I’m (quite:-) closely following the Morfik saga since late 2005 and did have many long discussions with Aram, Fuad, Mauricio and Shah about direction in the past three years, sometimes pretty heated chats about the vision and execution, so I’m not an unanimous supporter of everything Morfik, but version 2 is a very significant step in the good direction.

The new visual design concept with themes, states, popup customization, etc. is a major step forward for web application design. I see real innovation here.  Go and check out the tutorial videos, it is worthwhile! I wouldn’t say that I’m fond of the sixties look of the new demo apps (and of the new Morfik site itself), but the design flexibility is impressive and the functionality seems to be very well thought out.

I’ll be looking into the new version in the coming days and weeks (we do have some projects that now will have to be ported anyway, especially our PMAP controls and libraries), so stay tuned for new comments to come.

First of all, do read the comic strip by Scott McCloud and the Chrome Team. It is in itself a piece of marketing art and although its primary intended audience may be journalists and less technical people, it is a statement of how serious Google is about Chrome and full of hints for conspiracy theorists among us.

So some of my first thoughts will follow…

OS-within-the-OS

Chrome will be the OS-within-the-OS for Google: most productivity and line-of-business applications can now be successfully turned into web applications. Arguably more and more of them are better in ease of use than the original apps, due to the richness of the platform and due to being re-engineered from scratch (UI wise) with many usability lessons learned since.

The big issue is compatibility. There are at least four major players now: IE, Firefox, Safari, and Opera (plus the mobile editions), and their abilities are spread on a wide spectrum, to say the least. The unquestioned market share leader (IE) is trailing behind in almost all important areas (like performance, usability, standards conformance) and there are subtle but important differences among the others. This makes web application development very costly and time consuming. Even with frameworks like DOJO, Prototype and tools like GWT and Morfik you will encounter compatibility issues and missing functionality (e.g. lack of a consistent graphics layer, like SVG or Canvas).

If we had a browser that

• has #1 market share,
• is consistent among all the major operating systems (Windows, Mac OS X, Linux) and mobile OSes like Android and iPhone OS X (probably Symbian and Windows Mobile, but I would not hold my breath for those),
• is performant, secure and robust,
• and has some additional features like support for off-line operation, a strong graphics layer (for graphs and graphical apps), sandboxed native filesystem access, support for push technology (like COMET), drag & drop desktop integration, and a mature, efficient and familiar development platform (e.g. Eclipse/Java/GWT),

then most applications can be implemented on this platform regardless of the underlying native operating system.

Just think for a second when you used Microsoft Word the last time: I used to be in-and-out all the day, but recently it happens that I don’t open Word for weeks, and then only to edit a “legacy” document that originated from the “old era”. Most of my new documents are emails, Google Docs, or some other on-line properties (god, what that does to privacy, though, so don’t put all your documents on-line!).

Technology tie-ins

There will be technology tie-ins all over the place. Although Google is a huge animal and its projects are only loosely coupled (waving off the monopoly power arguments), saying that the Chrome team accidentally asked the Android team about WebKit love is amusing.

Gears integration is only for starters. I expect that GWT and Chrome will be “optimized together” pretty soon. Google Docs, Maps, etc. will gain in performance, stability and functionality if run on Chrome.

Then the primary business of Google is ads: now it will be able to collect even more information about us (although since gMail and Desktop Search they already have some data on you ;-).

BTW I wonder when Desktop Search will be integrated into Chrome…

Head start

Chrome may have a head start over all other browsers.

Current generation Firefox, Safari and Opera are pretty level on performance (relative to lackluster IE) and at least Firefox and Safari are engaged in further speeding up JavaScript with adding virtual machines similar to Chrome’s V8 (on paper); they are also keen to match each other in standards compliance and usability, but while Chrome addresses all these issues, it also brings a new architecture to the table with the promise of marked enhancement in security, memory performance and robustness, plus the native integration of Gears.

The others will have to play catch-up. And Google has the resources to compete — it is single-handedly financing Firefox at the moment.

Market share

In order to be successful, Chrome will have to establish market share.

In the consumer space all the good virtues (speed, stability, security) will play well, together with the hippie word of mouth marketing of comic strips and oh-so-accidentally-released-a-bit-ahead-of-time trickery.

The much harder nut is the corporate market. It takes years for corporate IS departments to certify products for use. Here being OSS will help (the corporate world is getting into love with OSS), but the primary message can be security: if Google can deliver on its promise of security (both process separation and malware filters), it will be salvation to IS departments fighting with the dilemma of supporting more and more intranet/extranet web applications and weak security of the very same applications.

