Dojo Toolkit

I was off to a good start with my project (dojox chartin improvements) with the candle plot. We basically followed the same model as in columns and did it. Actually it's clustered columns. I don't know whether this is a valid thing in these type of ... [More] plots... Anyway I thought is a nice feature to add so that you can actually add more than one series Add Series for chart looked something like below... chart.addSeries("Series A", [{lx:1,lq:2,md:3,uq:9,ux:10},{lx:1,lq:2,md:3,uq:7,ux:8}]); But Eugene Suggested that it would be more appropriate to have long names like.. chart.addSeries("Series A", [{low:1,open:2,median:3,close:9,high:10},{low:1,open:2,median:3,close:7,high:8}]); I also feel like that is a better way except for taking up little more space which is bad! Anyway your suggestions are most welcomed! So here is the sample screen.... [inline:candle.PNG=Candle Plot] Also you may call this box or finacial plot too. Eventhough I can't find any differences you might.. So please comment if you want to add/remove/modify any features.. [Less]

Today is my last day of finals, so after today I'll finally be able to focus on my project. So far, all I've been doing is studying to finish the quarter well, but I have been thinking about how to solve some of the issues associated with reusing ... [More] animation objects on dynamic elements. I had a brief conversation with my mentor about how to solve the "stale values" issue that surfaces when interrupting an animation. This is especially important for toggling animations. So, tomorrow I plan to sit down and finally get started on playing with a few ideas I have for working out the kinks. [Less]

Hi dojoers, I would like to share my experiences lately... This experience is very important for us, computer people who likes to sit in front of computer around more than 10 hours per day! Couple days ago until now... I am suffering heavy shoulder ... [More] pain (sometimes headache) why? Because I stay in front of computer everyday without any proper treatment... I quickly evaluate myself and made this checklist: - Doing sport? Minimum 3 hours per week... - Rest for every 1 hour in front of computer? - Drink water for 1.5 liters per day? - Check your lens (if you wear glasses) - How is your computer position? How about the lightning? If you dont do one of the checklist... I suggest you to do it... The pain that I am suffering is really awful... I remembered the worst time, I could not do my daily activities... I felt dizzy and wanted to throw up... It was very nasty... Yet... I am feeling shoulder pain... I know the deadline is coming in but your health is more important... What for the money if you are sick? Please beware... [Less]

Hi Dojoers, It seems optimization is my friend in this project with Dojo... This time I would like to share my idea about what I did (mainly optimization both design and implementation) for couple days. As remainder, I am doing the gfx library ... [More] enhancement, where real time processing is needed hence a fast and robust implementation plays cruical part in here... This idea is inspired from programming language called Haskell. This cute programming language is kind of late processing, where any calculation/procession will not be called unless it is needed for example, you say: foo1::Int = 1 2; <- at this point, it is not calculated foo2::Int = 3 4; <- not touched at all foo3::Int = foo1; <- foo1 is calculated then put into here... This feature snapped the idea and I applied it into two different areas in my library, in world transformation and view transformation optimization... Let me explain the background first, note that in the 3d graphic environment (especially OpenGL) environment needs to be drawn whenever view state changes (for example camera view changed or object transformation is moved or a new object is created and put into scene) Now lets see the old ways to draw/render the environment: // Main rendering process. draw: function(/*Scene*/ scene, /*RenderObject*/ list) { // Clear color and depth buffer either in client or server side. this._renderer.clear(this._renderer.COLOR_BUFFER_BIT | this._renderer.DEPTH_BUFFER_BIT); // set up our viewing parameters this._renderer.matrixMode(this._renderer.PROJECTION); this._renderer.loadIdentity(); this._renderer.gluPerspective(45, 1, 0.1, 1000); this._renderer.matrixMode(this._renderer.MODELVIEW); this._renderer.loadIdentity(); this._renderer.gluLookAt(1, 1, 10, 1, 1, 1, 0, 1, 0); // Draw the object // Apply world transformation this._renderer.translate(0, 1, 0); this._renderer.rotate(10, 0, 1, 0); this._renderer.scale(1, 1, 1); // Draw triangles this._renderer.drawElements(gl.TRIANGLES, indices.length, indices); // Show the processed buffer. Note this uses double buffering // technique. this._renderer.swapBuffers(); } Ah forgot to mention, OpenGL uses matrix for calculation structure such as translation, rotation or scaling... The implementations behind gluPerspective, gluLookAt, translate, rotate and scale are heavy duty multiplications and divisions (If you do not believe me, I could give you the design of the OpenGL implementation) and see that those functions are calculated everytime drawing process is initiated... // set up our viewing parameters this._renderer.matrixMode(this._renderer.PROJECTION); this._renderer.loadIdentity(); this._renderer.gluPerspective(45, 1, 0.1, 1000); this._renderer.matrixMode(this._renderer.MODELVIEW); this._renderer.loadIdentity(); this._renderer.gluLookAt(1, 1, 10, 1, 1, 1, 0, 1, 0); // Draw the object // Apply world transformation this._renderer.translate(0, 1, 0); this._renderer.rotate(10, 0, 1, 0); this._renderer.scale(1, 1, 1); Would it be nice if there is a way to calculate it once and just copy the calculation result into OpenGL straight away as long as it is not changed... Luckyly, OpenGL support a way to do that (Phew!) So what I did is basically change this partition code into more complex algorithm... this._renderer.matrixMode(this._renderer.MODELVIEW); this._renderer.loadIdentity(); this._renderer.gluLookAt(1, 1, 10, 1, 1, 1, 0, 1, 0); See that, whenever the view is changed then it is necessary to calculate (in function _calculate) it... Then the _execute function will load the calculated matrix by using loadMatrix (Thanks OpenGL!!!) // Helper function to execute camera transformation. Firstly, it will check // whether the camera is updated or not... If it is updated then calculate the // matrix for synchronization. apply: function() { if (this._isUpdated) { // Do the calculation. this._calculate(); // Set back the status. this._isUpdated = false; } // Execute the code. this._execute(); }, // Execute specific transformation by given properties. _execute: function() { this._renderer.matrixMode(this._renderer.MODELVIEW); this._renderer.loadMatrix(this._modelViewMatrix); }, // Calculate transformation so that they become synchronized each other. This method is kind // of pull data technique and it need to be implemented for each specific engine. _calculate: function() { // Insane calculation as OpenGL specified... }, This function is very effective and boosting the performance a lot, note that there are so many factors to change the environment state but it is kind of low chance to have all calculated everytime... so by using this design, it could save a lot of computation hence increasing performance a lot... I would like to design next optimization which is event to control when the right time to render the environment of 3d graphic... Note that, I described about the view states that cause environment needs to be drawn... It would be nice if there is a way to draw the envornment whenever view state is changed than periodic drawing. That is what I have been doing in couple weeks... Trying to set up high performance infrastructure for the gfx library... Of course, you guys may know better ways to do please put comment... I would really appreciate it... Even though just comments or suggestions.... Thanks! [Less]

For the ones who were not able to join the first dojo.beer() in Berlin last weekend (may 31st) I have uploaded some pictures: http://flickr.com/photos/27297860@N02/with/2548123434/ It was great to meet everybody in person, Berlin was a great host and ... [More] all we did - almost ;) - was drinking cold beer, lying in the sun and talking dojo. I am really looking forward to the next meeting (in the States, Europe, Japan, Berlin or any other place :) ). [Less]

There's been a lot going on in Dojo-Land recently, though our blog is no indication of that fact. Here's a meta-update encompassing all that I can recall: Zend Integration The last we heard was the initial announcement. Alex and I been lurking in ... [More] #zftalk and fielding questions on and off. There's been a lot of bleed over from #zftalk into #dojo, Zend Framework community members poking around, feeling the waters as it were. Lots of great questions being asked, lots of interest all around. We also participated in a Zend Framework webinar, talking with ZF Team members and the community about the proposal: Probably the most common question is: Valid HTML. I'd like to take a second to remind everyone there is nothing invalid about using Dojo. Some see the dojoType attribute in most of our tests and cry foul, "standardss heresy I say!". What most people overlook, however, is that dojoType, and the dojo.parser is an additive characteristic of the Dojo Toolkit. You have to explicitly dojo.require("dojo.parser") in order for Dojo to even understand what a dojoType is. If you want to write valid (x)html pages: you can. Like everything else in Dojo: you don't have to if you don't want to, though it makes for exceptionally easy prototyping. The djConfig attribute being the other invalid attribute we introduce, also entirely optional. To specify valid runtime configuration options for dojo.js: <script type="text/javascript">         var djConfig = { parseOnLoad:true, isDebug:true }; </script> <script src="http://o.aolcdn.com/dojo/1.1.1/dojo/dojo.xd.js"></script> ... which is an interesting point, because parseOnLoad:true indicates you probably