Discussions

News: Flex with Java: POJOs, Spring, Hibernate, and JMS

  1. Flex with Java: POJOs, Spring, Hibernate, and JMS (12 messages)

    Christophe Coenraet's new Flex Test Drive Server includes samples and tutorials that will show you how to create Rich Internet Applications for back ends such as Spring, Hibernate, ActiveMQ (for JMS integration sample), and POJOs. This is a great way to quickly get up to speed on Flex. The Test Drive Server includes: - A 30 minute Flex Test Drive for Java developers, including Flex Remoting, Web Services, real time and collaboration samples - Flex/Spring integration samples - Flex/Hibernate integration samples - Flex/JMS integration samples (JMS powered by ActiveMQ) - A Flex Data Management Services tutorial - A Real Time Market Data sample application - Other samples

    Threaded Messages (12)

  2. I'd love to buy it[ Go to top ]

    But... I can't afford the data server for $20K per server for each of my clients. .V
  3. Re: I'd love to buy it[ Go to top ]

    You don't have to spend a penny. The framework is free. The SDK is free. The compiler is free. If you choose to use the server (which you don't have to use) portion it as well can be free. Any other FUD? Dave Wolf Cynergy Systems
  4. Re: I'd love to buy it[ Go to top ]

    I think that Flex has great potential, and I've seen a great deal of FUD about it. To some extend the solution is proprietary, although some of the arguments for not using it is simply wrong. Check out: http://www.onflex.org/ted/2006/12/top-10-myths-about-adobe-flex-20.php
  5. Re: I'd love to buy it[ Go to top ]

    The article is all about using the data server, which, as the OP stated, is $20k per CPU. I also don't feel right giving that price to my clients. The OP was not spreading FUD if you read his statement ("for the data server") or if you consider that the article is all about the $20k/CPU piece of the flex stack.
  6. Re: I'd love to buy it[ Go to top ]

    If you don't user Data Service which provides ActionScript Java remoting capabilities Flex is free. Flex provides support for WebServices out of the box, ie no cost, and you can do an awful lot of stuff with its great XML support. We have build a great dashboard application without using Data Services. Adobe also have a range of licenses: If you need Flex Data Services, you can get an Departmental license for about 100 concurrent users for about $6K per CPU: http://weblogs.macromedia.com/pent/archives/2006/06/flex_20_now_ava.cfm regards Malcolm Edgar www.avoka.com http://click.sourceforge.net
  7. Re: I'd love to buy it[ Go to top ]

    I understand and love that the base package of Flex is free (if you don't use flex builder). It's a much better option than earlier versions provided. And the price for flexbuilder isn't prohibitive. But, as stated, this article was about specifically using the data services, as are many that I see out there and on Adobe's website. I understand why the data services are promoted like they are. They seem to be pretty useful and of course Adobe is entitled to make some money on this. But their prices are just way too expensive. $6000/CPU for a departmental application for 100 people? For a typical rack-mounted server with 2 processors (which is fast becoming the baseline), that's $12000. $12,000 for a departmental application. Not very realistic. And at $20,000/CPU for the default version, that's $40,000 just for the data services alone. That's absolutely ridiculous. You can spend that kind of money and get WebLogic with full support for 2 CPUs AND the physical server itself. That kind of money for a data enabling service is just way out of proportion to the feature set provided. If adobe were to charge $500-$1000 per SERVER (not CPU) for the departmental license (why charge per CPU for a departmental server?) and something more like $5-$8K/CPU for the enterprise, it would be much easier to justify. I can't even imagine trying to describe the advantages of Flex Data Server to a non-technical client at $20k/CPU. They'd never get past the cost unless the application absolutely 100% couldn't be "flex-like" without it. Again, though, this isn't FUD about Flex as a whole, which is a great product, but about the licensing for the data services. It's a scary cost that could prevent Adobe from gaining market share. At a reasonable cost, they could destroy their RIA competitors in both mindshare and sales.
  8. Re: I'd love to buy it[ Go to top ]

    I understand and love that the base package of Flex is free (if you don't use flex builder). It's a much better option than earlier versions provided. And the price for flexbuilder isn't prohibitive. But, as stated, this article was about specifically using the data services, as are many that I see out there and on Adobe's website.

