Below is a an internal Adobe employee intranet article that was posted with the launch of ColdFusion 8. It provides a bit of a "behind the scenes" peek into the making of ColdFusion 8 and contains a few notable quotes from some of the key Adobe executives that helped make ColdFusion 8 possible.
An internal Adobe staff writer interviewed a bunch of us on the ColdFusion 8 team, the following is the result, and I thought you might find it an interesting read.
ColdFusion 8: Lightning in a Bottle
July 30, 2007
Adobe announces ColdFusion 8, the highly anticipated new release of a product used by approximately 350,000 Internet application developers worldwide.
Adobe today announced ColdFusion 8, the latest version of a product used by hundreds of thousands of developers to create compelling Internet applications quickly and with ease. This is the first release of ColdFusion since Adobe acquired Macromedia in December 2005.
“This release is like lightning in a bottle – it’s something special,” commented David Mendels, senior vice president, Productivity Business Unit. “Thousands of customers can’t wait to get their hands on it, and our ColdFusion team has done an exceptional job making it the most powerful and versatile version of the product ever.”
“ColdFusion 8 rocks!” added Kevin Lynch, chief software architect and senior vice president, Platform Business Unit. “It has an impressive array of new capabilities, including integration with LiveCycle Data Services, support for Ajax data binding and presentation, smooth Flex integration, image compositing, PDF reports, and much more.”
“It clearly reflects how the ColdFusion team continues to listen to customers,” Lynch continued. “This release will enable hundreds of thousands of ColdFusion developers to solve real-world web application challenges more quickly and with even more engaging, effective user interfaces.”
For nearly all of its development cycle, ColdFusion 8 was part of the former Enterprise and Developer BU, which was recently merged with the Knowledge Worker BU to form the Productivity BU, led by Mendels. With the reorganization, ColdFusion is now part of the Platform BU headed by Lynch.
What Is ColdFusion 8?
ColdFusion 8 enables developers to quickly and easily create compelling websites and other Internet applications that can be smoothly integrated in virtually any enterprise environment. The version announced today is the fastest ever, even when running applications built with a previous version. It is loaded with new and enhanced features based on extensive, ongoing customer feedback.
“Customers absolutely love ColdFusion 8 because it’s easy to learn, fun to use, and powerful enough to create real-world, data-rich, enterprise-class applications, without having to go through a painful learning curve. In fact, most developers find the product fun to use,” said Ben Forta, senior technical evangelist.
Who Will Use ColdFusion 8?
ColdFusion 8 will be used by developers who want to create faster, more compelling Internet applications in less time, and IT project managers who want to improve the productivity of their teams. “And they want it immediately,” said Forta. “Indeed, many of the beta testers have already started deploying ColdFusion 8 even before the product is shipped.”
Now 12 years old, the product already has an intensely loyal following and is used by more than 10,000 organizations, including 75 companies in the Fortune 100. Well known enterprise customers include AT&T Wireless, Bank of America, Boeing, Caterpillar and U.S. Bank. Other institutions, such as the Mayo Clinic, the Peace Corps, and the Wharton School of the University of Pennsylvania also rely on ColdFusion.
By far the largest segment of ColdFusion customers work in national, state and local governments. In fact, many public-facing .gov websites such as the U.S. Senate, National Park Service, UK Parliament and European Commission have been built with the product.
Many customers use ColdFusion to build public-facing Internet applications. Others rely on it to build non-public-facing Intranet applications. For instance, both Adobe.com and Inside Adobe are built with ColdFusion.
Reasons to Buy or Upgrade
The ColdFusion 8 product marketing team and employees in Worldwide Field Marketing list many compelling reasons to buy or upgrade to ColdFusion 8, saying customers can:
Experience out-of-the-box performance gains – even for applications written with a prior version – because ColdFusion 8 is the fastest version of the product ever
Use the Server Monitor to see exactly what is happening in every application, and identify and fix bottlenecks via a sophisticated Flex interface
Debug applications more easily and effectively than ever
Enhance the end-user’s experience and productivity by creating personalized, multimedia-rich applications in Flex or Ajax, making it possible to interact with PDF forms and view rich-media, on-demand presentations
Smoothly integrate Internet applications with nearly any enterprise environment thanks to support for Java and .NET objects
Leverage ColdFusion 8’s integration with Microsoft Exchange Server to enable applications to access calendars, contacts and tasks
Deploy applications natively on leading J2EE server platforms including IBM WebSphere, BEA WebLogic, and JBoss
Building on a Successful Beta
Focused efforts to market the ColdFusion 8 release began months ago. The product’s public beta, announced on May 30, was extensively marketed by targeted email, blogs and on appropriate websites. Ben Forta and Adam Lehman, ColdFusion specialist, visited 41 cities in four weeks to meet with ColdFusion developer groups. They, along with Tim Buntel and Jason Delmore, also presented at dozens of events, demonstrating key features and benefits of the beta release.
