Flash for Internet Applications: The Evolution Continues.

by Timothy Appnel

Related link: http://news.com.com/2100-1046-994258.html

Macromedia announced an enhancement to Flash that will allow its content to operate independent of the browser on the desktop. Additionally, the company will be launching Flash Central, an associated marketplace that will allow Flash application developers to market and sell their applications. Macromedia will take a 20 percent fee on all sales -- a model very similar to that being employed with mobile applications in the J2ME and i-Mode space. (For more thoughts on Macromedia Central see this entry on the weblog of Macromedia's Chief Software Architect, Kevin Lynch.

This is a very positive step in the evolution of SWF files (Flash's native format) as a platform for lightweight Internet applications or, as some call it, microcontent clients. As I asserted in my article Flash MX and The Big Picture, the document-centric browser is an awkward solution to a growing number of emerging needs. the browser still remains quite useful and will continue to do so, however it would benefit from an application-centric mate. (The topic is one of the main focuses of O'Reilly’s upcoming Emerging Technology Conference next month.) The Flash file format and runtime engine addresses these exact needs making itself an leading candidate. As I stated, In many ways Flash MX promises to do for Internet applications what Visual Basic did for Windows applications in the early '90s. It could also lay the groundwork for redefining the desktop. Based on this announcement that seems to be exactly where Macromedia is heading and I applaud their foresight.

Despite its continuing evolution, some issues have yet to be addressed to date.

Recently in response to a question posed by Kevin Werbach, former Macromedia CTO Jeremy Allaire posted to his weblog some thoughts as to why Flash applications have not been more broadly adopted by developers. As often the case, Jeremy is spot on. Awareness and cognitive dissonance (the negative association of Flash with the gaudy overuse of animation) leads his list. Efforts have been underway and will take some time and persistence to address. (Certainly issues such as the recent release of the company’s new website have not helped this cause.) Jeremy also lists Flash's ties to the browser that will be rectified in the near term. What's left outstanding is the fourth reason he lists -- the programming model or more precisely in my opinion the programming tools. With Flash's impending freedom from the browser, the final point remains as the sticking point for myself and I suspect other programmers from buying in further.

Since reading Jeremy's post and follow-up by Kevin Lynch, I've been thinking about my ideal Flash development tool -- or at least the one I think will marginalize the barriers to entry seasoned programmers perceive and really let the market flourish.

What I want is a lightweight command line tool that will let me build Flash applications from text files containing XML and ActionScript code. In other words, what I want is a Flash application compiler.

This tool would make extensive use of existing standards. ActionScript already complies with the ECMAScript standard. The XML formats would excluded standard formats such as SVG and XForms (or perhaps XUL) with the conservative introduction of extensions through XML namespaces. These XML files would be used to define things such as objects, screens, form layouts, and would be compiled into native Flash format constructs. (In my estimation this is a good compromise in addressing the uproar of standards advocates while keeping the player lightweight.)

There are several lessor-known third-party Flash development tools that address some of wishes -- tools such as Lazlo, X-Wave, Swish and Screenweaver and DENG -- though not entirely or to my satisfaction. To grossly generalize, these tools gravitate towards Flash's animation heritage amongst other shortcomings. The Lazlo Presentation Server comes closest, but is a server based.

What these tools lack most importantly is that they do not originate from Macromedia. Providing freely available baseline tool for Flash application development is crucial to its evolution and simply too important to come from anywhere else. Hopefully I’ll get my wish.

UPDATE: Seems my wish could be coming true sooner then I thought. Macromedia's Mike Chambers writes on the tantilizing Royale project that was shown a the FlashForward conference today.

What are your thoughts on Flash applications and microcontent clients?


