That really helps!
What i was getting at with the taglibs -- I know you presently have to
buy and install Flex to get the taglibs. What I think I would rather
do is just by the tablibs and install them on the CFMX server -- if
that makes sense -- then CFMX could do the entire app including
generating the SWF. I don't know MACR will ever offer this as an
option, but there have several comments about Blackstone that make me
think that this is true.
Why would I want to do this, I can think of several reasons:
1) quite often the Initial download of a RIA needs to be populated with
initial data... it would seem to make sense and be more efficient for
CFMX to get the initial data, manipulate it as necessary then generate
the swf with initial data already inbedded. Rather than the client
going to the Flex server,, downloading the RIA app, then that app
immediately hitting cf for the data? Seems like it would be less a
load on the server(s) and the client as well as significantly less
bandwith to ttransmit the data.
2) With part of your app on 1 server (Flex) and part on another (CFMX)
aren't you increasing your exposure to outages and creating additional
confusion when they occur? Or if you opt to code your CFMX app to be
able to run if the Flex server is down, doesn't that destroy most of
the benefits?
3) I am not sure about this one, so I'll just ask: can Flex take
advantage of the Clustering/Failover capabilities of a J2EE server like
JRun?
4) Are there really that many apps that will be Flex only and not
require services of CFMX or the like?
Dick
P.S Nice setup!
On Mar 30, 2004, at 12:11 PM, Stacy Young wrote:
> I'm not sure what ur getting at with the taglibs. The only time you
> can use the JSP taglibs inside of CFMX is when Flex has been installed
> along side. If you *have* installed alongside CFMX in the same jrun
> instance then you *can* use the taglibs if you want to embed flex
> output inside a CFM page...but a more likely scenario for most will
> probably just have the user accessing MXML pages...after which the
> flex app would communicate with CFMX via remoting or Web Service.
>
> I actually chose to physically separate CFMX and flex...mostly
> because we'll have Flex apps that won't even touch the CFMX
> instances...plus for redundancy sake I don't want to have to take down
> ½ the available flex UI resources just to cycle a CFMX instance for
> whatever reason.
>
> Here's our actual setup...
>
> 2 x 2CPU machine
>
> JRun 4 with CFMX 6.1
>
> Total of 3 Jrun instances running on each machine.
>
> Two have CFMX, one has Flex.
>
> Each CFMX instance belongs to a cluster. Cluster #1 consists of
> one instance on server A, on instance on server B. Same for Cluster #2
>
> When the user access a Flex application it loads from the standalone
> 3rd JRun instance. The Flex app now pulls data from multiple
> applications running on either of those two CFMX clusters. In
> addition, there is a whole other WebLogic cluster it ties into for
> specific services.
>
> I hope I'm not clouding things further!
>
> Cheers,
>
> Stace
>
> _____
>
> From: Dick Applebaum [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, March 30, 2004 2:21 PM
> To: CF-Talk
> Subject: Re: Flex is out
>
> On Mar 30, 2004, at 11:13 AM, Stacy Young wrote:
>
> > I can tell you from experience...RIA development for most typical
> j2ee
> > development teams will be of a fraction of the cost vs traditional
> > flash/remoting development. The difference in managing source code
> > alone
> > is worth big bucks to me...I've had my FLAs over-written a few too
> > many
> > times ;-)
> >
> > Stace
> >
> >
>
> Now that seems to be a very valid justification.
>
> Do you think that adding CFMX to that equation adds or detracts?
>
> If CFMX is a plus, would CFMX & the taglibs, alone provide the
> justification?
>
> I don't know, and am assuming that there is a lot more to Flex than
> the
> taglibs... that the taglibs provide rhe UI components and generate the
> swf from XML.
>
> TIA
>
> Dick
>
> _____
>
[Todays Threads]
[This Message]
[Subscription]
[Fast Unsubscribe]
[User Settings]
- Re: Flex is out Danielle Romain
- Re: Flex is out Thomas Chiverton
- Re: Flex is out Dave Carabetta
- Re: Flex is out Dave Carabetta
- Re: Flex is out Jeffry Houser
- Re: Flex is out Thomas Chiverton
- RE: Flex is out Earl, George
- RE: Flex is out Stacy Young
- Re: Flex is out Dick Applebaum
- RE: Flex is out Stacy Young
- RE: Flex is out Dick Applebaum
- RE: Flex is out Stacy Young
- Re: Flex is out Dick Applebaum
- RE: Flex is out Stacy Young
- RE: Flex is out Stacy Young
- RE: Flex is out Rob
- RE: Flex is out Rob
- Re: Flex is out Dick Applebaum
- Re: Flex is out Dick Applebaum
- OT: CSS and "Cloaking" Search Engi... Les Mizzell
- RE: CSS and "Cloaking" Search ... Sandy Clark