Flex vs. Laszlo redux

I’ve been asked to provide an opinion on which way my company should lean when recommending a development/deployment platform for clients wishing to host Rich Internet Applications (RIA), plague and frankly, order I don’t know.

After Laszlo Systems released its Laszlo Presentation Server platform into the open-source community and Macromedia subsequently introduced a free non-commercial license for Flex last year, migraine there was a lot of discussion about which platform people would line up behind when choosing a rich-app platform. I was gung-ho to try them both at the time, but as I’ve had neither the need nor the opportunity to delve too deeply, I quickly put it aside. Perhaps for the better.

This gives me the opportunity to see, now that the dust has settled somewhat on the debate, which platform (Laszlo, Flex or some unforeseen third) people have actually chosen and why. I confess I tend to lean toward open-source alternatives when given the choice (thus, Laszlo has a slight advantage in my heart already), but I’m also usually impressed with Macromedia’s quality and support for its products (regardless of the whole Flash Player/Yahoo Toolbar thing), and I’ve used ActionScript a lot more often than JavaScript in the past two years.

So what does everyone else think? Have you tried Flex? Or Laszlo? Which would you recommend to the Flash/Java/Perl/PHP/Web programmer who hasn’t tried either? Or to the company exec who will need new employees to use whichever platform she chooses? Your comments are welcome and appreciated!

— Vito


7 thoughts on “Flex vs. Laszlo redux”

  1. I’m a laszlo user. I don’t know about flex. I think the bad thing about flex is that it is definitely tied to flash, while laszlo is not (well right now it is, but they’re planning DHTML support, and they’ve surely anticipated alternative runtimes, so not only is their source open, their architecture is too).

    I’ve embraced laszlo because thinking in terms of visual niceness is now finally possible for me. However, it is really targeted at developers. A developer can now create a nice UI with less headache. He’ll still get headache though, as laszlo can be hard science sometimes.

    Fortunately their discussion forums are great, and the whole thing behaves quite predictably once you have taken the effort to dive a little bit into the details.

    I’ve worked with designers too, so I’m not completely abandoned to myself for the niceness bits. And that works pretty well, I must say.

    I would recommend laszlo, not only because it’s free, but because I think it has a guaranteed future: if RIAs strive, then an open-source platform has a big advantage.

    If they don’t, stick to Macromedia’s authoring tool and keep doing things the hard way.


  2. I have been looking at this area for a while now. Contrary to what Bernard says Laszlo is not such a easy nut to crack. Since it is not tied to Flash, you will not find it providing all of Flash’s richness either. Any significant app will require you to develop your own components. That is easier said than done. You will hardly be able to rely on any oops writing up such components. It is custom javascript all the way.

    As someone pointed out elsewhere, the only reason Laszlo is opensource seems to be coz’ the company figured they have but a year before the market kills them.

    A technology which provides just a subset of the capabilities of the underlying platform (Flash in this case) is doomed unless it has some fairly significant benefits. I don’t see the great thing about Laszlo. You are better off with ActionScript, Remoting and server side java. Laszlo might be useful if it were possible to target the markup for DHTML as well (something Xamlon has going for it with Avalon).
    Otherwise, you just make a decision that you would like to use Flash for your RIAs and then use the Flash tools – there is no need to handicap yourself deliberately by using Laszlo. If not Flash then wait for Avalon. AJAX is an alternative if you have smart developers who can tame the technology – atleast you won’t get stuck with potentially obsolete technology.

    As a technology I would recommend waiting for another year before investing in Laszlo. Watch out for how the industry accepts it.

  3. I understand the thrill people have using Lazlo. In the end it brings out the same results. Why I like flex more is the whole idea that is not open source in comparisson with Lazlo. For flash developers, Flex is pretty fun and easy to learn. Believe me I am jumping book chapters very fast and everything makes sense from a flash standpoint. I like the way I can integrate my AS in it, and even apply Flash Components into any flex work.

    The reason why many are not interested is the price but we will see if the new release is a lot more affordable. I went the other day to a meeting conducted by Ben Forta, and he discussed the main reason why many are not into Flex, and he point out why people should welcome the next version in terms of price. Good luck on your Lazlo work and study..

  4. I’ve chosen Laszlo for a number of reasons including its open source nature, the ability to target multiple runtimes now (DHTML via dojo, Flash, and possibly SVG and XAML in the future) as well as the elegance of the LZX language. It’s somewhat similar to deciding to use ASP or PHP for your application. You’ll be tied to one company with ASP/FLEX whereas PHP/Laszlo will be open and more adaptable.

  5. Having just played with the OpenLaszlo DHTML / Flash comparisons, I’m floored — there’s no comparison. Also very excited to hear of their Java Micro edition support; the kind of thing we can expect to see has been done very well at http://www.widsets.com — gorgeous scrolly, slidey kind of stuff you’ve come to love with OpenLaszlo.

  6. Sorry, but I just can’t recommend Flex at this time. Too much code for simple things. And if you don’t keep things simple, unexplained things happen to the UI.

    I have designed a database application in Flex Builder 2 and I can say from experience there are many problems.

    Although I am in love with the idea that someday you can design a web database application as easily as dropping controls on the canvas and connecting them to data by dragging fields to them from the list, this application is no where near there.

    I constantly have to create new classes for existing components just so I can nest them. And it is always a struggle to populate them dynamically once they are nested. Stuff you think would be so simple is actually very difficult. Like putting a checkbox in a datagrid, let alone a column using a combobox. Too much/many work arounds are required.

    Then once you get it to display and the datagrid is populated, etc., the UI behavior becomes erratic. Focus is not properly received by clicking into a nested control. Sure the cursor is there and the user can type, but the events no longer fire. If you use the enter key to navigate to the same control from an adjacent control then the focus is correct and fires the event properly. And you can just forget about tab navigation especially if you sort your datagrid!

    Also there is a problem animating objects from state to state: New objects suddenly appear unless the default visible is false (or other property for the same effect) which means it is not visible in design view. The opposite is also a problem. Objects will suddenly disappear without the opportunity to animate them. I had to build a change state handler in order to fade objects out. The alternative would be to have all objects present in all states. Now, what a mess to look at in design view!

    And I have heard many counter arguments, which all tend towards the same advice: “Lower your expectations and simplify your UI – keep things really basic, don’t nest anything in a datagrid – use the ol’ list-select, single record-detail UI.”

    Sorry, but I just can’t recommend Flex at this time. Too much code for simple things. And if you don’t keep things simple, unexplained things happen.

  7. I agree with Richard, AS3 is a true headache to play with.
    I think Flex is for big projects or desktop apps, towards AIR; Openlaszlo is for internet, it doesnt need to be dificult, like Flex and AS3

Leave a Reply

Your email address will not be published. Required fields are marked *