2003-03-28 13:03:50
or just use Mozilla
which provides everything I need for a non document-centric local application. UI widgets, web services api's (http, soap), local caching, cross platform install.
2003-03-28 13:39:39
or just use Mozilla
While I understand where you are coming from[1], I don't entirely agree with your assessment. Mozilla's deployed base is currently not sufficient enough to make you proposal feasable in practice. I wish this where not true, but it is. (BTW: I use Mozilla as my primary broswer.) This is paramount to any Internet applications success.

[1] http://www.mplode.com/tima/archives/000018.html

2003-03-29 12:52:45
Positive step?
Why is any development of this worse-than-useless garbage a "positive step?"
2003-03-30 07:55:45
or just use Mozilla
I read your article "The Mozilla VM?" and it's very interesting. But it is not true you have to know C++ to write applications with Mozilla.
You can write local and/or remote application only with XUL and Javascript, using CSS for the layout (http://www.cfmentor.com/~faser/mab/).
Anyway, I don't see a future in Flash application. I's only a matter of time when MS will push web application with .net. If you want desktop application loadable from a browser why you have to use Flash, when you have .net or Java Web Start?
2003-03-31 10:54:48
or just use Mozilla
Yes, Mozilla exposes a tremendous amount of functionality to javascrpt. If you need features/performance that javascript cannot handle, then build that part in XPCOM in either c++ or python.

I think the advantage of the installed base for Flash is diminishing with the increase in access speeds. Even though Flash has such a large installed base, the barrier to getting another VM installed is so much lower than when Flash first arrived. And with XRE, Mozilla apps will be a very small download.

Actually, the future of Flash may be with Microsoft.

2003-05-08 14:04:36
flash for programmers?
Why? Are layers and timelines that scary to programmers? And besides, I think you completely miss the point. FLASH is an application to invite the non-programmer to come and create an application. It's conceptual use and conceptual user are a thoughtful match, and market.

Maybe we won't need thoroughbred programmers in the future web.

2003-05-08 16:23:58
Positive step?
Because it's more interesting then useless anonymous comments?
2003-05-08 16:24:54
flash for programmers?
> Why? Are layers and timelines that scary to programmers?

No one said they're scary anonymous. What they are is counter-intutive to how a programmer thinks and prefers to work. In otherwords, it's bad usability and bad user experience for what is a large potential audience.

> And besides, I think you completely miss the point. FLASH is an application to invite the non-programmer to come and create an application.

Actually I was thinking the same as I read your comment.

Why should Flash *only* invite non-programmers? Besides, I think you are confusing the Flash Studio tool with the SWF file format and the Flash Player. SWF file and the player are highly versistile and tool agnostic.

In a sense Flash lower the barrier of entry, but I have yet to a good visual tool the abstracts the user from having to understand logic, structure, constraints or system design. Read Joel Spolsky's Law of Leaky Abstractions fore more of what I mean: http://www.joelonsoftware.com/articles/LeakyAbstractions.html

> It's conceptual use and conceptual user are a thoughtful match, and market.

I'm not sure what you are saying here, but I don't like the sound of it. Did you work for a dot com by any chance?

> Maybe we won't need thoroughbred programmers in the future web.

Right. Maybe with Microsoft Publisher we won't need print designers and maybe with FrontPage we won't need to bother with HTML every again. And maybe if PhotoShop and Illustrator got their act together we wouldn't need designers either.

Tools change, but talented and experienced developers that can balance technical, business and user needs will always be needed. I would argue that there aren't enough of them around and that is part of the problem. Having better and more appropriate tools is the other. I'm quite optimisitic that programmers are going to be able to do great apps with Flash. I think Macromedia and others such as Laszlo do too because they are putting a lot of effort into just that.

2003-07-12 12:32:37
Laszlo Systems
I think you are selling Laszlo (http://www.laszlosystems.com) short. If you'd like to see what Royal may give you, someday, take a look at Laszlo today. We provide an XML and Javascript (not Actionscript) object-oriented language for developing web apps that have the power of desktop apps. There is a free developer evaluation kit available for download from our web site.

P T Withington
Software Architect
Laszlo Systems