have a dojoType somewhere. Leaving djConfig null (as a var or attribute on the script tag) leaves parseOnLoad:false ... In it's most simple form, including Dojo is: <script type="text/javascript" src="http://o.aolcdn.com/dojo/1.1.1/dojo/dojo.xd.js"></script> All that said, Zend and Dojo will be working very hard to provide all the available power of the Dojo Toolkit to the Zend Framework users, maintaining standards, though offering the option of the "invalid" convenience methods. It's all about choices and rapid development. I've updated the dojo.moj.oe demo to be valid HTML 4.01 - it involved removing the trailing '/' in my self-closing LINK elements, and wrapping a div around some inputs, which I find to be more difficult on large sites than remembering not to use the dojoType attribute. Speaking of the CDN Google has started their own Ajax library delivery system, offering Dojo >= 1.1.1 as a cross-domain build: Give it a try just like you would the AOL CDN: <script src="http://ajax.googleapis.com/ajax/libs/dojo/1.1.1/dojo/dojo.xd.js"></script> Note, there a some minor differences between the AOL and Google CDN, at least regarding this initial release. The theme CSS on the Google CDN has not been compressed or otherwise optimized, namely. Most examples you see using the AOL CDN, however, can simply be run by substituting the new URL for your dojo.xd.js - all the same X-Domain nuances still apply ... More on Google To follow up on Shane's recent Google dojo.data stores, I took the liberty of creating SMD's defining all the available Google Ajax API's: web, book, local, news, books, images, and videos, demo'd as a single test file, and a wrapper for the Google Translate API. Accessing the entire Google datacenter is as easy as: dojo.require("dojox.rpc.Service"); dojo.require("dojo.io.script"); dojo.addOnLoad(function(){     var goog = new dojox.rpc.Service(dojo.moduleUrl("dojox.rpc","SMDLibrary/google.smd"));     goog.webSearch({ q:"Dojo Toolkit "}).addCallback(function(data){         /* first group of results */     }); }); The google.smd library will be available in Dojo 1.2, though you can use the SMD with 1.1, and probably 1.0 versions as well. I'm also testing the lazy loading of google-analytics code over on my blog, in an attempt to speed up when the DOM is actually ready when including ga.js in a page. (For me, as often as I clear cache, google-analytics and ga.js (10k) can take upwards of 5 seconds to load, causing weird rendering issues on widget-riddled pages). So far, the test is producing data, but I'd like to let it run more than a day ... If everything goes okay, it will show up as a new tool in your favorite Toolkit, giving you the ability to dynamically track your static and Ajax-refreshed pages without adding any markup to the page, and simply specifying your Urchin tracking number. Currently, it's hacking it's way into djConfig, reading an urchin: param from there ... Ideally, you'd be able to specify it at runtime: new dojox.analytics.urchin("UA-123456-7") and the rest is taken care of. API details though, I'm just glad it's working. Updated Code Completion Todd @ ActiveState updated my makeCix.php script, which was initially announced back in January as an experiment. There were some issues with my attributes (and overall understanding of the schema), so a big thanks to ToddW for stepping up and working out the bugs. If you don't remember makeCix: It scans across the Toolkit source tree, grabbing inline summaries and function signatures, and generates a .cix file compatible with the free Komodo Editor ... It's a great editor, and having full Dojo 1.1 API completion is great. They've updated their official release of the code completion plugin, and should work with all newer version On a personal note I'd like to welcome to our ranks the newest (and youngest ever) Dojo contributor: Ethan Zachary Peller, Spawn of our esteemed Adam Peller (DojoX BDFL). Adam has been on hiatus since the exciting moment though will be returning to 'work' shortly. We're currently checking the validity of a CLA signed by a 3-week-old, to ensure his contributions (and influnce on the Toolkit vicariously through peller) are clean. Looking forward to his, and all of your (yes, you), contributions. Salud, peller, may health, happiness, and an unnatural ability to write code bless your child's life. You know, its hard to call what we do here "work" though, there is entirely too much fun involved ... which reminds me: dojo.workers() I secretly and silently added a new Dojo demo: dojo.workers(). I wanted to do something fun showcasing both the Dojo Toolkit's ability, and the hard working committed people behind it's maturity. We are a diverse bunch of people from all walks of life, and this demo gives them at least partial credit. There are a few omissions, unfortunately not everyone who has committed code to Dojo took the time to "approve" being listed (by adding themselves to the json file listing everyone) so it's a partially incomplete list. Feel free to update the users.json file in http://svn.dojotoolkit.org/dojoc/demos/skewDemo/ with your information, and alert me to the change ... viola. On an aside, the dojo.workers() page would validate by removing some trailing '/'s and removing an empty unordered-list. Hope that gets everyone up to speed, and I apologize if I missed anything new and pressing. [Less]