Being a consistent web application platform on all important OSes will also come handy – it makes corporate web app development much simpler.

Google will definitely push Chrome with subliminal tactics (e.g. “off-line mode and advanced features of our web apps working best with Chrome” splashes).

Still, it will be a hard sell — they’ll need some killer apps to get rolling.

Missing bits

I’m also missing a few things, the most prominent being additional web application security.

Chrome is a rich client platform, and rich clients run most of their code on, well, the client. This used to be the case with traditional apps, but those were compiled to binary, so poking around required some skills. Now Web 2.0 rich clients are generally made of JavaScript, which is quite readable and even can be changed on the run, making attacks against the code much easier than before.

Of course tools like GWT or Morfik will scramble and optimize the client code making it not a pleasure to read, but it is still the source code of the app that is downloaded and run in the browser. So it would be fine if some kind of run-time protection would be in place to prevent code morphing and allow code verification. The fact that each tab runs in its own process is promising, though.

Phew! So here are my first impressions — what if I stared thinking about this :-) Now it’s your turn!

Read on for technical details…

This is a prototype of a fully interactive (i.e. editable) Gantt (project) diagram created entirely with HTML & DOM, without any plugins/Canvas/VML/SVG whatsoever. We use only a minimum set of graphics (e.g. the project milestone diamond is one graphic image).

The client side was developed with Morfik, and we used Morfik M2 Beta 2.0.2.1 to create the browser application. It is hosted on Google App Engine, so the server side is written in Python (at the moment not much there:-). BTW since most of the code is independent from the Morfik Framework (does direct DOM manipulation), it is an easy port to straight JavaScript or to other toolkits, so let me know if you are interested…

Morfik M2 worked pretty well (it is a beauty), I needed it badly to create a client-only (or “Express”) version of the demo, so that I can host it on Google App Engine. Porting the code from the 1.x branch was pretty straightforward, even some issues have been fixed in the framework, although I still had to (re-)apply some of my patches :-) I did not use the fabulous effects library yet, as it does not work with IE in the current beta. Google App Engine was a snap to use – kudos to Shah Besharati for the integration sample! BTW I ended up using Komodo Edit 4.3 (free) for the GAE work – pretty neat.

Right now the Gantt is not really functional, all you can do is resizing columns, resizing the left/right side, expanding/collapsing WBS nodes, dragging and re-sizing task bars (without much effect on task data:-) and scrolling around (Alt+mouse scroll wheel will scroll horizontally). You can also switch between two visual themes. We will add more functionality later.

You can play with it freely, will not write anything back to the server. It is running on a high availability server, so I’m interested if you will have any issues.

In the next couple of weeks I’ll be working on the smarts behind the scenes to make it really functional. I’m open to ideas how to make it really useful, especially in a Web 2.0 way, so please keep the comments coming!

More later…

Our new Tree View will allow you to build trees of virtually any complexity, even containing HTML and the ability to command other parts of your user interface. It has been optimized to keep its performance sustained even with large trees. The PMAP code set contains both the full source code of the tree control and the continuously growing infrastructure functions we are using to build high performance controls and applications.

Some highlights:

• Full keyboard navigation.
• Multiple levels of customizability (style sheet, properties, events, etc.).
• Any tree node can contain valid HTML.
• Build the tree either through the control API, or with the help of XML or JSON.
• Performance optimized.
• Multiple selection support (with keyboard navigation, too!).
• Fixed height or elastic sizing based on content.
• Dynamic loading API for huge trees.

A few words about the demo:

• The default tree demonstrates a few things, like embedded hyperlinks, user interface commanding (will turn the pages in the Tab control in an event handler), and various icons.
• You can edit this tree or create your own in many ways
  • by building it up node-by-node,
  • by adding complete sub-trees defined in JSON format,
  • by specifying the entire tree in XML or JSON format, or
  • by generating a tree based on some parameters.
• You will notice that using the JSON interface is much faster than the XML interface. This is indeed a good comparison of the JSON v.s. XML support of your browser (see the Log for timing data).
• You will be able to customize many of the node and tree options.
• How to use the Items page:
  • First press “Set Items (XML)” (and be patient;-: it will set up a test tree that is in the editor on the right in XML format.
  • Then press “Get Items (JSON)”: the editor will be filled with the JSON version of the same tree.
  • Careful: press “Clear” below the tree control, to make the tree empty (it is lovely to have so many Clear buttons:-).
  • Now press “Set Items (JSON)”: your tree will be re-built much faster (just look at the log, on some configurations the difference can be 50 fold+!).
  • If you want to get back the original tree, just press “Default content”.

