Railo: fragmenting or fostering the ColdFusion community?
Community , Railo , ColdFusion Add commentsAn interesting perspective raised in ColdFusion Panel last week (https://admin.na3.acrobat.com/_a204547676/p34336028/) is that the introduction of Railo, an open-source engine that supports the bulk of the ColdFusion language (as well as introducing alternate features), will perhaps fragment the CF community.
For the record I disagree with this sentiment, though I admit to understanding where the statement draws from.
As I've posted on previously, I feel the life and death of CF is going to be determined by our efforts as a whole in developing an environment that new developers find welcoming, or that experienced developers can quickly see the tangible benefits of committing to. Learning a new development language is like learning a spoken language, with the added difficulty of breaking familiar patterns, conventions, frameworks and methodologies that somebody has been trained in and become accustomed to. I remember abandoning Perl many years ago because the help documentation just wasn't there; CF on the other hand was well-documented and an easy transition. That's no less true now, though the other development communities now have a much larger resource base to draw inspiration/knowledge/understanding from.
Because CF is (for now) a smaller community than say PHP or Java, the suggestion that it might fragment for any reason is a very scary one. Fragmentation leads to in-fighting, less compatible code, a great deal more work in validating applications, and mixed messages broadcast from the community outward. This is not an enviable position.
This said, I do not think Railo et.al. is going to have this effect. Sure there will be camps established that favor one or another, but they will be small and peripheral to the mainstream development community. Any commercially viable organization should approach the multi-engine environment with glee, knowing that by supporting both Railo and Adobe ColdFusion equally (and enforcing through expectation a high level of compatibility) they are going to have a broader audience to sell to. And this brings me to my central, very enthusiastic expectation that Railo will offer something that the CF community has never had: an easier road for the hobbyist or young developer to move from playing around with CF to becoming professionally competent and financially viable.
If you talk to many of the successful PHP, Java or Rails companies, you'll find many company leaders who cut their teeth developing small projects for themselves, friends, family, or small business. A this level any cost is going to impact what they charge for their work, and they sense this will continue even as projects get bigger and budgets expand. Many a ColdFusion development house can tell the story of clients lost only because they couldn't justify or understand having to pay for ColdFusion.
With Railo that situation goes away. Sure there are trade-offs, but the point is that there is now a path that never really existed at this level (not on the fully FOSS level, anyway). Railo will create opportunities where none existed before, opening doors that would otherwise have remain closed and putting another light at the end of that CF tunnel. Railo is their golden ticket to the best web development language around. This said, it is only a ticket ... the enchanted forest still rightly belongs to Adobe.
With Adobe you get corporate support, enterprise credentials and the security of vendor support from a large commercial organization. As a developer I also know Adobe will always focus on making ColdFusion a viable and competitive product. Its integration with other Adobe products will remain peerless, and their work in expanding the connectivity and foundation with other technologies (listen/watch out for for "Link") will continue to build on its unrivaled feature set. Unsurprisingly, a great number of my clients would rather pay for an Adobe license for these intangible benefits alone.
A friend recently said to me "Railo is the Halo". In some ways this is very true. Having a CF engine that a developer could use from birth to death will bring many more players to the field. But if Railo is the Halo, it is still Adobe that gives ColdFusion wings. I personally continue to use ColdFusion 8, will upgrade to 9, and will validate my applications against Railo for clients who choose that path. I expect every company that moves from small projects to a multi-tiered, multi-discipline development will find a need for both products on their shelves.
As long as the Railo (et.al.) and Adobe teams continue to work together (via the CFML advisory committee and other cooperative efforts) and set the example of cooperation and growth, I have no expectation that we will see any fragmentation. If we're voting, mine is for foster.
Aug 23, 2009 at 9:51 PM Grant, this is a great post, and addresses a really important issue. I've always seen Railo as not fragmenting the ColdFusion community, but adding to it. In fact, I really believe that Railo is the key to rapidly growing the CFML developer population.
In order for new (not currently using ColdFusion) developers to try out and potentially adopt CFML, they need an inexpensive (and preferably open-source) engine to work with. Railo's open-source version is available for free - it matches the model that developers coming from PHP, Python, Ruby, Java, etc. are used to.
Adobe's model of making the developer edition available for free is good, but not really enough to drive significant adoption. Because developers can use Railo for free on both their local development boxes as well as production servers, it gives them a risk-free toolkit to begin working with CMFL. Railo also makes using CFML a possibility for developers that are working on tight budgets; for these developers it can be tough to tell your clients that they need to purchase a ColdFusion license in addition to the developer’s fees.