SoC Week 1: Semi-random thoughts Here's a big brain dump of stuff I've been thinking about this week. As a reminder, I'm working on a drag & drop form editor. Heed warnings I'm fairly new to the dojo toolkit, so I figured I'd be doing lots ... [More] of exploration in firebug. Thus, I should pull in the awesome dojox.help.console during page load instead of always dojo.requiring it from the console. Great idea, right? Wrong. When you load help, there's a two-line warning. The first line: Script causes side effects (on numbers, strings, and booleans).Turns out they're not lying. I was getting inexplicable errors where dojo was using IE methods in firefox. Is the browser sniffing broken? Let's look at the dojo.isIE attribute, which should be the number 0. >>> typeof dojo.isIE "object" >>> dojo.isIE 0 __name__=dojo.isIEdojox.help turns everything into objects, which is great when you want to know more, but not so great when you want code to work. The next line of that warning? Call dojox.help.noConflict() if you plan on executing code.dojox.lang.aspect ftw One of the first problems I encountered with dojo.dnd is that there's no clean way to get a handle on the dnd object if it's created declaratively. It is possible to give it a jsId, but that introduces a global variable, and you can't do easy lookups as with dijit.{byId,byNode}. To work around this, I used dojox.lang.aspect to introduce a before method on the dojo.dnd.Container constructor that grabs the current object and node and sticks them in a dictionary for easy lookups in my code. Someone out there is probably shaking his head, muttering about a better way to do it, but I thought it was a neat way to add the functionality I needed. Javascript gets a lot of crap (much of it well-deserved), but I'm having a lot of fun discovering the powerful things it allows you to do. Working with real closures is awesome, and the dynamic features that libs like dojo.lang.aspect exploit open up a world of possibilities that aren't available in most languages. svn: patches not welcome Dojo, like a lot of other projects, is still using svn. This is great for, um, somebody, but I think it's more detrimental than good for open source projects. Reading through dojo sources this week to figure out how things work, I spotted little typos in comments and other minor things, but I didn't do anything about it because there's too much hassle. To get a fix in , I'd have to create the patch, make a silly ticket for this little fix, attach my patch there, and hope someone downloads the attachment and commits the code. It's a pain to share my code, a pain for someone to test/apply the fix, and at best I only get indirect credit for the patch, with my name in the commit summary. I'm doing this work for fun, and the abundant fame and fortune is just a plus, but there's something special about seeing yourself as an author in the changelogs for the first time. I really like the git style of sending the patch directly to the mail list. There's tools that make it easy for the submitter to get the patch out and easy for the committers to get the patch in. It's a beautiful thing, removing the barriers for contributing to your favorite projects. javascript style Javascript gives you lots of power by allowing function and object declarations almost everywhere, but this can lead to some funky-looking code. I'm still trying to figure out what my js style is, where braces and indentation belong (js2-mode has been a big help with its variable indentation support). This is strange for me coming from Python, where PEP 8 is the style guide, and there's not much opportunity to color outside the lines. Dojo does have a style guide, but there's a lot more room for interpretation. Code on the way Once I get my current quick-and-dirty code into a less embarrassing state, I'll be uploading it to the soc branch and writing another post about what's going on. [Less]

This coming Saturday, 31st of may, everyone who is in or near Berlin, or has a chance to travel to Berlin, is invited to join the first (hopefully at least annual) dojo.beer(). This will give everyone a chance to get to know each other, see and hang ... [More] out in Berlin and hopefully find some good time to talk dojo. We'll meet at 15:00 at the Weltzeituhr, move over to a nice Spanish restaurant at 19:00 and for the ones who like, we will attack the Berlin clubs right after that. I am not sure how much of a good internet connection we have, bring your mobile phones and we can connect to the interweb using those :P If you know that you will be there, please send an email to rsvp _at_ dojotoolkit.org so we can book some extra space. Otherwise you can ping us on IRC (klipstein, mccain, nonken). Exact details can be found here: http://wolfram.kriesing.de/blog/index.php/2008/the-timeline-for-dojo-mee... or http://dojocampus.org/content/dojo-meetings/dojobeer-berlin-may-31st-200... Hope to see you in Berlin [Less]

