In a move that has surprised some folks, New Atlanta has announced plans to create a free open source version of their J2EE BlueDragon product. They will continue to develop, sell and support a commercial version of it, as well as continuing the .NET version and the standalone JX version (neither of which will be open sourced).

This move makes a lot of sense for New Atlanta. They started out in the Java world and migrated to the .NET world and the latter became their primary development focus some time ago with enhancements being "backported" to the Java editions. Vince Bonfanti spoke at CFUNITED last year about the increasing focus on and innovation around the .NET version, such as tighter integration with IIS 7 and native administration via the Windows control panel etc - some very slick stuff! Vince also touched on the possibilities of the DLR (Dynamic Language Runtime) that Microsoft has developed which may allow a version of BlueDragon to be developed that targets the DLR and therefore allow CFML to be a first class language on the .NET platform in the same way "IronPython" has elevated Python (and other dynamic languages will follow).

My understanding is that New Atlanta took their original Java code and migrated it to J# to create the .NET version and I expect that since then the code base has continued to diverge, even with a common Java/J# core shared between all their versions. I've said several times that it would make sense for New Atlanta to open source their Java edition and focus on the .NET edition so this announcement comes as no surprise to me. I think it's a very sensible move.

The timeline in New Atlanta's press release and FAQ indicate that they have the move to open source fairly well planned out and they intend to create a steering committee to manage the direction of the project. I would expect this to follow the model of Sun's JCP - Java Community Process - with requests for change coming in from the community and being reviewed by the committee and, if approved, a small team of developers will implement those changes. Like all successful open source projects, the pool of committers is likely to be fairly small and, initially, will probably only be New Atlanta engineers. They have to maintain the integrity of the product, after all, especially since they plan to continue to develop and sell a commercial version of it.

I don't see their plans being much different to how Flex is being managed by Adobe, for example, so folks should not expect a free-for-all approach to just adding new functionality to the open source BlueDragon project!

Will this change the ColdFusion world? Some people are certainly claiming on their blogs that it will. I'm a bit more cautious. Smith going open source hardly made a ripple. BlueDragon J2EE going open source is a bigger deal but I think the reality will be only a small, core set of CFers will actually get involved after the buzz dies down. Apart from the obvious benefit of a free CFML engine that is feature-rich and market-proven, I'm not sure exactly how the open source aspect will change our world.

The real question is whether folks outside the current CF community will sit up and take notice. If the availability of an open source (& free) CFML engine is enough to entice a significant number of PHP folks (or maybe even Ruby on Rails folks) then that could be game-changing.

Bear in mind that the CF community has not exactly stepped up to help with CFEclipse - a Java-based open source project - or Smith and even the CF-based open source projects hardly get any contributions from the communities that use them (although I'm seeing some very small signs that things may be improving very slightly, very slowly...).

More reading:

• Vince Bonfanti's blog post
• New Atlanta press release
• BlueDragon Open-Source FAQ

Where very much on the same page me thinks: http://blog.daemon.com.au/go/blog-post/bluedragon-is-dead-long-live-bluedragon

Here's hoping Adobe rise to the occasion, and make the most of this opportunity.

Let's take bets: my money is on "no" -- people outside the CF community will not take notice...

Well written, as usual, Sean. I tend to agree with most of what you said. I'm also a bit concerned about adoption (I mentioned this in my blog entry, http://cfrant.blogspot.com/2008/03/you-just-became-more-value-aka-thanks.html). My biggest issue to date with Smith Project is I had a hard time picking up the source and using it, it just seemed like too much effort to get it to the point where I could use it in our environment. I sincerely hope is that this will not be the case with BlueDragon. That being said I agree with your sentiment not many people will contribute, but plenty will complain :)

From every account I've seen, Smith is not ready for prime time yet. That's just hearsay, as I've never tried it, but people in the know (e.g. you) have said it and those evaluations mean a lot to me. I think a lot of people only equate ColdFusion to Adobe, and before that Macromedia and Allaire. Will BlueDragon make a ripple? I think it has a better chance. It's been around a lot longer, and it has a better reputation, although it won't make the impact Adobe would if they open sourced ColdFusion. I think this WILL be significant, but it will depend on how this new version is received when it comes out.

It's a great time to be a ColdFusion developer.

I tried to help with smith - but received no response at all :-(.

