RunRev CEO Kevin Miller on revMobile and Apple's new iPhone SDK Agreement

 

revMobile and Apple's new iPhone SDK Agreement
As you may know, on the 8th of April, Apple changed the iPhone Developer Program License Agreement, which impacts our development of revMobile. The new agreement added a clause which required that applications be originally written in Objective-C, C++ or JavaScript. As revMobile applications are originally written in revTalk, not in one of these languages, their policy changes effectively prohibit revMobile on the iPhone/iPad. The new clause also prohibits frameworks and compatibility layers, which also describes revMobile in its present form.
This change took us, and thousands of other iPhone developers, by surprise. We greatly value our customers and we know how much you depend on us to provide an outstanding and reliable service. Since this announcement, we have been working around the clock to develop a strategy for that will allow us to continue to provide you with the service that you expect and deserve. The surprise from Apple’s announcement is heightened by the fact that Apple’s policy seems to be at odds with recent comments from Steve Jobs himself.
In response to a question in a recent Apple shareholder’s meeting, Jobs said that a HyperCard-like product for the iPad would be a good idea “though someone would have to build it”. Many blog posts suggested that Apple's intention with this change was to ban cross-compilers and ensure high quality, fully native iPhone apps. Jobs all but confirmed  [http://techcrunch.com/2010/04/10/steve-jobs-responds-to-iphone-sdk-complaints-intermediate-layers-produce-sub-standard-apps/] the analysis on John Gruber's site [http://daringfireball.net/2010/04/why_apple_changed_section_331]. The Mac platform in particular has a long legacy surrounding HyperCard and as virtually the last man standing in this space. We believed we were in a position to offer Apple something they wanted.
It makes perfect sense to have a high quality, rapid application development system available for the iPhone and iPad. Our customers agree. revMobile, even its pre-alpha form, received some of the strongest and most positive feedback from our customers that we've ever seen. We are aware of numerous exciting applications being developed even at this early stage. For example, one of our customers, EuroTalk interactive, have government funding to deploy thousands of iPads to schools in Malawi [http://www.scotland.gov.uk/News/Releases/2010/03/18130728] in a pilot project that will expand to tens of thousands of users. They intend to deploy hundreds of small teaching applications to iPads. It makes no sense for those applications to be developed in Objective-C or JavaScript, and EuroTalk will be looking seriously at other tablet manufacturers if they are unable to use revMobile.
Our Pitch to Apple
In order to support our active and growing revMobile customer base, we submitted an in-depth proposal to Apple that we create an iPhone-only product that uses native Cocoa objects, supports 100% of their API, works perfectly with multitasking and battery life, but uses a variant of the revTalk language to use these objects and APIs, and then translates those into native code. While a significant engineering departure for us from the current revMobile path, this solution would have resulted in perfect-quality iPhone-only applications impossible to distinguish from native applications. It would have been impossible to tell these applications apart from native iPhone applications because they would be native applications. As native applications running directly without a compatibility layer, there would have been no battery life issues, multitasking and iAds would work perfectly, and new APIs would be supported as they came out. In other words, we set out to offer Apple what they wanted by raising our game in response to their stricter requirements, while dropping the other mobile platforms we originally intended to support.
Apple's platform would have benefited immeasurably from a high-level, iPhone/iPad-only development tool that is 10x more productive than Objective-C and honors the HyperCard legacy still present on their Mac platform today. So strong is this platform legacy that millions of applications are shipped using Rev technology, and one of the investors in RunRev was himself a co-founder of Apple. With our proposed solution, tens of thousands of our customers would have been able to continue to use revMobile to develop these exciting applications and solutions in much the same way as was possible with our existing Rev product line, albeit at the cost of cross-device compatibly. There would have been no risk to Apple in any such agreement as we would comply with their terms in their entirety. Apple approves applications all the time for their store; ours simply would have been one more.
Steve Jobs has now rejected our proposal and made it clear that he has no interest in having revMobile available on the iPhone or iPad in any form. 
Obviously, this is a huge disappointment for us and our customers. The analysis we read both on Apple.com [http://www.apple.com/hotnews/thoughts-on-flash/] and on many blogs [http://developer.appcelerator.com/blog/] suggest that their main target might be Flash. Clearly this is not the case. We know that like any other company, Apple has the right to make decisions they deem necessary to maintain their competitive edge and protect shareholder value. However, had they made their decision public earlier in the evolution of the platform, it would have saved us and thousands of other developers what must add up to millions of dollars of wasted engineering budget. Their strategy makes it impossible for us to provide revMobile at the same feature set and service level currently in use by our loyal customers.
The Path Forward
Some of our customers have suggested that we continue to develop the version of revMobile that outputs entirely native code (as detailed above). However such a solution--even though it would create perfect applications--would be in violation of Apple's agreement, which states that code must originally be developed in one of their approved languages. As such, we cannot risk hundreds of thousands of dollars of further engineering budget to create a solution that does not guarantee Apple acceptance to the app store.
We will continue to offer revMobile as an excellent prototyping tool for iPhone and iPad applications. A large proportion of development time is spent during the design phase of any application and revMobile offers an unbeatable solution for that. We will also continue to offer revMobile to those who are members of the Apple Enterprise developer program (including educators who are part of the program) who want to deploy applications to their own devices and in-house.
However we cannot make any guarantee, now or in the future, that revMobile will be in compliance with Apple licensing provisions. It is impossible to make such guarantees as Apple reserve the right to change their terms at any point, and have clearly demonstrated their willingness to exercise this right to make sweeping changes without notice. It is up to individual customers to satisfy themselves of Apple requirements.
While we will be shipping revMobile as a product, we reluctantly have no choice but to scale back development on it. We can no longer create the native interface objects layer, but rather invite our community to develop a set of emulated objects. We also will no longer build the native-code compiler, as this is a vast engineering project that’s purpose was to facilitate submission of applications to the app store.
Given these business changes, we are pushing back our RunRevLive.10 conference to April next year, and are focusing on developing significant and exciting enhancements to the Rev platform. We have many exciting projects in the pipeline including an overhaul to our Unicode support, enhanced text display, improvements to the revWeb plugin, the rollout of revServer and much more. We are turning our mobile sights to the Android platform, and will unveil an aggressive strategy for supporting Android development projects. We will also roll out the updated Rev platform roadmap for our revEnterprise customers shortly.
The industry we work in moves quickly and we have always worked tirelessly to make sure the Rev platform enables you to take advantage of new and innovative technologies. That change has been particularly quick over the past month. But there are many new opportunities out there and we look forward to finding ways to enable you to take advantage of them. And all these changes notwithstanding, we still believe that revMobile remains the most sophisticated iPhone/iPad prototyping and Enterprise development tool in the world. We’ll be posting more updates and lessons for it shortly.
We have always worked hard to be a trusted partner for the many thousands of customers and businesses that use Rev. We’re going to continue to take care of everyone in our community. We will be redoubling our efforts to move forward in a new direction that enables you in new ways. I hope you’ll stay with us through this transition. Our customers are the reason we are here and we thank you sincerely for the unwavering commitment and support you’ve shown us.
Keep your eyes on exciting new developments from us in the coming months.
Kevin Miller
RunRev CEO, 10th May 2010

As you may know, on the 8th of April, Apple changed the iPhone Developer Program License Agreement, which impacts our development of revMobile. The new agreement added a clause which required that applications be originally written in Objective-C, C++ or JavaScript. As revMobile applications are originally written in revTalk, not in one of these languages, their policy changes effectively prohibit revMobile on the iPhone/iPad. The new clause also prohibits frameworks and compatibility layers, which also describes revMobile in its present form.

This change took us, and thousands of other iPhone developers, by surprise. We greatly value our customers and we know how much you depend on us to provide an outstanding and reliable service. Since this announcement, we have been working around the clock to develop a strategy that will allow us to continue to provide you with the service that you expect and deserve. The surprise from Apple’s announcement is heightened by the fact that Apple’s policy seems to be at odds with recent comments from Steve Jobs himself.

In response to a question in a recent Apple shareholder’s meeting, Jobs said that a HyperCard-like product for the iPad would be a good idea “though someone would have to build it”. Many blog posts suggested that Apple's intention with this change was to ban cross-compilers and ensure high quality, fully native iPhone apps. Jobs all but confirmed the analysis on John Gruber's site. The Mac platform in particular has a long legacy surrounding HyperCard and as virtually the last man standing in this space. We believed we were in a position to offer Apple something they wanted.

It makes perfect sense to have a high quality, rapid application development system available for the iPhone and iPad. Our customers agree. revMobile, even its pre-alpha form, received some of the strongest and most positive feedback from our customers that we've ever seen. We are aware of numerous exciting applications being developed even at this early stage. For example, one of our customers, EuroTalk interactive, have government funding to deploy thousands of iPads to schools in Malawi in a pilot project that will expand to tens of thousands of users. They intend to deploy hundreds of small teaching applications to iPads. It makes no sense for those applications to be developed in Objective-C or JavaScript, and EuroTalk will be looking seriously at other tablet manufacturers if they are unable to use revMobile.


Our Pitch to Apple

In order to support our active and growing revMobile customer base, we submitted an in-depth proposal to Apple that we create an iPhone-only product that uses native Cocoa objects, supports 100% of their API, works perfectly with multitasking and battery life, but uses a variant of the revTalk language to use these objects and APIs, and then translates those into native code. While a significant engineering departure for us from the current revMobile path, this solution would have resulted in perfect-quality iPhone-only applications impossible to distinguish from native applications. It would have been impossible to tell these applications apart from native iPhone applications because they would be native applications. As native applications running directly without a compatibility layer, there would have been no battery life issues, multitasking and iAds would work perfectly, and new APIs would be supported as they came out. In other words, we set out to offer Apple what they wanted by raising our game in response to their stricter requirements, while dropping the other mobile platforms we originally intended to support.

Apple's platform would have benefited immeasurably from a high-level, iPhone/iPad-only development tool that is 10x more productive than Objective-C and honors the HyperCard legacy still present on their Mac platform today. So strong is this platform legacy that millions of applications are shipped using Rev technology, and one of the investors in RunRev was himself a co-founder of Apple. With our proposed solution, tens of thousands of our customers would have been able to continue to use revMobile to develop these exciting applications and solutions in much the same way as was possible with our existing Rev product line, albeit at the cost of cross-device compatibly. There would have been no risk to Apple in any such agreement as we would comply with their terms in their entirety. Apple approves applications all the time for their store; ours simply would have been one more.

Steve Jobs has now rejected our proposal and made it clear that he has no interest in having revMobile available on the iPhone or iPad in any form. 

Obviously, this is a huge disappointment for us and our customers. The analysis we read both on Apple.com and on many blogs suggest that their main target might be Flash. Clearly this is not the case. We know that like any other company, Apple has the right to make decisions they deem necessary to maintain their competitive edge and protect shareholder value. However, had they made their decision public earlier in the evolution of the platform, it would have saved us and thousands of other developers what must add up to millions of dollars of wasted engineering budget. Their strategy makes it impossible for us to provide revMobile at the same feature set and service level currently in use by our loyal customers.


The Path Forward

Some of our customers have suggested that we continue to develop the version of revMobile that outputs entirely native code (as detailed above). However such a solution--even though it would create perfect applications--would be in violation of Apple's agreement, which states that code must originally be developed in one of their approved languages. As such, we cannot risk hundreds of thousands of dollars of further engineering budget to create a solution that does not guarantee Apple acceptance to the app store.

We will continue to offer revMobile as an excellent prototyping tool for iPhone and iPad applications. A large proportion of development time is spent during the design phase of any application and revMobile offers an unbeatable solution for that. We will also continue to offer revMobile to those who are members of the Apple Enterprise developer program (including educators who are part of the program) who want to deploy applications to their own devices and in-house.

However we cannot make any guarantee, now or in the future, that revMobile will be in compliance with Apple licensing provisions. It is impossible to make such guarantees as Apple reserve the right to change their terms at any point, and have clearly demonstrated their willingness to exercise this right to make sweeping changes without notice. It is up to individual customers to satisfy themselves of Apple requirements.

While we will be shipping revMobile as a product, we reluctantly have no choice but to scale back development on it. We can no longer create the native interface objects layer, but rather invite our community to develop a set of emulated objects. We also will no longer build the native-code compiler, as this is a vast engineering project that’s purpose was to facilitate submission of applications to the app store.

Given these business changes, we are pushing back our RunRevLive.10 conference to April next year, and are focusing on developing significant and exciting enhancements to the Rev platform. We have many exciting projects in the pipeline including an overhaul to our Unicode support, enhanced text display, improvements to the revWeb plugin, the rollout of revServer and much more. We are turning our mobile sights to the Android platform, and will unveil an aggressive strategy for supporting Android development projects. We will also roll out the updated Rev platform roadmap for our revEnterprise customers shortly.
 

The industry we work in moves quickly and we have always worked tirelessly to make sure the Rev platform enables you to take advantage of new and innovative technologies. That change has been particularly quick over the past month. But there are many new opportunities out there and we look forward to finding ways to enable you to take advantage of them. And all these changes notwithstanding, we still believe that revMobile remains the most sophisticated iPhone/iPad prototyping and Enterprise development tool in the world. We’ll be posting more updates and lessons for it shortly.

We have always worked hard to be a trusted partner for the many thousands of customers and businesses that use Rev. We’re going to continue to take care of everyone in our community. We will be redoubling our efforts to move forward in a new direction that enables you in new ways. I hope you’ll stay with us through this transition. Our customers are the reason we are here and we thank you sincerely for the unwavering commitment and support you’ve shown us.

Keep your eyes on exciting new developments from us in the coming months.

Kevin Miller

RunRev CEO, 10th May 2010