Are you building RIA's?

19 Aug 2004 - 4:53pm
10 years ago
1 reply
587 reads
Tony Lopez
2004

I'm in search of individuals and studios that have been designing and/or
building RIA's. I'm particularly interested in speaking to parties who have
been working with Macromedia Flash and Flex.

I'm in the process of building the user experience discipline for Macromedia
Professional Services and I want to establish outsourcing relationships with
top talent. Please contact me directly if you think that you or your company
fits the bill.

Thanks.

Tony Lopez | User Experience | Macromedia Consulting

Comments

20 Aug 2004 - 12:27pm
Clay Newton
2004

> I'm in search of individuals and studios that have been designing and/or
> building RIA's.

Hi there,

Here at my company we just completed a pretty significant technology
review for rich application platforms. Obviously, Flex was one of the
candidates, and in many ways was one of the drivers for the
realization that we could start building web deployed rich apps. A bit
of perspective here, I am the manager of my company's User Experience
group. My group is tasked w/ design and development of the user
interface, so the technology choice really affects me directly in that
it sets the conventions to be followed, adds constraints to the
interaction model, and is something I have to spend every day looking
at from a development/maintenance standpoint.

The final contenders in our analysis included Flex
[http://macromedia.com/software/flex/?promoid=home_prod_flex_111703],
Laszlo [http://www.laszlosystems.com/], DHTML, and ULC
[http://www.canoo.com/ulc/], though there are a number of other
platforms that we reviewed initially. We evaluated each of the
platforms against a number of technological and business requirements;
the tech requirements included:

Robust native widget library
robust event binding
Maintainability
Section 508 Compliance
Support of our existing Framework practices
Ease of development for application developers
Performance : Start-up : thick client
Performance : Start-up : thin client
Performance : Render
Performance : Large data-set
Deployment : server
Deployment : thick client
Deployment : thin client
Price Impact
Availability of 3rd party extensions
Price of 3rd party extensions
Long term vendor viablity
Industry Rich Client direction
Mockup / Demo Impact
Coverage range of UI Deployments
QA integration

In the end, the differences between Flex and ULC were really
statistically negligible. The decision really boiled down to two
factors: the cost of the solution and what I call the Java-factor.

The cost issue is an easy one: Flex has a runtime licence of $12K
versus a developer licence for ULC of about $1500. OEM exists for
Flex, but the final cost of the platform would be a much greater
percentage of our product price than preferable.

The Java factor, which to me is the more prescient issue, comes down
to the relative maturity and institutional acceptance of Java as a
standard. Flex uses a mixture of standards-based programming
languages, MXML and ActionScript 2.0. ULC is built purely in Java. In
the end, the relative stability and _openness_ of Java as an
implementation platform was perceived to be preferable to the nascent
MXML and ActionScript standards.

We are currently in the midst of building out the framework for our
rich client interface. While personally I initially had the gut
reaction that Flex was the right way to go for us, as I move forward
with ULC I am happily surprised with the flexibility and elegance of
the platform.

For anybody on the verge of making a decision about building rich
interfaces I would suggest that you try to divorce emotion from this
decision. In most development shops, someone will have a very hard-set
opinion about what is "right." Do your best to be objective. The
method I used in coming up w/ the rankings to the criteria above I
found to be quite effective. I filled in the ranking for each of the
above line items using squares (1, 4, 9); ranked the relative
importance of each of the metrics for our organization; polled all of
the major stakeholders in engineering; averaged out everyone's
answers. I multiplied the importance by the score for each metric,
added up the scores, and then compared that to a perfect score to come
up w/ each technologies score. The beautiful thing about following a
methodology such as the above is that it quantifies the qualitative
analysis.

It is going to be really interesting to watch the landscape evolve
over the next 4 years or so. Longhorn's debut (whenever that actually
happens!) will add some interesting complexity to the mix, but for the
time being there are definitely some solid solutions out there.

-Clay

Syndicate content Get the feed