And as we know, there are so many benefits to using CFML - rapid development, great performance and features, an incredible community, tons of great applications and frameworks, etc. Once a developer tries out and really gets a handle on CFML, they’re likely to stick around. And this can only mean good things for Adobe - as you pointed out, Railo is great, but there are times when Adobe’s CFML engine is the only choice. So while it’s true that Adobe loses out on license fees when a developer chooses Railo, Adobe can potentially see a net gain in license-based revenue if Railo is successfully able to increase the overall numbers of CFML developers.
And as application developers (we’re the team behind Mura CMS), we love the fact that we have 3 CFML engines that we can deploy on (ColdFusion, Railo and Open BD). It makes our development and debugging more complex, but it is great for our customers. More choice is better in this case. With Railo, we have been able to sell to clients that we couldn’t reach before, simply because they weren’t interested in purchasing ColdFusion (and we’re talking about some massively well-known organizations here). But we also are able to provide solutions for die-hard Adobe shops who are taking advantage of the extensive integration options that only Adobe can fully supply now. It’s a true win-win for us.
It’s a great time to be in the CFML development community - we think it’s on the brink of a huge growth in popularity. Railo will be the key to that growth, but Adobe is at the center of it all, and we’re excited to see what happens.
Aug 24, 2009 at 2:38 AM As both a developer working in Higher Education and an "out of work hours" freelancer, Railo has *already* been amazing.
At work, we can afford CF, and use it. Other departments can't justify the price of CF, especially when they've not got any CF devs on site - so it's easy to recommend Railo in that context, especially with the express version where they can just "try it".
For my freelance work, there's no way I can justify a CF licence for any of my projects - they simply don't need it for what are very small/personal websites - but as a developer, CFML is what I want to code in.
I now have the best of both worlds: I can use LAMP/VPS providers with Railo on for a *fraction* of the price even a shared hosting account with Win/IIS/CF is. I've literally cut my hosting costs down to a 1/3rd of what they have been already, and have more control, diskspace and bandwidth.
I'd argue, even for existing devs, especially smaller freelancers like myself, FOSS Railo is a God send. It's reignited my interest in all things CF related - beit Adobe/Railo/OpenBD. #Railo FTW!
Aug 24, 2009 at 8:28 AM Grant: Well said, I fully agree. The existence of both Railo, Adobe ColdFusion and Open BlueDragon in and of itself will certainly *not* fragment the CFML community. As long as all engines continue to work together via the CFML Advisory Committee for core compatibility, I see only fostering of the CFML community. As you and other commenters have said, Railo simply opens big new doors that did not exist before!
Before Railo, a PHP developer considering ColdFusion might be quickly turned off by the lack of a FOSS option. Some developers might even work on a CFML project and like it, but not use it for freelance/other gigs due to licensing costs.
Adobe will likely lose a tiny bit of licensing to Railo, but few enterprise customers will be likely to switch. On the other hand, a Java developer may be introduced to CFML via the FOSS Railo, recommend it to a customer/business and later result in a new enterprise-level customer for Adobe.
Businesses that feel more secure in the commercially backed non-FOSS support will be comfortable with Adobe, but more FOSS developers will join the CFML community via Railo. These same FOSS may work for a company that spends money w/Adobe and they simply stick w/Railo for their personal/freelance apps.
I'm extremely jazzed about Railo, but I still have apps running on ColdFusion 8, and I'm still very excited about ColdFusion 9. I see lots of potential for Adobe/Railo fostering the CFML community and nearly zero fragmentation concerns, so long as core CFML compatibility does not fragment.
Aug 24, 2009 at 9:50 AM Another thing to remember is that there is a *lot* of room for community growth. A magnitude of order is certainly a reasonable expectation, so even if the worst of Adobe's fears came true and only 20% of the community bought ColdFusion, that would still double what they are currently selling!
I can't speak to the Adobe corporate mindset, but they are hopefully looking at this from a "how can we benefit/leverage" perspective as opposed to "how can we compete" as the latter pursuit will certainly lean towards fragmentation and lost opportunities (i.e the Unix community).
Aug 24, 2009 at 2:21 PM I have been a very public a naysayer on this topic - not because I dislike or oppose Railo but because I think people to quickly jump on a bandwagon without fully thinking through the consequences.
All the stuff you said is great, except that Railo isn't "free as in beer" - the company hopes to make money off this application and somehow I doubt its hobbyists (who you imply quite rightly wouldn't be buying a license anyway). So if hobbyists and others who can't afford a full Adobe license go Railo - where do they make their money? I just think that if many of the folks like you writing about this are correct - they have an awful business plan...and if they have a good business plan, then your view of this issue isn't correct.
