effects.core&component=!ui.accordion&component=!ui.autocomplete&component=!ui.colorpicker&component=!ui.core&component=!ui.datepicker&component=!ui.dialog&component=!ui.draggable&component=!ui.droppable&component=!ui.grid&component=!ui.history&component=!ui.magnifier&component=!ui.menu&component=!ui.progressbar&component=!ui.spinner&component=!ui.tree&component=![meta]+ui.build&component=![meta]+ui.demos&component=![meta]+ui.design&component=![meta]+ui.docs&component=![meta]+ui.downloadbuilder&component=![meta]+ui.site&component=![meta]+ui.test&component=![meta]+ui.themeroller&component=!ui.selectable&order=changetime&col=id&col=summary&col=milestone&col=type&col=changetime&milestone=!1.5&milestone=!1.5.1&milestone=!1.5.2&milestone=!1.6b&resolution=fixed">http://dev.jqueryui.com/query?status=closed&group=component&component=!CSS+Framework&component=!effects.*+(individual+effect)&component=!effects.core&component=!ui.accordion&component=!ui.autocomplete&component=!ui.colorpicker&component=!ui.core&component=!ui.datepicker&component=!ui.dialog&component=!ui.draggable&component=!ui.droppable&component=!ui.grid&component=!ui.history&component=!ui.magnifier&component=!ui.menu&component=!ui.progressbar&component=!ui.spinner&component=!ui.tree&component=![meta]+ui.build&component=![meta]+ui.demos&component=![meta]+ui.design&component=![meta]+ui.docs&component=![meta]+ui.downloadbuilder&component=![meta]+ui.site&component=![meta]+ui.test&component=![meta]+ui.themeroller&component=!ui.selectable&order=changetime&col=id&col=summary&col=milestone&col=type&col=changetime&milestone=!1.5&milestone=!1.5.1&milestone=!1.5.2&milestone=!1.6b&resolution=fixed</a><br clear="all"> This will give you some 150 tickets that need to be check through and filled into the changelog. Thanks for any help you can provide tonight (answer to this thread if you do so please), otherwise I'll resume tomorrow. Thanks, Paul -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, I've digged through the current open blockers/criticals, here's what's left to do for some: <a href="http://dev.jqueryui.com/ticket/3952">http://dev.jqueryui.com/ticket/3952</a> - needs a attached test case (can someone convert the jsbin stuff and attach it?) <a href="http://dev.jqueryui.com/ticket/4188">http://dev.jqueryui.com/ticket/4188</a> - this one needs serious attention. I'm clueless, and I appreciate any input. <a href="http://dev.jqueryui.com/ticket/3994">http://dev.jqueryui.com/ticket/3994</a> - needs stripped down test cases (not using draggable) to progress <a href="http://dev.jqueryui.com/ticket/4166">http://dev.jqueryui.com/ticket/4166</a> - needs test case <a href="http://dev.jqueryui.com/ticket/4063">http://dev.jqueryui.com/ticket/4063</a> - needs em2px utility in core <a href="http://dev.jqueryui.com/ticket/4148">http://dev.jqueryui.com/ticket/4148</a> - has an accordion link but the title is slider - what's incorrect here? <a href="http://dev.jqueryui.com/ticket/4010">http://dev.jqueryui.com/ticket/4010</a> & <a href="http://dev.jqueryui.com/ticket/4180">http://dev.jqueryui.com/ticket/4180</a> - why are those priorized as 'critical' and not 'blocker'? <a href="http://dev.jqueryui.com/ticket/2973">http://dev.jqueryui.com/ticket/2973</a> - needs further research about how to properly configure shitty trac so this doesn't happen anymore <a href="http://dev.jqueryui.com/ticket/4158">http://dev.jqueryui.com/ticket/4158</a> - where are those links, and can't they be fixed already in the website trunk? Thanks on any input and/or help you can provide with solving those issues. Cheers, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, we recently had the discussion already about hat to do with the accurateIntersection option in Sortables. For those that don't know the specifics, the only change this option brings is that if set to true, outerWidth/outerHeight is used instead of offsetWidth/offsetHeight for offset calculations. Now it seems that jQuery UI 1. 7 will most likely include jQuery 1.3.2, which brings major performance gains in the outerHeight/outerWidth methods, so I propose we remove the option entirely for 1.7 already (and default to using the outer* methods).<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, see ticket: <a href="http://dev.jqueryui.com/ticket/4063">http://dev.jqueryui.com/ticket/4063</a> We need to find a way to either a) remove the margin from the slider entirely b) use absolute 'px' values in the CSS c) find a way to convert the 'em' value to 'px' Anyone?