So I don't think smith was the way of the future...

kind regards, Joel

I don't know, it's GPL, not WTFPL, so it can't be true Open Source :P

Seriously though, Terence Ryan is right (see his blog http://www.numtopia.com/terry/blog/archives/2008/03/yawn_blue_dragon_goes_open_source.cfm), what ColdFusion really needs is more Open Source projects so that people can choose those projects and extend them instead of one of the PHP or RoR projects out there.

You and he are also right that in by open source, many if not most people, really mean free as in don't pay for. Which of course is exactly what New Atlanta is positioning: "This free open-source version of BlueDragon will fulfill a very large niche"... "for little or no cost".

I don't think this will really change much at all, if anything.

As someone who used to do ColdFusion before moving into Python, I can tell you this is a good move, albeit too long in coming and by the wrong people (Adobe should be the one doing this).

The impression by most of the Open Source world is that when you do ColdFusion, what you get for purchasing your server is the same thing you get for free using the product of your choice.

The impression we have in the OS world is that with open source tools in the Python, Ruby, and PHP arena I can generate PDFs, images, JSON, AJAX, and everything that ColdFusion does. If I'm doing it in Python or Ruby and the core libraries don't handle it I can include foreign packages easily via Eggs or Gems in a manner much easier than installing Java bits.

BD going open source is interesting, but they don't follow the CF specification exactly.

As you can guess I've been of the opinion that Adobe should sell CF the same way that Redhat sells its products. They give you an OS, and then charge you Sun/Oracle scale prices for 'Enterprise Support'. This business model works great for them and a lot of other people.

Imagine if Adobe adopted it! The little guys play and convince powers-that-be to use your tool. The powers-that-be demand some way to spend money on the product. And Adobe steps in and says their focus is on service and offers an awesome support plan.

Again, works for Redhat, MySQL, JBOSS, and a bunch of other people.

@Daniel -- but those organization built service into their culture/strategy/operations from day one. Adobe has a few decades of being a packaged software company -- hard to just turn that around and go after an entirely different way of doing business. It's also not clear that ColdFusion would ever have a critical mass of "big iron" customers paying the big dollars for support -- it is a "middle market" product and has been pretty much from the beginning.

Now, don't get me wrong -- I'd love to see ColdFusion prosper and grow, and that may very well require some bold new ideas and approaches from Adobe. I'm just saying that you can't directly transfer the models of the companies you mention into ColdFusion.

I posted a response over at Ray's because I wanted to pick his brain on something that he said that triggered off of something you said. I want to pick your brain too.

You wrote: "Bear in mind that the CF community has not exactly stepped up to help with CFEclipse - a Java-based open source project."

I have 2 bones to pick with this statement. First bone is: Did the CF Community know that they can? What's the bottleneck keeping them from doing so? Not sure how one can complain and say no one helps when you haven't given people a grocery list of things of HOW they can help. Can you? Have they?

I've gone to cfeclipse.org, I can't seem to find a "contribute" link anywhere. I've even googled "Contribute to CFEclipse" and... I get various results, but none directly related to the CFEclipse website detailing what they're looking for or how to even participate.

This isn't the first time that I've seen you state that CF Community isn't contributing. Now I'm going to push back slightly and say I haven't seen any projects that asks for contributions or give any CLEAR directions on how people can contribute.

Are we all just supposedly to magically know subversion to make all this all happen? I'm learning to use subversion, but I don't call myself an expert on it and I don't think that I have enough confidence to blindly contribute code snipplet or otherwise to a global subversion. If I had things written down on a project website that wasn't buried in some obscure wiki 20 links deep somewhere that showed how I could contribute, I might be brave enough to do so. Until then, why bother?

My 2nd bone: No offense, but do you really want a cfml application developer submitting java code to cfeclipse? or do you want someone knowledgeable in java and familiar with cfeclipse to be submitting code? Are you barking (meowing?) up the wrong tree here? I would think that getting a cfml developer's feedback of how cfeclipse could speed up work and asking for support (money, donations, bug testing, etc) is much more important than asking web developer who is using the tool to submit java code.

Maybe I'm thinking backwards these days. o_O

@Nathan D,

Its a hard change, but look at Sun. The common perception is that no one ever got fired for choosing Sun.

But Sun has lost major market share to Redhat Linux and other Linux competitors arguably thanks to all the GCC toolsets added by a select few to the the Linux distros. IBM is a major supporter of Apache. In my world at NASA using Solaris is a hindrance, because out of the box Redhat just gives you so much more. The product is free, and yet management is kept happy by being allowed to spend oodles of cash on 'Enterprise Support'.