OpenBD on the other hand doesn't appear to be tied to a business plan - its more open source for the sake of open source (at least from what I can see). Yet, it somehow doesn't factor into people's views on this topic much. Is it just the baggage of the name?
Aug 24, 2009 at 2:52 PM @Brian, that all presupposes that Railo (the product) is a financial component of Railo Consulting Group's (the business) business plan. CFEclipse doesn't exist to make Mark Drew a ton of money, Model-Glue isn't making Joe direct income, and Mura doesn't fill the license coffers at Blue River Interactive. As with all those other examples, Railo's income will come from a) consulting, b) demonstrating through Railo a high level of competence in the industry, and c) being an obvious vendor for extensions and enhancements. The beauty is there's nothing stopping you, me or anyone else from competing against them in any of these areas.
As far as I can tell, Railo absolutely is "free as in beer". There is no more Enterprise version, and the only products they sell are enhancements and support (much as in nearly every commercially-fostered OS project does, from Red Hat to MySQL and onward).
Also, let me remind our viewers that OpenBD is a product that went OS only after its stint as a commercial alternative. I never smelled smoke there, and I don't smell it with Railo.
As to the bandwagon, I don't sense a lot of beating drums in the community, just excitement at the new opportunities Railo opens up. Personally I'm still an Adobe ColdFusion guy, have been for more than a dozen years and plan to be as long as they let me. But when the client and situation demand it, I'll gladly use Railo (the alternative being another PHP/Rails/etc. site) To me, that's just good business sense.
Aug 24, 2009 at 2:53 PM @Brian: To expand upon earlier comments, I believe "hobbyists" and other freelance developers, small businesses, etc. will be more likely to enter the CFML community via Railo -- that's not to say that there won't be enterprise Railo clients supporting the business. When you say "and if they have a good business plan, then your view of this issue isn't correct," are you suggesting that a good biz plan by Railo will fragment the CFML community or fragment Adobe sales?
An analogy to this situation might be RedHat Enterprise Linux, which supports a profitable business despite hundreds of other FOSS Linux distros, Windows, Mac, Solaris, etc.
I may just be misunderstanding your statement -- maybe you can clarify?
Aug 24, 2009 at 3:01 PM @Brian,
Railo is "free as in beer" just like every other open-source project. They make their money on consulting and support just like many other open-source companies, they give away a fully featured open-source product and sell services, support, customized extensions, and for those that need it, a license.
It's called professional open source and its a perfectly appropriate business model. there is no such thing as "open source for the sake of open source" there is ALWAYS a hook attached to the bait (except if it's staffed by volunteers). Companies are driven by profit, and I hardly doubt openBD is any different in it's role to drive profit to new Atlanta in some fashion.
Aug 24, 2009 at 3:06 PM Brian, I think your reply is actually somewhat off-topic (the question is not the validity of the Railo business model, but rather what effect will Railo have on the overall ColdFusion community). However, I'll reply to it because you bring up an interesting point.
But first, I want to know what you - someone that I consider a big champion of open-source projects in the CFML community - think that the addition of Railo to the CMFL community will bring? Could it bring a significant number of new users? Could Railo's affiliation with JBoss bring new opportunities for selling CFML solutions to Enterprise clients? Does the ability to sell lower-cost solutions benefit entry-level developers? And if any of these are true (and I believe all of them are), doesn't it benefit all of us in the community?
As for Railo's business model - it is not dependent upon entry-level developers adopting and using Railo. However, some of these entry-level developers are the corporate developers of tomorrow - and this is the audience that will be driving adoption of specific technologies within their organizations. But I'm off on a tangent of my own here.
I am not part of Railo, and my take on this may not be fully accurate, so please check with Gert and crew for the final word. However, I think that Railo's business model has many similarities to that of JBoss. JBoss gives away their software, and has multiple revenue streams - support, training, consulting, etc. JBoss is clearly a huge software success story - they were acquired by RedHat for 350 million, and they were very profitable at the time (and remain so today). Do you see any reason why this business model isn't a good fit for Railo? Remember; in software sales, increased adoption is fundamental to success, and this is much more likely with an open-source business model.
Aug 24, 2009 at 3:44 PM @Malcolm - Is BlueRiver/Mura still the exclusive retailer for Railo in the US? I know this was the case when Adobe and Railo met last year in Newton to discuss partnership opportunities (of which you were present), but didn't know if that had changed since Railo US was formed.