“We targeted meeting with 2,000 developers and ended up meeting with more than 3,000,” said Forta. “Many of them can’t wait to get their hands on the product, and survey results indicate that the vast majority of attendees plan on upgrading and deploying the new ColdFusion immediately.”
As a result of the tour and other marketing efforts, 14,000 developers took part in the public beta – nearly three times the target of 5,000 set by the ColdFusion 8 team. “ColdFusion 8 has so many great features, and we’re already seeing developers create innovative applications that display the power and versatility of this release,” said Jason Delmore, senior product manager.
Spreading the Word
Now ColdFusion 8’s marketing team is building on the momentum created during the public beta by aggressively promoting the launch, with an emphasis on the product’s existing base of customers and government as a key vertical. According to Andrea Woolner, senior marketing manager, tactics in the campaign include:
Targeted email announcing ColdFusion 8 to the product’s large base of loyal customers around the world
Paid search ads
A new banner ad for highly trafficked developer websites
Blogging (there are scores if not hundreds of ColdFusion 8 bloggers, including many members of the development team at Adobe)
Advertising on the ColdFusion 8 community website
Live events and training, including a launch event, House of Fusion, this week in Washington, D.C., and upcoming AJAX conferences
Sessions at Adobe’s MAX event in October
The product marketing team has also created a wealth of new content for Adobe.com, including a customer video, customer testimonials, demos, data sheets, white papers, and more.
Behind the Scenes – Developing ColdFusion 8
ColdFusion 8 was developed by Adobe employees in Newton, Massachusetts, and Bangalore, India. “We had team members in Bangalore and in the U.S.,” said Damon Cooper, director of engineering. “It was a bit of a challenge with the time difference, but we made it work. We had an outstanding team.”
Cooper said the research phase of development started in September 2005 and lasted about 18 weeks. Everyone on the engineering team – both in development and quality engineering – split into subgroups with about five or six team members. Each subgroup focused on a specific area related to the next release; for example, competitive products or the server environment in the enterprise.
After researching their respective areas, the subgroups reported back to the ColdFusion 8 team as a whole and communicated their findings and recommendations. “Everyone was involved,” explained Cooper. “As a result, I think everyone became a passionate champion for certain new features and enhancements.”
When considering the subgroups’ recommendations, the team as a whole looked at
Benefits to customers
The cost of building a feature
The potential return to Adobe
Integration with Other Adobe Products
After Adobe’s acquisition of Macromedia, which took place a few months into the development cycle, a big focus for the product and development teams became making sure that ColdFusion 8 leveraged other Adobe products.
“We had some great opportunities to leverage other Adobe technologies to enhance the power of ColdFusion, including PDF, Acrobat Presenter and LiveCycle Data Services,” said Delmore. “We wanted to make sure that ColdFusion 8 took advantage of everything Adobe had to offer, and demonstrate Adobe’s commitment to the product.”
Once the ColdFusion 8 team decided on a feature set and built early prototypes, they began meeting with customers through the Synchronous Development (Synch Dev) process. Synch Dev is a customer-driven development process that requires engaging with users early and often. It was broadly used at Macromedia and usually involves five waves of customer visits, totaling about 30 half-days.
“As we went through Synch Dev, we ended up refining and refining and refining to the point that we felt very good about our feature list and the future of the product,” said Cooper.
After each Synch Dev wave of customer product validation visits, the product pitch and prototypes were modified to take customers’ feedback into account, before the next wave of customer visits began,” recalled Cooper.
