Are RIM and Microsoft too secretive?

When you're a platform, how transparent do you need to be?

Last week we spent time talking to both RIM and Microsoft about their next mobile platforms. They're taking a remarkably similar approach to the future technically speaking, with RIM building on QNX with native code, C++, OpenGL, HTML5, Flash and Android and Microsoft building on the Windows kernel with native code, C# and C++, XAML, HTML5 and its own application packaging system. They both are emphasising a new design language that's radically different from what you see on their systems today: Microsoft has Metro, RIM has Cascades (which for those with a sense of irony is built on Qt, the system Nokia abandoned to switch to Windows Phone). They're both asking developers to make a major transition. Microsoft wants developers to give up Win32 and the desktop, RIM wants them to switch away from Java - and for both HTML5 is a big part of the switch, although there are plenty of native code options.

Often they're wooing the same developers; both companies have recently shown us how fast and fluid Cut The Rope is on their platform and they've both commented on how concerned ZeptoLabs was that its physics engine was fast enough. On the PlayBook it runs on the Marmalade engine and is a testament to the performance RIM gets out of the CPU and GPU. On Windows it runs in IE9 and IE10 and the ZeptoLabs team were impressed by the hardware-accelerated performance of the physics engine to the point of suggesting that it was maybe a little too fast.

One other thing that RIM and Microsoft have in common. Sometimes it can feel like pulling teeth to get information out of them.

RIM has promised several times to be more transparent but it's a culture shift for the company. It's only a few years since RIM started its first blog and for most of its history it's been working with carriers who didn't want to reveal which handsets they were going to sell and when or enterprises who cared more about security and support and predictability than when new handsets would be out.

When is the PlayBook OS 2.0 coming? This month. When this month? We're not giving the date yet. What about BlackBerry 10? Later this year. If you're a key partner. I'm sure RIM has told you the date but if you're a small developer wanting to do some marketing around the launch to push your PlayBook app, you'd probably like to have some warning about which day to do that on.

BlackBerry DevCon Europe last week was RIM's first European developer event and its largest developer event ever. It marked the first official appearance of new CEO Thorsten Heins (although he wasn't talking to the press). He chose a developer event for that to show RIM's commitment to developers "without you the BlackBerry experience would not be complete" and told the audience that he was listening. But he wasn't giving a firm date to them either.

Also last week, Microsoft filled in the details about Windows on ARM. Steven Sinofsky's near nine thousand word blog post covered a lot of ground and explained very clearly why you won't be able to put WOA onto an ARM tablet you just have lying around (they're all different). This is something developers have been asking about for a while and if Microsoft wanted to make sure everyone knew that the only way to be on every Windows device was to create a Metro app, it might have been useful to say that explicitly last September.

Of course Microsoft might not have been quite decided on that last September. Or they might have though that saying, as Steven Sinofsky did at the Financial Analysts Meeting that week, "We've been very clear since the very first CES demos and forward that the ARM product won't run any X86 applications" and "those applications aren't written to be really great in the face of limited battery constraints" and letting them realise for themselves quite how much work it would be to build app just for the ARM desktop (without the help that WinRT gives Metro apps to save battery) would steer developers away from the idea.

Microsoft developers have been twitchy ever since Bob Muglia spoke out candidly about how Silverlight wasn't the long-term future of Windows shortly before he left Microsoft, but there wasn't much Microsoft could say without announcing WinRT and Metro-style apps there and then. Could they have done that? The Windows team was certainly working on the ARM version of Windows long before that. But announcing things too early can cause as much confusion as making people wait.

Announcing Windows on ARM before there were any Android ARM tablets might have shown vision, but it would also have Osborned every Atom tablet PC that came out in the last two years. Given how few of these have been other than disappointing that might not seem like a bad thing (and it might have given Intel the kick in the pants it's been needing on low power processors that really are low power for far too long) but it wouldn't have made Microsoft's partners very happy.

If Microsoft had announced, say, the bare bones of WinRT earlier last year, when it didn't have details of everything that would be in the framework, there would have been questions it couldn't answer and it would have looked like its plans weren't fully baked. Every time I've been impatient for Microsoft to have told us more earlier, I've realised later that if they'd told us when we first asked there would have been questions they couldn't answer and we'd still have been frustrated.

Is knowing that Mozilla plans to bring Firefox to WOA as a Metro app useful to know when the developer team hasn't even decided what strategy to use to port the majority of the code to ARM? ("the assumption is that we'll be able to run as a Medium integrity app so we can access all of the win32 Firefox Gecko libraries avoiding a port to the new WinRT API for the bulk of our code" says the planning document, which could be quite an assumption.)

It's a difficult balance to strike, between disappointing people with a plan that's not ready and frustrating them by keeping everything inside the kimono until you're ready to have your big reveal - or indeed flooding people with so information that they can't take it all in. Given how many people commenting on the blog were asking questions that would have been answered by reading the original post, more information isn't always better.

RIM won't want to give a date for the PlayBook 2 update until it's certain it will be out on that date. Microsoft won't give a date for launching Windows 8 until it sees how much work there is to do based on the telemetry and feedback from the Consumer Preview. And we'll always want to know more, sooner, so we make decisions about what to do next with the most information possible. You can't keep everyone happy, whatever you say and whenever you say it.

Mary Branscombe