FWIW. I must disagree about the fragmentation, because I've already experienced it on several occasions. Albeit, I meet/talk with 1000s of ColdFusion developers a year so I'm a bit hyper-exposed, but sure enough, fragmentation is occurring. The fact that the question was asked in the first place, is a sign that the community is starting to feel it as well. Unfortuantely, Railo is spending all of their time marketing at CFUGs and CF conferences, so this promise of 'new developers' from PHP/Java doesn't seemed to be backed with actions. Hopefully, the CF community will keep them as honest as they keep Adobe.
-Adam
ColdFusion Product Manager at Adobe
(a.k.a. biased but honest)
Aug 24, 2009 at 3:56 PM I think that the Railo/Blue Dragon/Adobe situation is quite similar to the Firefox/IE/Safari/(etc) situation. While competition drives innovation, cross-platform compatibility becomes a problem. Eventually one platform lags behind the others in features and/or syntax, and each will have it's own unique bugs and patches, which becomes a problem for us developers, especially on open-source CFML projects. And there's no reason to do finger pointing, because they're all going to have problems at some point.
The easy answer is Adobe is the original vendor, so just support that. However, the CFML advisory committee has already been the catalyst for some new CF9 features that we might have never seen otherwise.
So, I don't think that the answer to this blog post's question is either/or, I think it's both. I also think that the benefits currently outweigh the headaches, as long as the communication and cooperation between all parties stays consistent and constructive.
Aug 24, 2009 at 4:44 PM @Adam: to be fair I actually I didn't raise this issue, you did (in CFPanel). What I did is respond to your comment in a blog. I do greatly respect your opinion, coming from somebody whose vantage point is much higher than my own, but down in the trenches at CFUnited I sensed only excitement and new possibilities...not a moment or shadow of 'which camp do you support?'
@Shannon here here and well said! Competition is the grease that keeps the wheels turning, and what would a programmer's life be without validating to separate specifications (well, more rested I suppose). It's actually a credit to Adobe that their products evolve so much without this stimulus.
The fact is, ColdFusion is an open specification. Adobe must have realized this day would arrive (as it did many years ago with OpenBD). It's a new world, so we *all* better get used to it. Railo/OpenBD have to accept that the vast percentage of ColdFusion programmers will continue to consider the Adobe version as the "true" CFML specification. Adobe will have to accept that their price model was a factor in community size, and that they may have have to adopt a new business model to remain competitive within this slice of their portfolio.
Both should realize that we in the trenches expect them all to behave and play well together, and that we will vote with our dollars and attention should one of them betray us. It might be their sandbox, but it won't be much fun if nobody's showing up to play in it.
Aug 24, 2009 at 4:46 PM oh and @Jamie: thanks for pointing out subscriptions weren't working. Hack complete ;)
Aug 24, 2009 at 4:49 PM @ Adam, Blue River and Railo are just friends now, with no official business partnership. I believe that the new Railo US organization is a much better structure for Railo at this point in time.
You bring up an interesting point about fragmentation, so maybe an interesting question to ask is what does it mean to you - it may be different for you than for me.
I think of fragmentation in the context of a scenario where a company or shop would "pick a side" and develop specifically for one engine or another. For us, as a software development company, we make sure we stay compatible and portable across the different engines. It would be tremendously counter-productive for us to create a Railo-only or Adobe-only version of Mura CMS, so our approach could be considered "anti-fragmentation". We believe that both are great engines, and we want to make sure that we are able to work with both (and with OpenBlueDragon too) for the foreseeable future.
We may end up creating some Mura CMS plugins that take advantage of specific features from CF9 or Railo that don't exist in the other engine, but these would only be add-ons, not core code.
I think there are two possible scenarios here:
1) Railo, if marketed correctly, can lead the way to significant growth in the CFML marketplace. If this does happen, individual developers and Adobe alike will profit in a major way from the increased awareness of and demand for CFML.
2) If the growth of Railo only ends up reducing Adobe's license revenue without increasing the overall number of CMFL developers and deployments, then that would be a real shame, as it would weaken Adobe's return on investment without providing any new benefits. This scenario, I would guess, is at the heart of Adobe's concerns, and this would be a bad outcome for entire community.
@ Shannon - I think you hit the nail on the head when you said: " the benefits currently outweigh the headaches, as long as the communication and cooperation between all parties stays consistent and constructive".
Hopefully the CMFL advisory committee can continue their work on a common standard that all engines will adhere to, and then users will choose the best engine for their needs.
