Is the Proprietary Nature of the Flash Player Keeping You From Using Flex?

It's relative dependency on a good IDE like FlexBuilder. It's like, if the Web depended on DreamWeaver, it would be at most 1% of what it has become. :-)

The application packaging is very unnatural when compared to the Web. Even with dynamic linking of code at runtime it just doesn't fit right. Imagine if JavaScript could be used in Flex, and the JavaScript could be downloaded just like it normally is with the Web browser. Compiling and deploying Flex applications is just to its own detriment when it comes to universal adoption.

The programming model is not competitive enough. As it is, the false sense of security by using Flex can be dangerous even. By opening the server back end too much applications developers could cause problems down the hill.

1) you don't need to use FlexBuilder, you can compile everything from the command line for free.

2) JavaScript can be called by Flex applications. For instance, in order to "mesh" Google maps, you create a web page with your Googlemap on it and you create your rich application in Flex which calls the JavaScript to control the Googlemap.

3) What do you mean by false sense of security? Flex and LiveCycleDS has a security mechanism. When you create a RemoteObject on the server you can specify that a Flex session has to have particular security credentials. The flex "session" can only be manipulated on the server so Flex apps are not able to "spoof" security. Of course, if you're not using LiveCycleDS (or BlazeDS for that matter) then you are at the mercy of whatever back end framework you have chosen to use.

Flex's apparently shallow learning curve intices developers like me only to reveal something much more deep than is immediately apparent. I'm sure if you looked deeper in to the documentation or went on a training course you would see the power and security of Flex applications.

Also remember, that until recently Java wasn't "open" either, that didn't stop people using it.

Cheers,
Chris

... the only thing that persuades them to use a technology is a critical mass of peers also using it - in other words reluctance to invest learning time in something that becomes defunct. The technology will only get better through use, so any arguments regarding familiarity of Run Time / Compilation / Dev Environment are irrelevant.

Following this idea through to its logical conclusion, Open Source or Proprietary, the only thing that really matters is whether or not the technology is achieving what the market demands.

It's so hilarious how some people tell us that it's open since the build tools are open, as if someone was really simple enough to think that you could really do Flex projects without buying the proprietary tools for which there is only one major vendor.

The market has progressed past the point where technology lock-in goes unnoticed and is tolerated.

Fine. I'm going to hope for something like Flex but that works like I prefer to come up on the scene in the future. Who knows where it's going to come from? As being proprietary is not a problem necessarily, it could come from Microsoft or Sun again. Or maybe from a little startup in the sea of startups out there. As far as replacing the browsers, it's difficult to imagine that. But I figure something simple and efficient could come from anywhere, even from Adobe.

Partly, yes it does. But things seem to be going in the right direction, so I am not so worried now. For the mainstream, I think people think its open enough, so what I think probably doesn't matter.

Who cares if it is open sourced! To think why so many people are using .Net and are ready to use Silverlight, you'll know your questions are so naive. Most developers don't care if a technology is open source or not, otherwise, M$ would not be so lucky to have so many developers using their Visual Studio, .NET, Silverlight, and applaused for them.

Also remember, that until recently Java wasn't "open" either, that didn't stop people using it.

I don't want to develop Flex applications because I don't like using flex. Every time I come across a site that uses flex (which is luckily rare), I just groan in disappointment. It just doesn't work the way the web works. Often it's difficult or impossible to copy and paste information from the application. Bookmarking (while possible) is a pain, and is therefore rarely implemented. It's slow, with long load times between clicks. Maybe these are developer issues -- but it's so common that I have to put some blame on Flex. Obviously idiomatic Flex development does not yield usable high performing applications... Kinda like how applets were always ugly. ;-)

Hi Clinton,

Can you post some of the sites you feel are good examples of what you are saying? Personally I find experiences like Buzzword and Picnik to be much better than their non-Flex competitors.

-James

Also remember, that until recently Java wasn't "open" either, that didn't stop people using it.

I've created both swing, and flash apps, and in both cases, I've been happy with the results, but both of them were for internal applications. In each case I asked if there were any thoughts of turning them (or some aspect of them) public. In each case I got a very flat no. The reason - "I don't want the client to be required to have 'x'".

I don't want to develop Flex applications because I don't like using flex. Every time I come across a site that uses flex (which is luckily rare), I just groan in disappointment. It just doesn't work the way the web works. Often it's difficult or impossible to copy and paste information from the application. Bookmarking (while possible) is a pain, and is therefore rarely implemented. It's slow, with long load times between clicks. Maybe these are developer issues -- but it's so common that I have to put some blame on Flex. Obviously idiomatic Flex development does not yield usable high performing applications... Kinda like how applets were always ugly. ;-)

My experience is completely the opposite of everything you said in your post (except the book marking issues).

