MBS Berlin Developers Conference Keynote

Geoff Perlman, CEO of Xojo Inc gave his keynote address at the Berlin Xojo Developers Conference today.  The conference is hosted by Monkeybread Software.  In his hour long talk, he discussed the future of Xojo and, in particular, what’s scheduled for the rest of 2017.

In the past two years, Xojo releases have fixed over 1,000 bugs.  They’ve also add 200 major features.  This includes a new Language Reference, HiDPI support for Mac, Windows, and Web, Raspberry Pi support, 64-bit builds, and iOS additions.  Uploading in Xojo Cloud is now 400% faster and new data centers were added in New York, San Francisco, London, Amsterdam, Bangalore and Singapore.

Geoff spent some time talking about things coming up in 2017.  In Release 2, due out in July, 64-bit builds will no longer be beta status.  64-Bit debugging for Windows will be a reality.  Geoff said that it was almost available for R1 but the version of LLVM  for Windows they needed was too new.  XojoScript will be 64-bit as well.  The split and join string operations have been reconfigured to make them comparable in speed to the 32-bit versions.

For Windows, 64-bit builds will have app icons and version information.  Neither of these are available in R1 and earlier 64-bit builds for Windows.  Linux will now use GTK3 which will allow a 64-bit HTMLViewer.  A stretch goal might be HiDPI support and Geoff said that this might be later.

Later in the 2017, after R2, here are the goals:  64-bit builds will be the default in the IDE (32-bit will still be available).  The IDE itself will be 64-bit.

Also scheduled for 2017 is a new plugin format.  This new format will require Xojo Pro for creation.  It will still support libraries written in C/C++.  You can add images, sounds and other resources.

One of the more interesting things about the new plugin format is that it’s project-local.  Currently all plugins are global which means you cannot have multiple versions of the same plugin installed.  This new plugin format is project based, so it’s easier to handle different versions for different projects.  The new Xojo plugins are compiled into an intermediate LLVM format so there’s no need to ship classes with encrypted source code.

Even though the existing plugin format will be around for many years the new format is definitely the wave of the the future.  This will be the ONLY format supported for mobile.

The IDE is being redesigned.  Geoff admitted that the Navigator isn’t great for large projects and in some cases works against you.  The current Navigator is a custom canvas control and later this year it will be replaced with a standard ListBox.  This will make the next step easier and it will make tabs work as a hybrid between the old Real Studio and the newer Xojo IDE.  Geoff showed some screen mockups and design document so it felt like they know exactly what they want, they just need the time to implement it.

Next up was Interops.  Currently working with Declares is challenging.  Differing versions of OS SDK versions are hard to deal with.  There is also no type conversion.  Interops promises to make all these issues go away and use nothing but Xojo data types.

Geoff showed an example of an SDK call in Swift for iOS.  Then he showed the ugly Xojo declares for it, which didn’t look like Xojo code at all, and then showed the Interops version.  The Interops version looked nearly identical to the Swift versions.  This is the technology used to develop Android.  Interops will be developed first for mobile and then MacOS.

Finally, Geoff talked about Android.  He said a beta of Android will be out by the end of 2017.  It will be using Interops to make development easier and will create native code and controls.  Apps will be compiled into machine code which should give Xojo Android apps comparable performance as Google’s own native apps.   Support will be for KitKat version 4.4 or later which is roughly 80% of the installed user base.

Geoff wrapped up his talked with a brief look back at the competitors to Xojo when the company was first started.  Apples MPW, Microsoft Visual Basic, Delphi, MetroWerks CodeWarrior, Sun’s Java, Symantec Think C were all the rage.  Some of those tools no longer exist.  Some of those tools have been sold multiple times with different strategies.  For over 20 years Xojo has stayed with the same company and continues to evolve and change to meet the needs of users.