Aug 24, 2009 at 6:36 PM I have no doubt that Railo is adding huge value to the CF community for a number of reasons. Railo has made us think about what we use to develop for a number of reasons, so we are split between Railo / CF as I'll explain.
From our end CF has some major limitations in the way we can deploy the product on an On-Demand platform whereas Railo is a much better architecture. So for a hosted service platform Railo is superior IMHO, even if Adobe release an Amazon instance you are still trapped in CF Enterprise sandboxing for multiple web site hosting which just didn't work for the model we are trying to deploy. Adobe I'm sure will react and chance this in the future but it would be to their timeline which might be too long for us.
On the other hand it's hard enough to sell CF with Adobe backing into larger corporates without having to work through what CF is etc, so the Adobe brand and more stable / tested and CF can be an easier sell.
It's also not that easy to do a major product port to Railo (like porting to another DB server), 95% is fine out of the box but you need to test, adjust lots of small bug and update documentation etc, etc. You also then need to support 2 CF servers going forward which has a cost. But for us this is better than moving to another language to achieve our goals which may have been the case if Railo wasn't around.
We commit to spending well over 100K a year on CF licenses with Adobe each year and have done so for the last 5 years, in that time I've never seen a sales or account person visit our office so I don't think this is that big deal for Adobe. On the other hand Gert has visited us (we're in New Zealand so reasonable trek, although granted he was going to webdu). We also get direct access ( chargable of course) to the Railo engineering team which helps us build something specific for us. Maybe there is a way to do arrange this with Adobe but I"m not sure. Money does play a part in this but we've never had issues selling ColdFusion before or costing the license into our business model,although when you go to an On-Demand model it is different, as you dont need restrictive or cost prohibitive plans stopping you scale. So until we see what Adobe have in this space we couldn't build or plan anything or have any idea what timeframe this will happen in.
So yes, Railo will fragment the community is some way, but it may also keep a lot more people in it as well, and ultimately over time it will add new users for sure. It will give us options to explore new ways of doing business that might not be possible from an Adobe perspective and puts pressure on Adobe to really come to the party to keep us as customers. I"m sure Adam will be able to get more resource, budget and people on board with the competition as a lever. How is that bad?
Aug 24, 2009 at 7:25 PM @Brian Rinaldi: Have you bothered to look at the license for Railo? LGPL v2.1 - If you want to package up Railo and sell it, feel free. The 'Railo Team' is more for consulting and offering services. The 'enterprise bundle' is a bundle of all the paid plugins. You too could sell plugins with Railo as well if you wanted. Railo is 'free as in beer' and then some.
If you think that OpenBD isn't tied to a business plan, I think you need to look at where the dotted lines are again and remember where it came from in the first place?
Aug 24, 2009 at 8:25 PM @Malcolm, great points, and I too agree that Shannon hit the nail on the head with "the benefits currently outweigh the headaches, as long as the communication and cooperation between all parties stays consistent and constructive." You have pointed out the only possible negative effect I can think of: Railo takes away enough of Adobe's business while overall CFML community growth stagnates, which could diminish Adobe's investment in ColdFusion and therefore maybe eventually have a negative impact on the community.
@Adam: Adobe has been reporting strong growth numbers, so hopefully this trend continues and all positive for everyone?! Also, I'd tend to agree with Malcolm's assessment of "fragmentation" here -- do you agree? This discussion, as I understand it, is about fragmentation within the CFML community, not just possible effects on Adobe ColdFusion profits. I think we're all doing a ton of speculating here, but with positive CFML community growth (albeit slower than we all might prefer to see) and all the buzz about both Railo 3.1 and ColdFusion 9, I just find it hard to find any negative. Have you really witnessed fragmentation that you'd expect to diminish rather than grow the CFML community in the foreseeable future? Any examples you might be able to share about the CFML community fragmentation (due to Railo/Adobe) you've already seen on several occasions?
One last thought... I'm only jumping in to this conversation because I have hope! Like others, I've not heard a peep about what I'd characterize as CFML community fragmentation around the existence of Railo, Open BlueDragon and Adobe ColdFusion. I truly hope that core CFML compatibility is respected and achieved by all, and I hope I'm fortunate enough to continue to work on exciting projects that involve each CFML engine.
(PS: @Grant, I am receiving comment e-mails now, thanks!)
Aug 24, 2009 at 8:36 PM I believe @Brian R made very good points related to ongoing commercial viability, whatever we say it is a pre-requisite of any businesses survival. I am not saying Railo does not have it, just that they need it, Gert is an incredibly nice bloke and very genuine human being and I would always wish him the best of success. It would be incredible if OpenBD and Railo encouraged developers to jump in, it peeves my no end that Ruby on Rails got so much traction. My last thought is this, if there are any differences between ColdFusion, OpenBD and Railo which mean that any application cannot run seamlessly on any one at all times, then that is fragmentation in my opinion.
