Tuesday, March 30, 2004
Flex is cheap as chips
Flex is not so much about rapid Flash development (although this is one of the advantages) as it is about collaborative Flash development. Flash IDE, the FLA, and everything about it is the bane of largescale development -- it sucks. This is why Macromedia Flex had to be built.
Without a very sophisticated knowledge of Flash development, it is all but impossible for a development team to put together a large scale Flash architecture that is not bottle-necked by key developers and FLA binaries. If it were a piece of cake to externalise everything as actionscript files and work collaboratively in Flash everyone would be doing it.
Few projects have the budget to resource sufficient prototyping and scoping for an appropriate RIA application framework. Fewer projects have escaped unscathed from the experience.
Then as we move into the longer term cycle of enterprise development, large Flash applications typically leave you with all kinds of maintenance difficulties. Not to mention a colourful mix of binaries and texts to version.
Rest assured that no amount of training is going to give developers an insight into large scale RIA development using the Flash IDE -- its a mystical art learnt only through painful experience. Of course its possible, and many teams are successful, but at what cost??
This is why Flex is a compelling product for large scale RIA development. And a very cost effective solution for the right portfolio of projects. Flex is really not all that expensive in comparison to other products in the space Macromedia is shooting for. Furthermore, Flex represents good ROI when applied to a project of relevant scope.
The more pointed question might be: is there a place at all for these sorts of large scale RIA apps? And is the only reason we don't see more of them because the tools to develop them haven't been around?
It will be very interesting to see the sorts of Flash apps people can build given the ability to apply more traditional development processes to what has historically been a very eclectic development environment.
Posted by modius at 07:42 PM | Permalink
Trackback: http://blog.daemon.com.au/cgi-bin/dmblog/mt-tb.cgi/227
Geoff, you're spot on mate. Flex is Enterprise, enterprise is not the small dev teams which have been repsonible for the bulk of Flash RIA's to date (I know it seems obvious, but from all the rants about Flex and pricing recently, well maybe it isn't so obvious?).
The questions Geoff raises are the more pointed questions: these are the questions that will determine if Flex lives or dies - not the price.
Posted by: Scott Mebberson on March 30, 2004 09:00 PM
"Flash IDE, the FLA, and everything about it is the bane of largescale development -- it sucks. This is why Macromedia Flex had to be built."
I would like to say it differently:
Flex consists of two parts: the MXML language which lets developers create stunning RIA without all the Flash hassle and a server that is just a very expensive on-the fly SWF compiler. Most businesses/developers do not need the server feature. All they want is a new tool, an alternative to Flash Pro, that let them create RIA's in no time. MXML is the future of rapid ria development. My conclusion: Macromedia should provide developers with a MXML-based RIA development tool that could be used to develop traditional SWF-based applications as well as SWF-applications generated by Flex. This is what Brady should be instead of a narrow-minded Flex oriented tool
Posted by: Vinny Timmermans on March 30, 2004 09:43 PM
What I find spot on in this entry is this:
"The more pointed question might be: is there a place at all for these sorts of large scale RIA apps? And is the only reason we don't see more of them because the tools to develop them haven't been around?"
I think it's a question of maturity in terms of hardware and software. Friends I've shown Flex apps either said "I dont have flashplayer installed" or felt that the apps were slow. Also, Macromedia claims that many users of a e-commerce app are lost before a transaction is made because of the bad "experience" of HTML, but how many will be lost in a Flex app due to hardware or software restrictions?
I believe Flex has it's place and this will be more true with better hardware, and a faster flashplayer(?) It sure will be interesting to follow the evolution of Flex and the RIA market.
Posted by: Nisse Bryngfors on March 30, 2004 09:46 PM
Just because you can't see something doesn't mean it doesn't exist (3 negatives...oops, lemme try again).
I have built a few RIA's, but most are for private use. Therefore, they do not get public approval and viewing, and therefore do not help facilitate others to want an RIA. The perception appears to be that RIA's need to be on a public website; if they aren't, then there must not be any RIA's. That is not the case.
My company, for instance, could definately utilize Flex in that our software is installed at hospitals (not their public website, but inside of them). Our development team could probably work a lot better (both the 5 developers here and the 10 or more in India) with a .NET version and SourceSafe.