Heck, Novell ditched their proprietary OS and now are the maintainers of SUSE Linux. Their web site is powered by all OS tools and they are working on giving back to the OS community. I know cause I've talked to them. Now they make their money providing service for a tool that they give away for free.

Now there are rumors flying that SUN is going to make the next Solaris another version of Linux and they might go the way of Novell.

I can pull out some @#$%ed statistics and point you at the infamous Tiobe index. No matter how good CF is or how well the community supports it, it won't get past the tier its currently sitting in until the business model changes. it doesn't have the MS branding, so as a paid-for product it just can't advance any further.

I doubt that Adobe will go this way. CF8 seems to have done well. But to really move up in presence and respect I'm of the firm belief that CF8 can't move far until its opened up.

The CFEclipse teams over the years have begged, repeatedly, on their blogs for people to help contribute. Mark has repeatedly asked on the CFEclipse mailing lists. The response has been almost universal silence. Saying you don't know how to contribute is lame - if you're on the mailing lists, you can easily volunteer (and you would have seen Mark asking for help). There have been clear directions published several times on how to contribute.

The same goes for Fusebox which has repeatedly reached out to the community with very specific asks (as have the other frameworks) and almost no one responds and most that do only contribute very briefly and then disappear.

I think your comment sums up the attitude perfectly: "why bother?"

As for your second bone... Smith and BlueDragon and ColdFusion are all written in Java. Following your logic, I assume you don't care that about them being open sourced since you don't think CFers are capable of contributing anyway?

And the CFEclipse project has very specifically asked for donations and bug testing and documentation - and been met with almost complete silence on those fronts.

Wait, I see a "donate" link buried under support and how you can log a bug. What the heck is your definition of "stepping up" here?

The announcement from Blue Dragon is perfect in any sense. There is no (free) alternative for any hosting company that wants to offer CF to its customers.

I truly believe that one of the reasons (among others) why PHP/MySQL is so well know and spread, is not because it is the best language, but simply because it is offered with every hosting package and freely available to everyone.

As a informal announcement I can tell you that my company will release its Enterprise Digital Asset Management, previously only available in closed source and with a price, soon to the public in a open source version and based on Blue Dragons J2EE offering. We are rewriting all CF8 code to fit BB code (mainly the ExtJS Ajax stuff). This will then be the Alfresco alternative but written in CF.

Sean, c'mon dude. If you WANT a serious contribution there should be a page DEDICATED to how someone can contribute! If you're looking for contributors, it should be right smack dab on the FRONT page. It's called marketing, no?

Yes, if I used cfeclipse all the time (and, I don't, but I don't bad mouth them and I have referred it to people), then I would be on their mailing list just like I'm on other mailing lists and trying to get involved.

By your logic, everyone that cranks out cfml code is an awesome java coder that is being LAZY! Crap, wait, when did I learn java again?

I know I can read it and figure things out, but not enough to contribute. I'm going to call myself about the middle-of-the-road cf developer here in the "global" scheme of things.

So, you might think that Mark Drew & crew have done enough. I personally think they can do more and all it would take is a link on their homepage and a grocery list of how to get involved.

Anyway, I keep seeing you bring this up time and time again and I wanted to figure out your brain (and, apparently I failed again). :)

@Daniel -- I don't disagree that the ColdFusion business could benefit from "big changes" if they want it to be a lot bigger, but remember that CF 8 is, as you say, doing well -- and it's a small piece of the Adobe pie. Hard to shake things up substantially under that circumstance.

Also, for Sun they were in trouble with their core business at a fundamental level, and they were highly unfocused when things started going badly (and you could argue they were not paying attention to market changes). This graph says it all about that particular comparison, me thinks:

http://finance.google.com/finance?chdnp=1&chdd=1&chds=1&chdv=1&chvs=maximized&chdeh=0&chfdeh=0&chdet=1205265600000&chddm=1000960&cmpto=NASDAQ:ADBE&q=NASDAQ:JAVA

Adobe is actually making a number of fairly substantial moves to avoid major problems as shrink-wrapped software becomes a maturing and shrinking part of the overall ecosystem -- making lots of noise in various online services efforts. But, you are suggesting building a massive professional services organization, something that Allaire, then Macromedia both made some attempts to do with mixed success. Adobe is even further afield than those companies were given the core culture around designers and printers rather than enterprise IT support, so I am simply saying that it's doubtful that a) they will invest the kind of resources it would take and risk what they see as a small but relatively healthy line of business and b) be successful even IF they tried to do it.