Aug 24, 2009 at 8:56 PM @Mike: Very interesting thought on applications running seemlessly on all 3 engines. My initial reaction is that it's a bit rigid, or extreme, but I can certainly see where you're coming from. My hope is for a very seamless core language compatibility. The vendor differences are inevitable, but if they're non-core, and only some applications are vendor-specific, then I think that would be okay. For argument's sake, think of existing ColdFusion apps that require LiveCycle ES -- such an app is not even "compatible" between two ColdFusion licenses, as it must also have a LiveCycle ES license. This is what I'd expect Railo add-ons to be like (ones that you may or may not need to pay for and may or may not be developed by Railo Technologies, the company). There are also apps that run only with a MS SQL Server database, or only MySQL, and others run on 5 different database engines. There are already some Windows-only features in ColdFusion.
So, I feel a certain degree of differences between CFML engines will be fine, however, I must admit that it could be tricky to draw that line for "core language." But I'm hopeful!
I'll also submit that a certain degree of what might be considered CFML language "fragmentation" could be healthy. If Adobe was the only game in town, and their business decisions don't line up with every customer's use case, then that customer is off to RoR, rather than Railo, which is still in the CFML community. For such situations, Railo and Open BlueDragon provide the possibility of direct interaction with a customer to grow a feature. Adobe may still refuse to add it, or may be encouraged to do so.
As much as I love Adobe's products, they're already making many decisions that make my life difficult. I've chosen to use Linux as my development environment, but Adobe has another type of "fragmentation" in support -- no ColdFusion Builder for Linux, no Flash Builder for Linux, weak Flash player support. You can see how this is fragmenting the Flex development community:
http://bugs.adobe.com/jira/browse/FB-19053
Luckily CFEclipse and IntelliJ IDEA exist, so I can remain in the CFML and Flex development communities. I know there are reasons and tough decisions that need to be made at Adobe, so I'm not here to knock any of the amazing stuff they've done -- they've gotta run a business while also trying to please a varied community. I fully respect and appreciate that. I also fully respect and appreciate the professional open source software model. As I said earlier, I truly hope I can continue to be a part of both for years to come!
Aug 24, 2009 at 11:05 PM On a daily basis I switch between CF7/8/9 and Railo and at the end of the day Railo is my new home. The engines are so similar in many regards that to work in CF and hours later switch to Railo is a trivial experience. This idea of fragmentation frankly sounds like FUD from the camp that poses to lose the most in this deal. In all reality the closed and expensive nature of CF is what has hurt it most over the years, Railo fixes that issue. I was considering developing several things in RoR last year, and it was Railo that brought my full attention back to the CFML community. If anything Railo has brought me deeper into the fold. Just a few weeks ago I launched linkm.in AKA Linkmin there is no way I would have payed for a CF license for that site, it makes me no money and was just a toy project. Additionally I'm building an app called shoutEngine that also would not have been possible without Railo as the licensing would be quite expensive if the projected growth happens. At the end of the day Railo has reaffirmed my commitment to the CFML community and I've been far more active because of it.
Aug 25, 2009 at 6:15 AM So I think I screwed up my "free as in beer vs free as in speech" analogy...so let me clarify some things.
First of all, I am not a Railo hater (or out to promote OpenBD specifically). I consider Peter, Sean and Mark friends and wish them the best of luck with their business (and Gert seems like a nice enough guy). I just want to add a healthy amount of skepticism to the Railo lovefest that has happened across many ColdFusion blogs where Railo seems to be the solution to all our problems and a cause of none.
First, why is their business model off-topic? (and thanks @Gary for the somewhat condescending education on POSS - I know what it is thank you...though I am not sure you understand it clearly). You are claiming that this product is uniquely important to our community, correct? We debate Adobe business as it relates to ColdFusion and the future viability of the product. Why then would Railo's viability be off-topic to a discussion of its potential impact on the community.
There seems to be a reflexive defensive nature when you criticize or are skeptical of a company with an open source product. While Gert seems nice, open sourcing Railo was not some grand altruistic gesture. This was a business decision - he thought his company could make more money and compete better with an open source license than not. The plan originally seemed to be going after an untapped market of Java developers who are looking for a simple web UI language. That doesn't seem to have panned out. It then changed to targeting existing CF customers and making money off consulting...in fact, that plan seems a little in flux as an outsider too. I don't blame Gert for continually changing his plan to adjust to the realities of the situation but I think its fair to say that the viability of this product and company, which is new on the scene relatively, is unproven.