Today marks a very exciting milestone for two great Open Source Products: The Zend Framework and The Dojo Toolkit. Zend aims to provide out-of-the box Ajax functionality capable of communicating directly with the Framework, and Dojo aims to provide ... [More] the solution. The parellels between the projects make this integration a perfect match. Both follow a "use as needed" creedo, have wonderfully active communities, and crystal clear licensing. As a Dojo developer and PHP enthusiast, I'm proud to be apart of this integration effort, working directly with a number of Zend Framework team members for the best possible results. So what exactly does this all mean? Here is a FAQ written to help: Zend Framework and Dojo Partnership FAQ 1. What are the Zend Framework and Dojo Toolkit teams announcing? Zend Framework and Dojo are announcing a strategic partnership to deliver an integrated solution for building modern PHP-based Web applications. In order to deliver an out-of-the-box experience Zend Framework will bundle the Dojo Toolkit and will feature Dojo-specific components. 2. Why did the Zend Framework and Dojo teams decide to work together? There are many synergies and similarities between the two projects and their communities, including: a) Licensing Zend Framework and Dojo are both licensed under the new BSD license, allowing end users to integrate, alter, and distribute each project as they wish. In integrating with Dojo, Zend Framework continues to deliver business-friendly licensing along with its full Ajax support. b) IP Purity The Zend Framework and Dojo project both require all contributors to sign Apache-style Contributor License Agreements, which mitigates the risk of accepting contributions that infringe upon third parties' intellectual property rights. c) Design Affinity Both projects have similar design philosophies, including a strong emphasis on use-at-will architecture. Additionally, each has rigorous quality guidelines with strict unit testing and coding standards. d) JSON Format While Dojo can accept XHR responses in a variety of formats, JSON is the preferred response format. Zend Framework fully supports JSON for Ajax interactions, and already has a variety of helpers to facilitate data transmission via JSON. JSON is a lightweight format, can be evaluated directly in Javascript, and presents an elegant solution to the problem of data representation in XHR requests. e) Comprehensive Ajax Solution Dojo provides a comprehensive solution for rich web user interfaces. Many other toolkits either abstract common DOM-related actions to make remoting more efficient or focus solely on the UI layer; Dojo provides utilities for all of these. f) Use of Standards Dojo not only implements published standards, but also drives them. For example, members of the Dojo Foundation are working on draft versions of the JSON-RPC, JSON-Schema, and Bayeux Protocol specifications to promote interoperability among JavaScript libraries. In addition, Dojo is adopting and implementing standards driven by the OpenAjax Alliance including the OpenAjax Hub for interoperability. g) Support There are dedicated organizations behind both that allow customers to benefit from a fully supported stack. Zend offers support for PHP, Zend Framework and its application server offering while SitePen has support offerings for Dojo. Depending on customer demand the companies may also create joint support offerings in the future. h) Communities Both projects foster very strong and active communities that can support each other. Visit http://dojotoolkit.org/community and http://framework.zend.com/community for more information on how to participate. 3. What if my favorite Ajax toolkit is not Dojo? How does this fit in with your use-at-will philosophy? Zend Framework will continue to be largely Ajax toolkit agnostic. While we will ship Dojo with Zend Framework as our preferred Ajax toolkit, only those who seek out-of-the-box Ajax functionality in the standard library will require Dojo. Additionally, we expect that the various Dojo-related components and helpers added to Zend Framework will serve as a blueprint for similar components serving alternate Ajax toolkits developed by the Zend Framework community. While we don’t have immediate plans to support them directly, we may ship such community contributions in the future. While the Zend Framework team feels that Dojo is the right choice of JavaScript toolkit to build our Ajax experience on, it is not necessarily the case that Dojo is the right toolkit for you or your project. In addition, it may not be worthwhile to refactor existing code to standardize on Dojo. You may find that features found in other JavaScript toolkits far outweigh any benefits of our collaboration. The Dojo Toolkit project will, for its part, also continue being server-side framework agnostic. In essence, this collaboration should not be taken as a move towards exclusivity in either project; rather, it adds features in each project to facilitate interoperability between Zend Framework and the Dojo Toolkit. 