One of the bottle necks in our last project was me. If we could fully utilize all of our developers, we'd get true RAD development, and I believe, at the time, Flex could have possibly helped us do that. Hopefully in the future, I'll be on a Flex team as the Flash support as the teams I work with are not Flash developers, and don't intend to be. Flex, however, is right up their alley in terms of how they work; and condsidering there are at least 30+ people here (including India) that I believe have the ability to use Flex with minimal training, it'd make a lot more sense to upper management when utilizing existing resources to take on new RIA projects.
Posted by: JesterXL on March 31, 2004 12:20 AM
Agreed, Flex will have it's place on enterprise RIA development, but as any other new technology, specially on the server side, it will take some time before it reaches significantly beyond the current early adopters.
One of the reasons I didn't apply for the beta is because I lack that "Enterprise applications development" experience they were asking for in the applications form and the adequate development environment, which doesn't mean I didn't have a lot of interest on "Royale" when it was first announced.
However, today is clear that programming for Flex isn't rocket science, and although it's existence is due to distributed application developers, it is probably easier for Flash IDE users to learn Flex, than for hardcore developers to deal with the Timeline. Let's don't forget that Flex only generates RIAs based on Flash content already existent in the form of SWC files on the server.
I agree that $12k might not be too much, in the long-term, for what Flex has to offer, and the new doors that it will open to mainstream RIAs development. It certainly isn't the tool for small multimedia shops catering small-medium businesses, at least not now. I guess some time Flex would be incorporated in hosting plans, just like FlashCom and then we'll see what the 'little guy' can do with it.
Posted by: Oscar Trelles on March 31, 2004 01:55 AM
The problem is that it is never 12k for enterprise deployment. When was the last time you deployed an application in an enterprise on 1 server with no more than 2 CPUs?
Posted by: Calvin Ward on March 31, 2004 04:29 AM
"then we'll see what the 'little guy' can do with it."
Actually, the little guy would like to see what he can do with it already, but can't because there's a massive price tag attached to it.
It's all relative.. Those in enterprise don't see the price as a problem - the rest of the world does. I'm not in enterprise development, but Flex is still relevant to me - I just can't afford it. I hate being excluded.
Posted by: Cliff Rowley on March 31, 2004 04:29 AM
Personally, I think FLEX is more than just an Enterprise only product. Its got more capabilities then just the Enterprise side of things, Macromedia have shown in the past that they have assumed one thing only to find out its got more capabilities than they originally thought - hence CENTRAL dilemma.
I would bank my house on the fact that if FLEX were to be accessible to the current *existing* market, you would see some very radical and more robust RIA applications vs some basic mini applications. You would also see a more vibrant FLEX community aswell.
I dunno, long term FLEX looks very good in price model, but not many people think past yearly budgets :) and when you get a 50K funding proposal together for a product, you better have answers come the following financial year/qtr and how the product is reducing costs.
Its all about the initial startup sale, once you get past that, your on the FLEX honeymoon nothing but fun :) - the trick is to convince management its a long term investment.
Posted by: Scott Barnes on March 31, 2004 02:22 PM
Geoff, you're spot on mate. Flex is Enterprise, enterprise is not the small dev teams which have been repsonible for the bulk of Flash RIA's to date (I know it seems obvious, but from all the rants about Flex and pricing recently, well maybe it isn't so obvious?).
The questions Geoff raises are the more pointed questions: these are the questions that will determine if Flex lives or dies - not the price.
Posted by: Scott Mebberson on March 30, 2004 09:00 PM
"Flash IDE, the FLA, and everything about it is the bane of largescale development -- it sucks. This is why Macromedia Flex had to be built."
I would like to say it differently:
Flex consists of two parts: the MXML language which lets developers create stunning RIA without all the Flash hassle and a server that is just a very expensive on-the fly SWF compiler. Most businesses/developers do not need the server feature. All they want is a new tool, an alternative to Flash Pro, that let them create RIA's in no time. MXML is the future of rapid ria development. My conclusion: Macromedia should provide developers with a MXML-based RIA development tool that could be used to develop traditional SWF-based applications as well as SWF-applications generated by Flex. This is what Brady should be instead of a narrow-minded Flex oriented tool
Posted by: Vinny Timmermans on March 30, 2004 09:43 PM
What I find spot on in this entry is this:
"The more pointed question might be: is there a place at all for these sorts of large scale RIA apps? And is the only reason we don't see more of them because the tools to develop them haven't been around?"