This is all very important to Railo's impact, and I can assure you the plan isn't to get penny ante hobbyist developers to sign on...there's no profitable business model in that. The issue is, if you as a company sign on to adopt, say, Mach-II and Mach-II suddenly stops being developed any longer, you probably have the in-house expertise to continue to maintain that product internally indefinitely. If the Railo engine was suddenly no longer supported and built by Railo (the company) - most companies probably do not have the expertise to maintain this in-house.
Lastly, I mention OpenBD to express my concern that if this is truly purely about having an open source and free alternative - why is OpenBD so quickly dismissed? AFAIK, OpenBD is not tied to New Atlanta (and a quick look at the developers working on the project I think that claim that this is a bait and switch for New Atlanta falls apart - http://www.openbluedragon.org/about.cfm). From what I can tell, this is just open source (even if the reason for originally open sourcing it was it wasn't a viable for sale product anymore) and not tied to a business model. OpenBD is in fact a much more proven product than is Railo...but its trashed and dismissed in ways that smack of pure Railo "fanboyism" completely detached from any real tangible connection to a concern for an open source and free alternative that the hobbyist developers can use.
Ok...so I am sure the same people will find all the more reason to go after me for this comment but while I am excited about the growing openness and options in the ColdFusion community, I think its healthy to be skeptical.
Aug 25, 2009 at 7:50 AM @Brian: Everything in your last comment seems spot on to me, but I'm not sure exactly what your "healthy amount of skepticism to the Railo lovefest" is. Sure, there are some blog posts showing some great excitement about Railo, and this is usually because it does everything that said blogger needs it to do in the same way or better than ColdFusion, and it's free and open sourced -- it's hard not to get excited about that! That's not to say that there aren't still compatibility issues, and I certainly wouldn't say that it solves all our problems, but if it perfectly solves a given developer's problem, for free, then cheers to that.
I can certainly understand a discomfort with "fanboyism" of any sort -- I for one try very hard to not put down alternatives when I'm excited about a given solution or tool. Some may not take the time or consideration to do so, but that doesn't mean it's a bad thing to be excited about Railo -- I hope there's nobody saying anything like "Railo is great, so long Adobe" -- I have not seen any of this. Then again, this may be the case for certain developers or development shops, and the same goes for ColdFusion, and the same goes for Open BlueDragon. If a company chooses one engine, because it's a great fit for a given product, super -- this does not fragment the CFML community.
If open source framework authors choose to ignore one or more vendors, then this has the potential to fragment somewhat, but so far I think there's a good effort to provide cross-CFML-engine support. Maybe a little less so with OpenBD, but I believe that's because OpenBD is lagging in compatibility a little bit. Part of the excitement around Railo was the huge jump in compatibility with ColdFusion 8, along with a great architecture and the jump to full OSS.
Regarding OpenBD, you said: "it's trashed and dismissed in ways that smack of pure Railo "fanboyism" completely detached from any real tangible connection to a concern for an open source and free alternative that the hobbyist developers can use." Really? I must admit it's often ommitted from conversations involving Railo and ColdFusion, but I haven't seen it outrightly trashed or dismissed -- any examples? Again, my best guess is that compatibility is the main reason for OpenBD omission from some CFML engine conversations. Well, I do now recall it being somewhat dismissed due to old bad blood with New Atlanta -- I can't even keep track of that drama, so it hasn't affected me :) I've liked a lot of what I've seen with OpenBD -- for me, it's just not quite as compatible and feature-rich as Railo and ColdFusion. I'd love to see OpenBD catch up on compatibility and improve performance, and who knows, maybe that will happen very soon!
So, again, I agree with your comments. Sure, Railo (the company) is a business and I'm sure they hope to make money, and I'm sure open sourcing Railo was a business decision (at least at some level). But, what's the problem? Is your skepticism simply regarding Railo business motives? If so, and since you've brought this into this conversation, does this mean you fear that Railo business motives will somehow fragment the entire CFML community?
Regarding your analogy about a given shop likely being able to take over support for Mach-II, but not Railo if either was dropped by current supporting development teams -- again, how would this fragment the CFML community? I'd agree that most shops would not likely take over Railo support to support a product, however, if the CFML community continues to grow around ColdFusion/Railo/OpenBD I bet there would be a strong group of developers out there that *would* take over support of Railo in the community. This is a risk that we all take when using an OSS product -- there are some really big enterprise products running on OpenBD today, and they probably wouldn't take over support for OpenBD either. Again, my hope would be for continued core compatibility, so in the worst case scenario, I'd expect to be able to convert my apps from any one engine to another. Some apps would migrate extremely easily, others not so much, but this is a decision we all make every day.
