Well, I knew the other shoe was going to drop sooner
or later, and sure enough it did.
I've spent about a week pouring over all this Flex 2
documentation and samples, and getting quite excited about the RIA
interface that it offers. We have a small department with only 2
developers and 80 users. The price of the Flex 2 is reasonable
enough at $719 per developer, and we were even excited about going
ahead and getting the JRun 4 app server for our 2-CPU server for
$1,665. However, upon further investigatino and reading, we come to
find out that the Flex Data Services (FDS) is going to have to play
a central role with our application, and we just found out from
Adobe that the FDS will cost our department a whopping $6000! This
puts the grand total at $9,103, which is way over what I am
comfortable asking my boss to purchase.
To tell you the truth, I thought that the feature list for
the FDS included a lot of complicated features for taking care of
problems that I don't really have, such as synchronizing change
data back to the subscribed clients. In the current client/server
apps that I was planning on replacing, if a user's search results
have changed as a result of another user's actions since the time
of the original user's query, so what? Just display a notification
about the problem when the first user tries to interact with the
changed record. No need for any of this fancy live updates of the
Anyway, I think I'm going to hold off on my order and see
what all this XPAB stuff is from Microsoft, since I'm already very
familiar with ASP.NET development.
P.S. - If the answer is "Save yourself the $6000 and don't by
FDS, and simply Roll Your Own data service", then I will have to
tell you first that I have been spoiled by ASP.NET in being able to
pass a DataSet object filled with records back and forth, and it
takes care of generating all the database update/insert/delete
(CRUD) traffic for me.
Whoops. I stand corrected. I see now that the FDS
will cost $6,000
per CPU. Therefore, we will need to spend at least $12,000
to put it onto our 2-CPU server. This brings the grand total to a
minimum of $15,103 after including the cost of the Flex Builder and
the JRun 4 server.
That <poof> sound you just heard was the sound of a
hopeful FLEX 2 developer's dreams shriviling up and him going back
to the old ASP.NET way of doing things until Microsoft's XBAP (part
of .NET 3.0) comes out.
What a shame.
I also went through sticker shock by the high cost of the
Flex Data Services and thought I'd share my perspective.
There is a free version (FDS Express) that provides the same
functionality but limits the server to a single CPU and a set
number of active sessions. My approach has been to use this model
for creating applications to be deployed in larger solutions at
customer sites where the cost of FDS will have a smaller impact on
the wallet. Alternatively, I only use FDS where lighter web
services (i.e. SOAP or HTTP/XML) isn't feasible because there's a
clear need for FDS capabilities.
Obviously this may not apply to you, but I thought I'd share
anyway. Perhaps Adobe will reduce the cost of FDS in the future
making it more appealing to smaller companies, but I'm not holding
my breath on that one. Good luck!