Once the picture of the final product and the must-have features became clear, the engineering team began feature-design and development work. The development process was broken into “Dev Blocks” averaging about 12-15 weeks each in duration. As each Dev Block was completed, the entire engineering team went into a two week closedown period – for example, just prior to the Alpha 1 release, the Alpha 2 release, the Beta 1 release, the full public beta (internal release candidate), and the GMC/final releases.
“At each stage, we invited customers to participate. We didn't want to miss anything major,” said Cooper.
All Hands on Deck
Prior to the public beta and every milestone release, the ColdFusion 8 team had a close-down period that required “all hands on deck,” recalled Cooper. “For two weeks, every development engineer, every quality engineer, put their feature development pencils down and focused solely on making sure all aspects of the product were the absolute best they could be.”
“Everyone on the team understands that quality is a feature and probably the most important feature that we ship,” Cooper continued. “Knowing how many customers rely on our technology, we all wanted to get it right.”
Early feedback from beta customers and developers at conferences indicates that the team has, indeed, got it right.
“Members of our team recently attended the CFUnited Conference, and after breakout session on Ajax, they were mobbed,” recalled Cooper. “Many of the 600 developers who attended those sessions, as well as others at the conference, wanted to know how ColdFusion 8 was built, how we did it. Some of our guys felt like rock stars.”
Delmore added, “As we went through the Synch Dev process, customer feedback went from great to ecstatic. We’re keeping our fingers crossed that this launch will be something big. This release could have a huge impact.”
“I agree,” said Cooper. “I think the ColdFusion 8 team has done something pretty extraordinary here.”
Employees Referenced in this Story
Adobe has a fundamental issue with their pricing model and this needs to be sorted because you need to grow CF use.
Even the simple solution of including sandboxing in the standard edition to enable hosting companies to securely provide CF facilities at a lower cost level would seem logical.
There are companies out there producing Open Source and low cost CFML engines and I think this is where the growth in CF will be.
Now I know that the development time is fantastic, but Adobe really needs to look at the bigger picture and grow their market share.
Great work by the team, but you need to sort the pricing and in particular you need to sort the difference in pricing between Europe and US. Adam
The pricing difference is to be expected. It IS lower than it was if you compare US/UK prices for Enterprise between 7 and 8, but there are costs to supporting the UK customer base and unless you want to see the price change on a daily basis, Adobe has to cover the off chance that the Dollar will eventually pick up from it's current low.
I hope that Adobe - Denmark is reading it. As the only news about CF8 is that it has been placed on the front page on adobe.dk
After Adobe's acquisition of Macromedia there has been very quite about "Macromedia" products Flash, Flex etc. But there has never been much about Coldfusion. I believe that after the Macromedia office in Denmark/Scandinavia closed, Adobe Denmark did not take over.
Hope that Adobe does not forget that there are users outside US.
I have mentioned it to Ben Forta and Tim Buntel aswell, but the more who knows about the problem the better :-)
I think to be taken seriously though operators such as ++ in the script version need to be implemented a.s.a.p. in a point release along with many other tags.
Standard is a way to keep the developer base large enough to be interesting, so we get plenty of new features for the price of an upgrade - including gateways which is a great step forward.
Adobe are literally punishing non-US users of their products.
Now to CF. The issue is that if I download CF, there isn't a specifically different version for the UK and the US market. These are the same "downloaded" product. I cannot see why I am paying extra.
Now on to the CF is "reasonable priced because it is now J2EE and look at the other players in the market". The roots of CF are as a competitor to ASP/.NET and PHP. With the move to Java (which was facilitated by another company in Scotland initially), suddenly it becomes this enterprise level solution. (I honestly believe it is the fastest way to develop J2EE apps and have no qualms about the US ent price).
This still fails to address the fact that the standard edition is NOT an enterprise class solution, and that it should be priced correctly (and uniformly) within the world market.
So does this mean CF is going to die? No. Will it grow within the UK/European market? Not really because small start-ups will discard it as a viable solution as it's price is not comparable with PHP/.NET (i.e. perceived free). We're trying to convert people here.
It's people like this http://www.smithproject.org/ who are producing an opensource CFML engine that have a fighting chance of growing the market at that level.
Adobe needs to come up with a pricing model/solution that makes CF "just" another option a hosting company can offer. Standard doesn't cut it. No sandboxing, cost is prohibitive (remember we are competing with PHP/.NET here).