<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, this morning, I've introduced a couple of changes to our server and domain structure that will help us with the traffic. The details: - <a href="http://ui.jquery.com">ui.jquery.com</a> + <a href="http://jqueryui.com">jqueryui.com</a> now point to a entirely new server only used for the production website - <a href="http://jqueryui.pbwiki.com">jqueryui.pbwiki.com</a> has been replaced by <a href="http://wiki.jqueryui.com">wiki.jqueryui.com</a> - <a href="http://ui.jquery.com/bugs">ui.jquery.com/bugs</a> now forwards to <a href="http://dev.jqueryui.com">dev.jqueryui.com</a> (along with the query string and paths) This is how our domain stack will look like from now on (and should be communicated to the outside): <a href="http://jqueryui.com">jqueryui.com</a> - main jQuery UI website <a href="http://dev.jqueryui.com">dev.jqueryui.com</a> - bug tracker <a href="http://blog.jqueryui.com">blog.jqueryui.com</a> - jQuery UI blog (not live yet) <a href="http://wiki.jqueryui.com">wiki.jqueryui.com</a> - jQuery UI planning wiki The server seperation: old server: trac + ui-dev new server: ui website Cheers! Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hi everyone, I've spent quite a few hours on the website today, it's getting in much better shape now. Here are the changes: - removed applications folder entirely - removed 40% of rewrite rules - finished download page (dependency checks etc) - changed documentation/ links to docs/ - split the "Demos & Documentation" tab into "Demos" and "Documentation" again - "Documentation" links to docs/, which is a fully standalone 100% replica of <a href="http://docs.jquery.com">http://docs.jquery.com</a> with a new left hand navigation that includes more documentation focussed links (i.e. effects) The demos section still features inline documentation of course. The standalone docs are just needed for anything not browsable through the demos (i.e. effect documentation). It still needs heavy styling though, and the demos need to be loaded. It'd be great if Todd/Maggie could work on that today/tomorrow. Together we can get that release out in no time! Cheers, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey team, all embedded documentation from the wiki, - static documentation pages (theming pages, changelog etc) - theming tab on widget pages - all other tabs on documentation pages (these come from the processed xslt) are now synced automatically. The first two use my caching and fetching script, meaning they're updated every 10 minutes right now (we can set that to a higher/lower time for performance reasons on production). The scripts that are generated through the ant task are updated using the general cron job (I think every 5 min). Cheers, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey guys, when investigating why some of the jQuery UI demo pages didn't work with 1.3rc2, we found out it's related to 1.3rc2: We were doing the following: $('> :eq(0)', context) this worked in 1.2.6, in 1.3 you receive: Syntax error, unrecognized expression: > The bug only happens if 1) The first character is > 2) the immediate following part is a pseudo-selector (i.e. "> *:eq(0)") 3) The pseudo-selector has brackets (i.e. "> :input" works) This is definitely a regression. Any ideas?<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey guys, I figured recently that the :input selector matches select boxes but no options. Since we use it to cancel events in jQuery UI, some users reported select boxes not to work when draggables are used, for example. See the bug report here: <a href="http://dev.jquery.com/ticket/3791">http://dev.jquery.com/ticket/3791</a> Do you see any reason why adding <option> to the matched elements would be problematic? Thanks, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey everyone, to speed up our website, I've added caching to our Apache configuration. It's still pretty conservative right now, with expire headers set to 1 day only - if we've pushed out 1.6, we can set that for the production website (in the case of images at least) to 1 month. This also means that if you work with <a href="http://ui-dev.jquery.com">ui-dev.jquery.com</a>, don't be surprised if your new commit doesn't show up immediately - clear your cache and try again. If there are any concerns or issues, give me a reply. Cheers, Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>
