FlexLive.net

This article is applicable to LiveCycle DS, LiveCycle DS Express, and BlazeDS.

Remote Object Service, one of the keys services in LiveCycle DS and BlazeDS, enables Flex applications make remote procedure calls to the Java server via the AMF3 protocol.

AMF3 is architected similar to SOAP, but magnitudes faster because it’s a pure binary protocol. If your Flex application loads large amounts of data, and speed/bandwidth is a priority, you should consider leveraging the Remote Object Service.

Continue Reading »

The web community has been buzzing since IE8 Beta was made public during MIX. The elephant in the room is obviously ACID2 compliance. This is a blessing for AJAX developers, but not without consequences. Under IE8, GMail no longer support GTalk, GCal is broken, and Google Reader is just plain fubar. This just shows the fragile nature of AJAX applications and the inherent flaw of browser dependence.

Ever wonder how much the CS3 package weigh? Well, wonder no more. The folks at PC Magazine have put it on a scale. Apparently the CS3 Production Premium alone weighs 17.58 pounds. That has got to be some sort of world record. I guess Adobe should stop printing CDs in solid gold.

IBM, one of the most influential players in the enterprise application space, has adopted RIA for its enterprise product lines. Personally, I believe this is a big deal because it finally validates the value of RIA in the enterprise space. When I think of IBM, the first impression comes to my mind it big and slow. It represents the early majority on the bell curve at best, and that’s why its every move carries so much clout. Just like how it brought Linux into the enterprise IT mainstream, IBM has given enterprise RIA its final nod.

Like many enterprise vendors, IBM adopted RIA not because it was hip, but it made business sense. The end users of enterprise applications are highly paid information workers. A well designed RIA can significantly reduce training cost and boost worker productivity, both of which could have direct impact on the bottom line. Businesses simply cannot ignore the benefits of enterprise RIA. With solid platforms like Flex and the proliferation of SOA, it’s now more cost effective than ever to build enterprise RIA.

Looking back, IBM’s decision to embrace Linux instead of building it’s own OS was definitely a stroke of genius. It’s highly likely that IBM will settle on a existing RIA platform than building its own. With the Flash and Flex moving towards the OSS direction, it could be the most attractive choice for IBM.

TechCrunch has just put up a new post on Apollo. Compare to the comments from the previous post, I think the people are less skepstical about the technology this time around. However, I noticed there were two reaccuring questions in the comments: How is Apollo better than traditional desktop applications and what is the security model?

So how is Apollo better than traditional desktop applications? Platform neutrality. Apollo is rooted in the Flash runtime, and Flash currently supports Linux, Mac, Windows, as well as most legacy Windows. The Apollo runtime is also extremely light weight, which means it would have a much more rapid market penetration rate compare to the mammoth Java runtime. Since Apollo supports both HTML and Flash content, it would help organizations protect and leverage their existing investments in AJAX and Flex. To sum up, Apollo is better than traditional desktop platforms because it’s platform neutral, light weight, and backward compatible with existing rich internet applications.

The security issue is something I’m interested in as well. Unfortunatly, since Apollo is still in its early stages, the security model is not solidified yet. But be sure to check the Apollo FAQ regularly for updates. The Apollo team is also very open to community feedback. You can send your feature requests to wish-apollo@adobe.com.

Update: digg away, folks!

In Ryan’s recent MAX update, he mentioned Adobe is putting up $100,000,000 in venture capital to fuel the Apollo echo system. I didn’t take it seriously until I saw this confirmation article on CNET. Although I still believe that reliance on venture capital is not healthy for startups, nevertheless, I think it would be a fun exercise to brainstorm for some cool Apollo ideas. I’m going to throw a few of mine up here. Feel free to leave a comment or follow up with your cool ideas

The first thing that came to my mind was a Picasa clone, but then I remembered that photo and video sharing has been done to death… scratch that one.

How about an instant messenger and social networking hybrid? It would look just like a standard IM client. Users would be able to chat with their friends’ friends or introduce their friends to each other…

If you are not into the “social networking” fad, maybe something more practical… what about tooling for Flex/Apollo developers? The transition from browser to the desktop means Flex has to do more heavy lifting than before, which represents many opportunities for third party vendors. Here are a few ideas,

Flex profiler: the flash player (debug) has a built in profiler, but the output is pretty raw. Maybe you can write an Apollo app that takes that profiling data and represent it using Flex charting components.

UML tool for ActionScript: most Java developers must have heard “round-trip engineering” – tools like Rational Rose let you generate UML diagrams from Java code and vice versa. It’s extremely useful for large projects. If there is an UML tool for ActionScript in the market today, I would definitely buy it.

Object layer generator: a lot people have been asking for direct SQL access from Flex. It wouldn’t be long before these Flex SQL client components pop up all over the place (I think Ted Patrick is working on a MySQL client). Why not go a step ahead and build a tool that generate ActionScript classes mapped to database tables?

Alright, it’s getting late and I think I had enough fun. I hope my solo-brain-storm session has piqued your entrepreneurial spirit. Keep in mind that these are just bits and pieces of ideas. It would take a lot more to build a successful business. But then again, if YouTube can sell for 1.65 billion, anything could happen

Ryan Stewart was blogging about how Apollo is going to change the internet application paradigm. It reminded me of the same promise made by the Java Web Start technology. The Java camp has been trying to push web application to the desktop for years with Java Web Start. Unfortunately, Web Start was plagued with poor performance and instability in the first few releases, making it unattractive to the developers. The biggest roadblock for Web Start was the fact that Sun was having a hard time getting the public to upgrade to the latest Java runtime environment. The 20mb installer and a confusing download page have proven to be too much for the casual computer users.

Do all these remind you of anything? That’s right, the Java applet technology. Java applet was the first serious attempt at rich internet application – a concept way ahead of its time, but it was hindered by the same problems. If you ask any developer familiar with Java, he or she would tell you that applet is too slow and most people don’t know how to install the JRE. Fast forward to the present, Flex is really just another attempt at the original Java applet vision. Unlike the Java applet, the Flash platform enjoys a tremendous install base and it’s very easy to upgrade to the latest version. Flex also loads faster and looks way better than its ugly cousin. It’s no wonder the Flex Builder 2 is just flying off the shelf as soon as it’s released.

Looking into the future, the Apollo shares a common vision with Java Web Start, but can Apollo overcome the challenges plaguing its predecessors? Looking at what Flex has done for rich internet applications, one can’t help being a little optimistic. Hopefully, the Apollo team can learn from the Java Web Start’s mistakes and finally capture the market.

Earlier today, a blog entry from Uneasy Silence accused the Adobe Reader for being too bloated. It was getting a lot of attention on digg due to the inflammatory nature of the post. My comment regarding this issue was getting a lot of diggs as well, so I decided to repost it here.

First of all, I agree with the fact that Adobe reader is bloated. The reason being that Adobe Reader contains many features people normally wouldn’t see (e.g. commenting, digital signature, etc.). There is a server side product called LiveCycle Reader Extensions that allows the document author to selectively activate those features. There used to be a stripped down version of the Adobe Reader, but it was dropped to avoid confusion. If there is enough demand for it, they just might bring it back.

In the meantime, if you are unhappy with the performance of your Adobe Reader, you could disable some of the plugins by following this short tutorial posted on DWTIPS.