10 thoughts on “MBS Berlin Developers Conference Keynote

  1. Thanks Bob for the update! As I’m trying to get up-to-speed using Xojo, sometimes the learning curve causes concerns about even choosing Xojo.
    Geoff’s comment about companies that have come and gone really struck home with me. I have been orphaned so many times. DBaseIII, Clipper, Foxpro, Visual FoxPro, VB6, Clarion, VO, PowerBASIC, and there are many others that I have forgot.
    I bought them all. Bought the books, shirts, hats and support over the last 27 years, all gone. Did enjoy many conferences, met some great coders, and some of the weirdos.
    Thanks again

    • WAAAAH! We are DOOMED!

      Obviously you got the “reverse Midas” Touch … 😉

  2. Did Geoff give any more details?

    Would one be able to do more in a Xojo coded plugin than allowed in a class now…. Like being able to show visually control state visual changes in the IDE depending on properties set in the inspector?

    About the new plug-informat.. Did Geoff say if one would one would need Xojo Pro to develop a written in Xojo plugin or just “compile” it for distribution/sale?

    I could envision a number of ways that could be done (without Xojo Pro not able to ‘compile’ and make the source unencripitable for that project type and/or lock the execution of even the debugger version to that machine or a specific lower Xojo license. The necessary information could be imbedded in a an encrypted – By Xojo Inc- specific type type project file freely allows editing the source code in the IDE only )

    That could help jumpstart usage and could eventually result in more plugins available… and might get some hobbyists/”Citizen Developers” (like me) who produce something useful that may want to distribute it, to upgrade to Xojo Pro.

    Just wondering…

    • I’ll take your points one at a time.

      “Would one be able to do more in a Xojo coded plugin than allowed in a class now”

      Yes, it will be more than what you can do with external plugins and with a namespace right now. We will be able to bundle resources (images, sounds, etc). Obviously we won’t know the final details until it’s actually out. So stay tuned for more information.

      “would need Xojo Pro to develop a written in Xojo plugin or just “compile” it for distribution/sale?”

      You will need Xojo Pro to developer the Xojo plugin. Any license will be able to use it.

      These Xojo plugins will create LLVM intermediate code. Which is compiled and not pure source code. So presumably that since it’s already ‘compiled’ it won’t need to be recompiled like current plugins are.

      The goal of this new format is to jumpstart the third party market. Without being able to play with the new format I can’t really give much of an opinion.

      • “You will need Xojo Pro to develop the Xojo plugin. Any license will be able to use it.”

        That is a shame… If I created a plugin that I thought might be salable it would be a justification to pay for a pro license to be able to distribute it.

        In theory I know I might be able too do that with the right idea as I have made enough with my listbox class to justify a Desktop license – but I had that long before I created that control…

        I have used Xojo for about 16 years so I am reasonably experienced with it, but I don’t need anything but desktop, and don’t make my living with Xojo …

        With the current price to maintain/renew a Xojo Pro license, it would be hard for me to justify a Pro license just so I MIGHT be able to develop a saleable plugin someday …

        But If I did develop one I sure as heck would get the Pro license!

        If Xojo inc could allow that safely with lower licenses (make sure that it can’t be distributed and only used in the IDE (not compiled apps) by ONLY that developer (locking it to a specific lower user license even for use in the IDE), it could help both sell pro licenses and jumpstart the plugin market even more.

        But I know it will likely never happen…

        Oh well my dream of supplementing social security a little with income from using Xojo when I retire in a few years is probably just a pipe dream!

  3. Can’t wait for the new xojo native plugin format will make using shared class and libs must easier across large multi app projects….or so I hope

  4. Thanks Bob, it’s great to hear these updates as always. The completion of 64 bit support is something to look forward to. And it’s good to hear that the new IDE will borrow some ideas from REAL Studio… because even though I can barely remember what the REAL Studio IDE looked like, I can definitely remember liking it!

  5. InterOps sounds like real progress, can’t wait to see it on Xojo. Most of my frustrating hours on xojo are spent with the UI and trying to extend the naturally limited (lowest common denominator) set that comes with a multplatform language. Easier access to ready made native UI elements on different platforms could be a life-saver.

  6. I’m exited about being able to make my own plugins. I didn’t have the time or the will to learn C.

Comments are closed.