Hey John, first off, nice job with the live query / jQuery.support commits! I'm almost /feeling/ the improvements, without even testing them :) I'm curious about a couple lines in the commit, thought you might explain them to us: - ajax.js: It seems instead of a substitute, you simply removed the safari check. What was the reason for this? - core.js: removal of opera/display fix, color() method with safari fix, the whole display none logic - does this mean we simply don't support getting some style values for display:none elements? (which is fine, just wanted to check) Other than that, it looks all good, a lot of nice fixes / better logic in there, especially in event.js. Thanks! Cheers, Paul
Hey everyone, since it's on the roadmap in the UI wiki and needed for the grid component, I went ahead and created a infinite scrolling widget, which I believe will become incredibly popular (since no other library has such a widget in line apparently [disclaimer: probably Dojo, they have always stuff I don't know :D]). While the spec isn't written yet, the script's usage is a simple as possible as possible, yet flexible. First, you start with a markup like the following: <div id='myTable'> <table> <tbody> <tr><td>Sample row</td></tr> </tbody> </table> </div> A couple of instructions on the markup: 1) 'myTable' must have a overflow-y/overflow value set to 'scroll' 2) The table must be prefilled with a first set of items, in this case, one sample <tr>, so the script can detect the correct row's height (relative to styling etc) That's it for the markup, here's the initialization code, for a easy configuration: $('#myTable').infiniteScrolling({ total: 1000, // This tells the script the total rows coming from your backend chunk: 20 // if set to 20, 20 rows are fetched at once, reducing requests scroll: fn }); now, in order to make it work, the scroll callback must be loaded with a function that handles the ajax part for retrieving the actual rows. The callback works like this: function(e, ui) { // ui.chunk gives you the actual requested chunk of data // ui.start gives you the item to 'start' with, from the actual chunk // now get your rows, if you set options.chunk to 20, retrieve 20 rows, from the start index ui.start .... // after the rows are requested, use the method ui.fill() to give the rows back to the script // 'data' can either be: // a) a flat array with <tr> nodes or their string representation // b) a flat array template objects, i.e. { 1: 'something', 2: 'something else' } // If you want to use option b), you also have to set the option 'template' on the initialization to // something like '<tr><td class="{$1}">{$2}</td></tr>' ui.fill({ chunk: ui.chunk data: [...] }); } and that's it, you're done! The script calls scroll() the first time upon initalization, and then everytime the user scrolls horizontally and loads chunks accordingly. Of course, there's a little more that you can configure: 'preload': default is 1, tells the script how many chunks should be preloaded before and after the visible offset 'delay': default is 100 (ms), tells the script to wait a little once the scroll callback happened before your scroll function is actually executed - this is much better, because then you won't have thousands of requests if you scroll fast, only one once you end scrolling. Of course, it's not production ready yet, and misses a few things / has a few quirks: a) right now you need to pass a total row value, so it's actually finite scrolling, not infinite - as next feature, I'm going to implement that if you don't use the total option, rows are not allocated in advance, and you can scroll as long as the callbacks retrieve data. b) cache revalidation: there's no way to revalidate the cache right now - once a chunk has been filled, a cache value for this chunk is set to true and it's never requested again. There needs to be a mechanismn to manually rerequest it, and/or to request it after a certain amount of time. c) There's a strange firefox bug, causing the actual chunk offsets to be off a bit - the rows in the visual test have a height of 22, in all browsers - however, in FF, the *actual* height, if you measure it, is 23 for some rows (for example, the forth) - this is insane - even more so, since Firebug correctly reports '22' in the layout panel, but 45 as offset for the next one (should be 44) feel free to check it out at the jQuery UI SVN in branches/dev/grid, the visual test is at tests/visual/infiniteScrolling.html, and it's already implemented and up and running in the grid demo. Cheers! Paul<br clear="all"> -- Paul Bakaus UI Architect -- <a href="http://paulbakaus.com">http://paulbakaus.com</a> <a href="http://www.linkedin.com/in/paulbakaus">http://www.linkedin.com/in/paulbakaus</a>