    I understand why the data services are promoted like they are. They seem to be pretty useful and of course Adobe is entitled to make some money on this. But their prices are just way too expensive. $6000/CPU for a departmental application for 100 people? For a typical rack-mounted server with 2 processors (which is fast becoming the baseline), that's $12000. $12,000 for a departmental application. Not very realistic. And at $20,000/CPU for the default version, that's $40,000 just for the data services alone. That's absolutely ridiculous. You can spend that kind of money and get WebLogic with full support for 2 CPUs AND the physical server itself. That kind of money for a data enabling service is just way out of proportion to the feature set provided.

    If adobe were to charge $500-$1000 per SERVER (not CPU) for the departmental license (why charge per CPU for a departmental server?) and something more like $5-$8K/CPU for the enterprise, it would be much easier to justify. I can't even imagine trying to describe the advantages of Flex Data Server to a non-technical client at $20k/CPU. They'd never get past the cost unless the application absolutely 100% couldn't be "flex-like" without it.

    Again, though, this isn't FUD about Flex as a whole, which is a great product, but about the licensing for the data services. It's a scary cost that could prevent Adobe from gaining market share. At a reasonable cost, they could destroy their RIA competitors in both mindshare and sales.
    Man, seriously. For a profitable organization, $20K per CPU is hard to justify? Maybe some garage startup would have a problem, in which case they have the choice to either not purchase the extra services and/or use Open Laszlo. They spend that much a month on a few developers.
  9. Re: I'd love to buy it[ Go to top ]

    Yes, for a profitable organization, $20K/CPU is too much to justify. When a product that makes things slightly easier costs more than your application server per CPU, that's too much. While they may spend that much per month on a few developers, why should they also have to spend it on something else that doesn't seem to provide enough utility to justify it? Don't the developers already cost enough? We are talking _profitable_ here, aren't we?
  10. Re: I'd love to buy it[ Go to top ]

    Some funcionality offered by FDS, like shared objects, is supported by red5 for free. http://osflash.org/red5
  11. Re: I'd love to buy it[ Go to top ]

    There's no doubt that FDS is a major breakthrough for Flash and thereby Adobe. However 20k/CPU for an enterprise license will restrict adoption. The person who said that 20k/CPU doesn't mean much for a profitable corporation has obviously never been responsible for profitability and can never have made any real purchasing decisions or controlled pricing and budget. That's a huge price tag to carry around. I work in a business where we sell software supported consulting. 20k/CPU comes directly out of our margins, or we have to pass on that cost to our customer and bump our price up. Either way it is a serious problem. I am very unlikely to be able to sell a 20k/CPU proposition to our board which means I have to exclude FDS from my architecture on the grounds of margin, not just cost. The cost of developing an alternative would almost certainly be a lot less. Adobe have definitely done the right thing by giving the express version away for free but if it wants adoption of Flash and Flex to the commercial ISV sector it is going to have to look again at its pricing. I for one hope they do.
  12. Problems with Flex Builder 2.0[ Go to top ]

    My first experiences with Flex Builder 2.0 reveals two problems, one of them a major pain. The Flex Builder 2.0 plugin for Eclipse leaks memory like a sieve. I've allocated 750 MB for Eclipse and get an out of memory error about every day and a half. Eclipse has never given me out of memory errors before installing the Flex Builder 2.0 plugin. The major issue with the Flex Builder Eclipse plugin is that it expects your Eclipse project to be a Flex project. If you add Flex 2.0 to an existing Java project you'll have to refactor the project to put the Flex stuff at the root of the project or you won't get the benefits of Flex Builder. The hack work around is to create a seperate Eclipse/Flex project that contains the subset of your Flex files from the source repository at the project root. You'll have to work both in a Flex project and a Java project. Once you've learned how to set up both a Flex and Java project Flex Builder works together with the Eclipse Java support pretty smoothly. You can step through the Flex code with the Flex debugger, step into the Java code in the Java debugger, and then return back to the Flex code in the Flex debugger. I've used other plugins, like Exadel Studio Pro, that play nicely with Eclipse's Java projects without demanding that they take over everything. Flex Builder needs to learn that it can't demand to take over your project.
  13. I think, if the price is a problem, to look at the openLaszlo project. It offers an interesting alternative and it is free. I've done an important project with Flex a month ago (Spring1.2,hibernate3,Java5,...), the mean problem is to manage an MVC architecture with re-used of components. When you switch from a struts vision to a Flex one, the difference is enormous and the time spent to have the same functionnal result is very very different ! It's obvious that Flex is not choosen for its facilities but for its visual result. In my opinion, to have a great visual result it's THE technology to adopt. Nevertheless, to have only a good visual result, the time spent is too important : dhtml is better.