I really do agree with your comments, Brian, but I'm not sure I fully understand what your skepticism is really about, or how it might be related to fragmentation of the entire CFML community. Maybe we all have a slightly different idea of fragmentation in our heads here? Maybe I'm missing something else? I'm still really excited about CFML, in general, and would love to see ColdFusion, Railo and Open BlueDragon all continue to flourish! Sometimes typing these discussions can lose a lot of meaning, eh? I'd love to have a quick IM chat or Skype voice call if you'd like to chat and better convey our thoughts. DM me on Twitter (@jamiekrug) if you have time. Cheers!
Aug 25, 2009 at 8:18 AM @Brian, I fully agree that skepticism is a very healthy attitude, and I should be clear that this post wasn't to push the Railo agenda (whatever that might be). I'm a Adobe ColdFusion developer, have been almost my entire professional life, and I have never used Railo in a single commercial project, much less done much more than install it and play around a bit.
I think the current focus on Railo is a result of a few things: one the fantastic team Gert has assembled, another that there is a strong sense of a driving engine behind the project. For whatever reason OpenBD has dropped off the radar screen for many of us, and while that doesn't make it any less viable it does explain why Railo is the hotter topic.
Whatever Gert's motivations were for releasing Railo, the fact remains it is GPL2.1 now, and as such has a great deal of *potential* for changing the game. As to personal skepticism, I consider their participation (along with OpenBD) in the CFML Specification a move in the right direction. The fact that Sean and Mark specifically are involved, both of whom (like yourself) have a *very* respectable track record in the CF OS realm is another. (And anybody who attended Peter Bell's CFUnited presentation is bound to be impressed by his abilities).
I am certainly no Railo "Fanboy", but I am very excited about its potential effect on the growth of the CF Community. Some of the replies here do certainly hint that aspects of our community may fragment according to preferred base engine, but the fact is those who choose Railo or OpenBD or Adobe ColdFusion are doing so because something about the product is particularly appealing beyond just the "free"...if that weren't the case, we'd all have moved on to "free" RoR or PHP long ago. If the arguments are only going to be about "loyalty" or "getting on board the OS train" then the decision process itself is broken, so I restate my opinion that the camps that form behind these positions will be peripheral and irrelevant in the face of what the community really wants: growth, support, commercial viability and mutual respect.
Personally, I'd love to see a CFPanel discussion on this topic sooner rather than later. There's no point in hiding from it, and conversation is the best way to move from assumption and expectation to understanding (for good or bad).
Aug 28, 2009 at 9:13 AM Grant,
interesting and important blog post !!
I think everyone should decide on his own, if he offers his technology open source, commercial open source, on a SaaS basis or as a license product. Each model has its advantages, but of course each model has its disadvantages, too.
The guys at Railo did a great job in the last years, but - as everyone knows - they have to earn money to survive (or is it a Railo foundation? ;-)
But please don't forget: the guys at Allaire / Macromedia / Adobe (we are using ColdFusion since more than 10 years) did a fantastic job as well. They developed a technology that was (and is) the basis for all of your and our CF businesses.
We at CONTENS would be very happy, if Adobe, Railo, NewAtlanta and all the upcoming CF products in the future would co-exist in a friendly way. For a CMS vendor like us it's simply not possible to concentrate our activities on one exclusive CF partner (although this partner "recommends" this...). Would be the same as if a BMW would only run exclusively with petrol from Exxon or Shell. Or vice versa Shell petrol would only run BMWs...
We will see what the CFML advisory committee can do for us partners.
Jan 18, 2010 at 11:22 PM Nice post indeed. I just bookmarked this blog.
However, until today, this user group community has been fragmented. Adobe Groups gives the Adobe community a place to connect, ... They are a tremendous resource to the Adobe community, building connections and fostering learning ...
first off, Fusebox is dead because ColdFusion is dead (or at least ..... fragmentation of the CF community.
Jun 29, 2010 at 3:14 PM I don;t think it will be fragmented, there will just be squabbles and bickering over which is best as there are with other languages and technologies.
If anyone wants to try out railo but can't gets to grips with the setup, they can get some FREE Railo or indeed FREE ColdFusion 9 developer hosting over at www.cfmldeveloper.com and compare the two engines and see which they prefer.