I actually don't really disagree with you on that point. I was trying to state why I think there will be some trouble with adoption and often that choice is not left up to developers. Also - Flex and Flash do need to be separated a little bit. Many people don't have the latest plugin, and that is required for Flex apps. For enterprise applications, this may be a non-issue. Additionally, many of my other points are still valid, and are reasons why I personally avoid Flex for public applications.

In the end I think it depends on the contract you engage in with the user. Is the user in a controlled environment? Do they have to pay for the app? Does the app allow them to do something they can't get anywhere else? I think if user wants video, they need to be willing to have a plugin. And right now, flash is definitely the best plugin for that job. If its a more data-centric app, maybe Flash/Flex isn't the right tool.

If you haven't already done so, take a look at Appcelerator (www.appcelerator.org). It is an open extensible platform for developing RIAs that can communicate with all major back end languages (java, ruby, PHP, .Net and even perl). It is standards based, HTML, CSS and Javascript, but you do not have to write JavaScript. The product is licensed under the GPL and involves some interesting people like me:) Disclaimer, I work for appcelerator.

1) you don't need to use FlexBuilder, you can compile everything from the command line for free.

I don't think being proprietary is a problem. The real question is: does Flex offer enough value to developers to get them to learn MXML and ActionScript. For me, the answer is no. Here's why:

1) MXML and ActionScript don't provide me with anything beyond what I can do with standard HTML, CSS and Javascript.
2) I can leverage Flash without Flex
3) The communities around Javascript are more widespread and active than those around Flex (e.g., ExtJS, YUI, Prototype, JQuery, Scriptaculous, and many more)

One benefit of Flex is that it offers all of the building blocks required to build RIAs in one fully-integrated solution. It gives developers:
- A set of pre-built widgets
- The ability to consume services written in any programming languages
- The ability to control the state of the user interface (e.g., animation, hide/show, enabling/disabling, etc.)

At Appcelerator, we built an open source RIA/SOA platform that is fully-integrated, like Flex, but based on open standards (e.g., HTML, CSS and Javascript for RIA development + Java/Ruby/PHP/.NET/Python and Perl for service creation). A standards approach has some significant benefits:

* Third-party libraries and toolkits can be easily integrated (e.g., we have integrated widgets from YUI and ExtJS)
* Developers can use skills they already have = low barrier to entry = easier to build a large community of developers

Flex is certainly a viable option for building great RIAs, but I think the approach that will ultimately get the most traction is one that is fully integrated (like Flex), but based on open standards.

Its not just the proprietary Flash player that keeps me from using Flex, but also their proprietary back-end services. Didn't until very recently they charged a ridiculous amount of money to use their back-end services? Yes, they just open-sourced it, but it only works with Java. That rules out the many, many people who deploy successful web applications using PHP, Ruby, and even .NET.

Hi Nolan,

I do think that Appcelerator is pretty cool but targeted at a different type of applications than Flex. I have a few questions / comments about your reasons for not using Flex:

1) MXML and ActionScript don't provide me with anything beyond what I can do with standard HTML, CSS and Javascript.

2) I can leverage Flash without Flex

3) The communities around Javascript are more widespread and active than those around Flex (e.g., ExtJS, YUI, Prototype, JQuery, Scriptaculous, and many more)

That rules out the many, many people who deploy successful web applications using PHP, Ruby, and even .NET.

Hi James,

We do use Flash for our charting widget - but that's not Flex. The fact that we use Flash for our charting capabilities really underscores my second point - i.e., I can leverage Flash without Flex, and I agree with your statement - that is a very good thing. Flash is great for doing graphically intensive tasks like charting or drawing vector graphics, etc. We will continue to use Flash in cases where it makes sense.

The Flex community numbers are great. My point was not to say that Flex does not have a large and active community, it was simply to say that the communities around Javascript are larger. For example, according the ExtJS Forums, they have over 23,000 members - and that's just one JS-based project.

I do think that Flex's fully-integrated approach to building RIAs is powerful. One of the problems in the HTML/CSS/JS world is that you have to piece together different parts of the RIA puzzle to build an application, which gets to what we're trying to do - provide a fully-integrated, standards-based solution that makes it easy to build RIAs

I would love to talk with you off-line about any ideas you have regarding how Flex and Appcelerator could be combined. Feel free to shoot me an email - nwright@appcelerator.com

Nolan

good

I cannot help anyone who refuses using the - in my opinion - great Flex Builder IDE. I think Adobe did the right choice to use the Eclipse Platform for the Flex IDE. I'm also using Eclipse (WTP) for JavaScript and XHTML/CSS, and maybe therefore I'm a bit biased. ButI can not see any limitations that the Flex IDE has in comparison to any other simple code editor or development environment.
Your comparison to Dreamweaver is quite far fetched, or haven't you realized, that there is a "source" tab in the Flex Builder, when editing MXML files?