@Todd, how many open source projects have you worked on? :)

I just did a quick tour of a number of big open source project sites. A *few* of them had a link on the home page about contributing. Most of them had contributor information fairly buried.

I never said CFers were "awesome java coders" - I was just responding to your comment that you didn't think Java projects would want CFers to contribute code. There are lots of ways to contribute and everyone can contribute at their comfort level. Some CFers *are* Java programmers and some of them could be great contributors to the code. Everyone can contribute to testing and documentation which are always in need.

You seem to think the fault is with the open source project teams. I think the fault is with the community. I suspect we just have different opinions (but mine is based on a couple of decades of contributing to open source projects so I might be a bit biased :)

Sean, I'll bite. I just posted a pathetically small code to riaforge and that's my first one I've released under a "license" and I'm ashamed of it. Have I shared code freely otherwise? Sure, before I took my domain down in 2005, I had some code I gave out to the world.

Wait, what's open source again? Contributing to something with a specific license? Shame on me. I've made zero contributions (other than referrals and discussions) to CFEclipse nor to Fusebox, Mach-II, Model Glue, COOP, OnTap, and and and... do we really need all these frameworks? o_O Which one are we supposed to contribute to again? Can you answer this without saying, "Oh, just freakin' pick one and contribute?"

That being said, I don't use Fusebox, Mach-II, Model-Glue and such.

I'll even back up a little and concede. I don't think the fault is _entirely_ on the open source team. I do blame their marketing/community manager and whomever is running their website for not making it prominent on how to contribute.

I do agree that riaforge needs to be filled up by the community and even more promotion. Once I get back up and running (as my domain slowly shows progress), who knows... maybe I'll get off my ass. :)

@Todd, LOL... I'm not going to suggest you pick *one*... I'm going to suggest you contribute to all of 'em! :) :)

Sean,

Interesting comments. To me what I see the issue in the CF world is that many developers, I think, don't know HOW to contribute. Currently I am working on some gov't work and I see CF developers who have nary a clue of the CF world, let alone th e open source one.

I don't blame them, its just the way it is. Me personally I have at various times contribute to many things either as a beta tester or developer or curmudgeon but what becomes hard is managing that and so in many instances things fall of the radar.

So maybe that too happens to people, who knows. All I know is the blue dragon thing works well this could be a real asset and a nice way of introducing some clients to the beauty of CF. Heck, the thought of being able to code internal CF apps on a free server (CFMAIL, baby!) could have a real benefit.

Todd, I just rolled out the new CFEclipse site and one of things I will be adding there (or more likely on the Wiki) is a list of ways you can contribute back to the project.

I also just blogged about how to contribute.

http://www.thecrumb.com/2008/03/16/giving-back-to-the-coldfusion-community/

Usually in most of these projects there is no 'marketing team' or 'community manager'. We are so lucky to have people like Mark Drew and Ray Camden (and many more) in our community but they all have 9-5 jobs in addition to their CF projects. What little free time they do have goes into the project, not 'marketing'.

My blog goes into more detail but the bottom line is if you want to contribute (and I encourage you to do so) be proactive!

I am thinking of build a new website with very strong possibilities of growing, I have already a very small part of the application in CF, but I want to go to an open source. Please I need a detailed opinion about which one to use and could work better for me (I mean a very big website to spend money into with television publicity) between Open BlueDragon Vs PHP. Those are the choices. In the future I will need Scalability, support and reduce costs. You can email me at ruecast2007@yahoo.com. OPEN BLUEDRAGON VS PHP

@Rer, there are now two strong open source options for CFML: Open BlueDragon and Railo. I'll leave it to others to comment on OpenBD in terms of scalability because I have not yet used it under load but I know Railo performs very well.

You might consider joining the OpenBD and Railo mailing lists for a while to get a sense of their respective communities and to ask specific questions about high-traffic sites running on those two CFML engines.

Look seriously at railo

it is neat.

- Joel

Name:
Email Address:
Comments *** Please note that all comments require moderation so it may be some time before your comment posts to this blog! ***
Remember My Information: