Hi Intel Team/Community,
I stopped by the forums yesterday evening after a short development hiatus to discover that the XDK and the Cloud build system are being deprecated and set to be retired next month! I have been an incredibly satisfied user of the XDK even before Intel acquired it from AppMobi many years ago and wanted to reach out to understand the future of the tool. I have been very grateful for the improvements and continued customer support that the Intel team (including the members of AppMobi that followed their product) gave to the XDK (including the new GUI, crosswalk, better build logs, transitioning to the more standard cordova system, constant updates, etc.) I had hoped that the transition from a smaller company like AppMobi to Intel would mean a longer life/support for the great toolchain (and I am quite pleased with the support and attention it got), but sad that this is coming to an end.
For me, the XDK was a great way to centralize by development of the interface of my app (with quick view testing) and of native multi-platform plugins that I could use to integrate more advanced features and build for multiple platforms with one click. There was a perfect blend with the simplicity of a GUI tool for common tasks, and the ability to customize everything else with powerful configuration files and options. It served as a wonderful GUI layer to manage icons, splash screens, certificates, and other build assets/code on top of what could be done with the Cordova CLI and other tools that we are being suggested to move toward setting up. It was also tremendous that this could be backed up and retrieved by loading the XDK from any computer and not having to worry about a full toolchain setup. Probably the biggest loss here is the marvelous ability to build for any platform (including iOS) from any computer without a complicated setup of the toolchains and configuration of compilers and environments.
I wanted to ask, on behalf of everyone, what the future holds...
- Is there a possibility that the XDK GUI, native testing app, and the setup for the cloud build tools could be open-sourced? I would personally consider running a community instance of the cloud build servers (after a brief discussion with Intel about the historical associated costs/usage, etc.) I would love to hear the thoughts of the community on this as well.
- What do you see as the closest replacement for the XDK (development GUI, cloud builds delivered OTA to devices, and native testing app)?
- Other than the focus on IoT and the costs associated with running/supporting the XDK, what is the reason for this decision to quickly remove the XDK? Was there waning usage, technical complexities, and/or something else? Will there be any extra communications/suggestions to users who may have not been as lucky as me to stop by the forums/docs since March and may not be aware of the sad news that is happening so fast?
- Is there an opporitunity for someone to takeover the service entirely (code, tools, etc), including the user database/forums such that they could continue to support the service for the community? Perhaps a crowdfund campaign if we could get a reasonable estimate of a target that Intel would like for this.
I appreciate everyone who takes the time to read this and really welcome input from the community that still very actively uses the XDK. I look forward to learning more about the situation from the Intel team and what the future of the XDK will be. I also want to thank Paul and the rest of the Intel team in advance for being so responsive to users and for advocating for this transition to be as smooth as possible for everyone. My goal here is just to understand the reasons for these changes and explore all the ways this community of developers may continue to use the great tools that were created (that appear to be unrivaled by any other offering).
Thank you very much!
P.S. I have read the, now outdated https://software.intel.com/en-us/forums/intel-xdk/topic/700951 which says that some supporting components of the XDK will not be open-sourced, though this was before the sad news that the cloud build tools are also a casualty of this change in focus. I would really like to learn more about why the cloud build tools do not fit into the future IoT plans and what the complexities that resulted in this change were.