4. What components in the Zend Framework will be affected by this integration? Will any of this work benefit integration projects for other Ajax libraries? Currently, we intend to add the following components: o A dojo() placeholder view helper to facilitate Dojo integration in your views, including setting up the required script and style tags, dojo.require statements, and more. In essence, this work will support and enhance Dojo's modularity at the application level. o Zend_Form elements that utilize Dijit, Dojo’s widget collection and platform. This will simplify creation of Zend_Form elements that can be rendered as Dijits. For instance, highly interactive widgets such as calendar choosers, color pickers, time selectors, and combo-boxes will be provided in the initial integration project. o A component for creating dojo.data-compatible response payloads. dojo.data defines a standard storage interface; services providing data in this format can then be consumed by a variety of Dojo facilities to provide highly flexible and dynamic content for your user interfaces. o A JSON-RPC server component. JSON-RPC is a lightweight remote procedure call protocol, utilizing JSON for its serialization format; it is useful for sites that require a high volume of interaction between the user interface and server-side data stores, as it allows exposing your server-side APIs in a format directly accessible via your client. Dojo has native JSON-RPC capabilities, and Zend Framework will provide a JSON-RPC implementation that is compatible with Dojo. These features will be added to Zend Framework; no components will be re-written to make use of Dojo. With Dojo support in Zend Framework, we hope to see ZF community contributions that follow this blueprint to add similar functionality for other Ajax toolkits. 5. I have feedback regarding the proposed method for integrating Dojo and Zend Framework. How can I deliver this feedback? The Dojo integration will undergo the standard Zend Framework proposal review process. Please watch the main developer’s mailing list in the coming days for a proposal. You will be able to give feedback as with any proposal. 6. Could I contribute support for my favorite Ajax toolkit to Zend Framework? Absolutely. However, we will only officially support Dojo components for the foreseeable future. 7. Will Zend Framework ship Dojo? Yes. 8. Is Zend joining the Dojo foundation? Zend has signed a corporate CLA with the Dojo Foundation in order to enable Zend staff to contribute to Dojo as needed and has begun the process of becoming a new Dojo Foundation member. 9. Is the Dojo team joining Zend Framework as contributors? Yes; the Zend Framework project already has CLAs on file for Dojo contributors. 10. If I have signed a Zend Framework CLA will I be able to contribute to the bundled Dojo library? We will not allow contributions to the bundled Dojo library through the Zend Framework project. We will bundle the latest, unmodified version of the Dojo library in Zend Framework; all contributions to that library should be done through the Dojo Foundation according to their policies. However, we may create custom modules to extend Dojo that contain contributions from Zend and the Zend Framework community. The Zend Framework team does not expect to ship custom extensions as part of our initial Dojo integration project. 11. What license governs Dojo? It is dual licensed under the modified BSD License and the Academic Free License version 2.1. For details see http://dojotoolkit.org/license 12. Will Zend Studio add support for Dojo? Will Zend Studio also support other Ajax toolkits? Zend Studio will continue to enhance its Ajax support in upcoming versions. As part of these enhancements it will likely also support individual toolkits including Dojo. We are evaluating enhanced support for Dojo widgets used in Zend Framework components. 13. I have questions which you haven’t answered in this FAQ. How can I ask them? Alex and I will be participating in a Webinar on Tuesday May 27th with Zend Framework team members to deliver a short overview of the proposed integration, followed by a Q&A with the audience. In addition, both Zend Framework and Dojo community members are encouraged to communicate on thee main development lists of either project. A w00t is in order here. [Less]

Hi I'm Osandi Malage, from Sri Lanka. I'm going to work on charting improvements for dojo this summer. It includes adding some new chart types like candle plot & bullet plot. Also I'm planning to work on time series support, multi axis support ... [More] and better integration with dojo data. These days i've been playing with the existing system. Also I must thank dojo community (specially Eugene) for their great support! Since gsoc official coding starts from next week, I will keep the progress posted here. I'm always open for your suggestions! :) [Less]