I think it's a question of maturity in terms of hardware and software. Friends I've shown Flex apps either said "I dont have flashplayer installed" or felt that the apps were slow. Also, Macromedia claims that many users of a e-commerce app are lost before a transaction is made because of the bad "experience" of HTML, but how many will be lost in a Flex app due to hardware or software restrictions?
I believe Flex has it's place and this will be more true with better hardware, and a faster flashplayer(?) It sure will be interesting to follow the evolution of Flex and the RIA market.
Posted by: Nisse Bryngfors on March 30, 2004 09:46 PM
Just because you can't see something doesn't mean it doesn't exist (3 negatives...oops, lemme try again).
I have built a few RIA's, but most are for private use. Therefore, they do not get public approval and viewing, and therefore do not help facilitate others to want an RIA. The perception appears to be that RIA's need to be on a public website; if they aren't, then there must not be any RIA's. That is not the case.
My company, for instance, could definately utilize Flex in that our software is installed at hospitals (not their public website, but inside of them). Our development team could probably work a lot better (both the 5 developers here and the 10 or more in India) with a .NET version and SourceSafe.
One of the bottle necks in our last project was me. If we could fully utilize all of our developers, we'd get true RAD development, and I believe, at the time, Flex could have possibly helped us do that. Hopefully in the future, I'll be on a Flex team as the Flash support as the teams I work with are not Flash developers, and don't intend to be. Flex, however, is right up their alley in terms of how they work; and condsidering there are at least 30+ people here (including India) that I believe have the ability to use Flex with minimal training, it'd make a lot more sense to upper management when utilizing existing resources to take on new RIA projects.
Posted by: JesterXL on March 31, 2004 12:20 AM
Agreed, Flex will have it's place on enterprise RIA development, but as any other new technology, specially on the server side, it will take some time before it reaches significantly beyond the current early adopters.
One of the reasons I didn't apply for the beta is because I lack that "Enterprise applications development" experience they were asking for in the applications form and the adequate development environment, which doesn't mean I didn't have a lot of interest on "Royale" when it was first announced.
However, today is clear that programming for Flex isn't rocket science, and although it's existence is due to distributed application developers, it is probably easier for Flash IDE users to learn Flex, than for hardcore developers to deal with the Timeline. Let's don't forget that Flex only generates RIAs based on Flash content already existent in the form of SWC files on the server.
I agree that $12k might not be too much, in the long-term, for what Flex has to offer, and the new doors that it will open to mainstream RIAs development. It certainly isn't the tool for small multimedia shops catering small-medium businesses, at least not now. I guess some time Flex would be incorporated in hosting plans, just like FlashCom and then we'll see what the 'little guy' can do with it.
Posted by: Oscar Trelles on March 31, 2004 01:55 AM
The problem is that it is never 12k for enterprise deployment. When was the last time you deployed an application in an enterprise on 1 server with no more than 2 CPUs?
Posted by: Calvin Ward on March 31, 2004 04:29 AM
"then we'll see what the 'little guy' can do with it."
Actually, the little guy would like to see what he can do with it already, but can't because there's a massive price tag attached to it.
It's all relative.. Those in enterprise don't see the price as a problem - the rest of the world does. I'm not in enterprise development, but Flex is still relevant to me - I just can't afford it. I hate being excluded.
Posted by: Cliff Rowley on March 31, 2004 04:29 AM
Personally, I think FLEX is more than just an Enterprise only product. Its got more capabilities then just the Enterprise side of things, Macromedia have shown in the past that they have assumed one thing only to find out its got more capabilities than they originally thought - hence CENTRAL dilemma.
I would bank my house on the fact that if FLEX were to be accessible to the current *existing* market, you would see some very radical and more robust RIA applications vs some basic mini applications. You would also see a more vibrant FLEX community aswell.
I dunno, long term FLEX looks very good in price model, but not many people think past yearly budgets :) and when you get a 50K funding proposal together for a product, you better have answers come the following financial year/qtr and how the product is reducing costs.
Its all about the initial startup sale, once you get past that, your on the FLEX honeymoon nothing but fun :) - the trick is to convince management its a long term investment.
Posted by: Scott Barnes on March 31, 2004 02:22 PM