I want to sell CF to new clients as a good solution with a fast turnaround. Tender processes at this level are along the lines of "Hi I'm looking for a developer to develop a system. Technology is PHP/MySQL".
Why the LAMP stack? Because it's perceived to be 'free' and it is EASY getting a hosting provider.
Adobe needs to look at the pricing model Railo (another stunning CFML engine) is using. It's the ol' price per website model. Adobe need to provide a hosting version of CF that provides sandboxing and allows the host to have a very small investment.
Not sit there trying to explain to some bloke who knows nothing about web technology but has found a host who provides PHP/MySQL for X per month why they should also consider CF oh and it's going to cost him the best part of another grand.
It's rather screwed up.
I can accept the whole Ent CF is worth it, but I can't accept the price differences (we're talking $1400+ difference).
I also feel that CF needs a hosting product that would allow easier "hosting market" penetration. This is the gap in the market Adobe does not appear to recognise.
Hell you could even take an idea that google maps is using when people want to put maps on their websites. A company would be able to provide CF hosting (with no charge from Adobe), however a key bought from Adobe would enable CF to run for a specific site. Certain keys would enable 'more' features. In effect a subscription service.
Have an incentive programme where a percentage of key revenue is returned to the host.
Keep the cost in line with other solutions (we're talking say $20-50 dollars per year) and you may make it extremely attractive.
Remember though that this is all about shared hosting environments and these are your low-end, low-performance environments with say 100-200 sites (or more) per server where the enterprise features are not important( exchange etc).
Remember that there is no initial charge to the hosting company.
Hell if Adobe are worried they could even invoice the hosting company after one year for the cost of the hosting product and offset any key revenue against the bill. This would 'encourage' the hosting company to sell CF solutions.
All I'm asking is that Adobe look at companies like Railo and where they see a hole in the market. Adobe CF is the gold standard, but they really really need to look back at the roots of CF and understand that standard CF has to compete with PHP and .Net in the Small to Medium Business market and it doesn't feel like it is winning.
Apologies if I rambled. I've been writing intermittently over the day so rather a long post which started off as a small response.
Enterprise New License $ 10,003 (ouch!) Upgrade from CF7 $ 4,641 (ouch!) 24 month new subscription $ 4,125 (ouch!) 12 month subscription renewal $ 2,062 (ouch!)
Standard New License $ 1,787 Upgrade from CF7 $ 893 24 month new subscription $ 715 12 month subscription renewal $ 358
All prices are AUD inc GST (GST is our magical rip-off tax like the UK's VAT - but only 10% - I understand VAT is much higher). If you do not already have a subscription you cannot buy a renewal.
Most of these points have been covered elsewhere, but to catch the highlights:
1. We're looking at a 2-3year time horizon for the pricing. I'm not sure what it'd cost to hedge projected sales for 3 years at a 1.90 exchange rate, but it might be more expensive than you think.
2. CF competes on features. Problems selling to a client CF over PHP? No problem. Price what it'd cost you to develop the project in both and then give them the shared or dedicated hosting cost difference and let them decide. If they need sharedhosting, it's only going to be a few pounds a month more than an PHP cheapie for a similar quality host. If dedicated, additional cost is the Standard license for the box. I don't know how much you charge an hour, but if you have a client who needs a dedicated server and you're not going to save them 15-20 hours of development time on the project (to cover the cost of the license), you SHOULD be developing the app in PHP. Only reason to pay for a product is if it saves you or the client money. I am always happy to provide PHP and C# quotes on my projects. To date, other than for trivial scripts ("I need a screen scarper that . . .") it has always made sense for us to develop and deploy on CF and while I don't like spending money, even another few hundred bucks on the price of standard wouldn't have affected the math for my projects to date.
3. Nothing wrong with an SPLA style approach with a monthly fee instead of upfront or a pay per site model if Adobe goes there. I use SPLA from Microsoft for SQL server and my Win boxes and as a small hosting provider appreciate it.
2) I agree. I use CF because the ROI is there. I'm not talking about us, the converted, I'm talking about Joe Schmo. As a developer I can be more productive in CF. I know that, you know that
3) It's really a no brainer in the end.