Feel free to play with the control and let us know of your findings (either by commenting here or by dropping a mail).

Dan Webb has some important thoughts about the Flash v.s. Ajax debate and the comments on his site and on the Ajaxian site are also thought provoking.

Indeed it is anyone’s guess how WPF/E, Flash/Flex/OpenLaszlo, AJAX (Backbase, YUI, Dojo, Prototype, Qooxdoo,…), GWT, Eclipse/RAP and Morfik will mature but this year will be definitely very interesting for web application development!

I won’t repeat the press release at the MorfikWiki site. What’s important, though:

• Morfik has improved tremendously in the last two months;
• stability is much better;
• the Framework is streamlined and got rid of some fat (more about this later);
• there is now good documentation in the form of the Developer’s Guide by Mauricio Longo (excellent, even I could learn a few new things;-) and the MorfikWiki site (that contains some of my tips and overall will be a perfect platform for documentation).

Stoicho just released a blurb at his “The Morfik Watch” blog about the release and he is also amused by the name (Morfik 07) and how close it is to 007: IMHO it would have been more appropriate, considering the disruptive lethalness of the new technology to the current, stone-age way of developing web applications :-)

Notable items are: the context menus, a direct-mode graphics canvas, inline popup windows, and server-sent
events.

The latter is especially interesting for load/performance considerations: fundamentally all web applications have to maintain a channel to the server to learn about events that are important for the user (think of a chat app, for example). At the moment this is done by the client (browser) periodically polling the server, which is a recurring load for the server (it happens even if you are not doing anything). The server-sent events technology promises to turn this into server-push: the clients will not have to poll, and the server will rather push the information to clients that are interested.

Opera seems to be the first to implement the technology in the 9.0 browser. It seems Opera is working hard to become the darling of web applications developers: first their rendering and editing improvements, and now this starts to make Opera a very compelling platform for web apps.

Morfik is (in my view) not one technology, but rather a synergic integration of technologies with the ultimate goal of making web application development much more efficient than it is today.

There are many specialist tools on the market, some for web UIs and AJAX, others specifically for efficient browser-server communication, then frameworks for server-side operations, database access, etc., which are very professional and high quality in what they are doing but these are all separate initiatives and integrators are very few and far between. Then there are things like Ruby on Rails, J2EE, .NET, etc. that do play the integration notes with varying success. Sometimes not enough coverage, other times big complexity or prohibitive costs inhibit application for small to medium real-world problems.

That’s why creating web applications today is a challenge: you have to master many separate technologies and make sure that they work together harmoniously that puts a great burden on the developer, and instead of focusing on the business case we have to deal with low-level technical details (which is fun in itself, just a bit frustrating when deadlines loom).

Just think of this question for a moment honestly: how many times did it happen to you that you had a great plan for an application, but as the deadline approached and show-stopper technical issues arisen you had to cut-cut-cut the really valuable functionality to deliver on time/within budget? The resulting solution is OK, but far from the vision.

Morfik right now integrates the web UI, efficient browser-server communication, server-side business logic, database access, the web server and database system itself, interactive UI editing, the build process, debugging and profiling in an integrated development environment. This is already a powerful combination: all you need is to develop first-class web applications is a machine with an operating system and the Morfik WebOS AppsBuilder. Can any other tool match this simplicity? If they make sure that the platform is open and ready to absorb additional third-party technologies with ease (as it is demonstrated with the current UI toolkit integration) and consistently, then they can embrace the OSS community (some OSS-friendly steps would be welcome:-) and commercial toolkit developers too, to augment the system with additional components so that we can have the same kind of component ecosystem that makes Windows RAD so efficient.

There is still a lot to do, but the potential is there.

“Google Web Toolkit (GWT) is a Java development framework that lets you escape the matrix of technologies that make writing AJAX applications so difficult and error prone. With GWT, you can develop and debug AJAX applications in the Java language using the Java development tools of your choice. When you deploy your application to production, the GWT compiler to translates your Java application to browser-compliant JavaScript and HTML.”

It must be a superb moment for the Morfik team to see that one of the biggest players in the software industry found Morfik’s technology very valuable and decided it was an important piece of its own AJAX software portfolio (and probably a basis of their own WebOS?).

I am really interested in how this develops!