After Surface, what is left for OEM’s ?
So now that Microsoft is going to offer their own Windows 8 tablets , aka. Surface, what is left for OEM partners to develop ?
Don’t copy !
I don’t think just trying to copy Surface will be enough. There needs to be some “thinking outside the box” on this one. So what is left ? Plenty if you have imagination. So let’s give it a try.
A little imagination, a developers ideas
Now I am a bit out of the mainstream when it comes to software development, so my concepts may seem out of the ordinary, but at this point anything would be worth a try. I am still not crazy about the idea of the Windows RT version of Widows, since two different CPU platforms create problems because they just don’t work exactly the same. ARM and Intel (x86) are so divurgent in core design. One was built for a minimal CPU command set and the other for a more high level command set. Second, ARM is not backward compatible with previous versions of Windows and for that matter neither is Metro. I have always strongly encouraged backward compatibility, so let’s stick with the Intel version of Windows 8 for now.
Metro is yet to be fully tested in the real world, meaning it looks great in the preview versions, but few are using it full time as their core software environment. Sure some who have been experimenting with Windows 8 may be using it regularly, but are they doing their everyday work with 100% Metro apps, rather than a mixture of Metro apps with desktop apps ? There is no doubt that Metro is here to stay, but since Microsoft is putting all their emphasis on Metro, one area left open to OEM’s and their software developers is the desktop. So a combination of development of OEM apps for their tablets using Metro and the Desktop may offer some more choices.
Rather than spout theory about making a Windows tablet, I will stick with what I do know. I currently have an ExoPC Windows 7 Tablet. I really like it, but it has two weaknesses. First it is large and is heavy. It is 2.09 pounds. I think that it is too heavy. Now the Windows 8 Pro Surface tablet will be about 2 pounds from what I have read, so it won’t be light weight by any means. This is the first opening for OEM’s, weight. The second is that it only has a single core Atom CPU which is not a powerhouse, so anything more powerful would decrease battery life significantly, but the Atom isn’t a great processor for much of todays software. Yet, I like the Atom CPU’s since they continue to get better and they have enough power if used correctly (I will get to than in a moment).
Let’s start with the tablet design first.
So I’ll throw out my idea of the perfect tablet. I think most people will have multiple computers, so the idea for this tablet is not to be the “something for everyone”, but simply for something different which will fill a nitch that I doubt Microsoft will even consider. The two biggest concerns I have are price and mobility (meaning small and light). So let’s start with the size of the tablet. I really like the 7 inch form factor. Now a 7 inch screen is kind of small and many will simply view that size only as an ereader. So let’s start with an 7 to 8 inch screen, but find a way to have a minimal edge around it, so it is close to the 7 inch form factor as possible. I like the 16:9 aspect ration Surface is using, so lets also use that with a 1280 x 720 resolution. Cheaper versions of this tablet can use a 1068 x 600 resolution. Of course manufacturers would use a resolution which allows them to use stock parts readily available, but close to these. I will define two versions of this tablet concept, one for as low a price possible device and the other will more optimal features.
The cheaper version will use 1 gig ram and the better one 2 gig, which is plenty for a low cost device. There are a lot of different Atom CPU’s available, so I won’t mention a specific model, but the low priced version could use a single core Atom CPU and the better one a dual core Atom CPU. The purpose of the cheaper version is so that schools could afford to buy lots of them and also businesses could have a cheap alternative when they want a large quantity and cost is critical.
So let’s look at our two tablet specs so far:
- 7 inch display 1280 x 720 resolution
- 2 gig RAM
- 32 or 64 gig SSD
- two USB ports (2.0 or 3.0)
- HDMI mini port
- SD slot
- speaker jack
- 7 inch display 1068 x 600 resolution
- 1 gig RAM
- 16 or 32 gig SSD
- WIFI or non-WIFI
- one USB port
- HDMI mini port
- mini SD slot
- speaker jack
The price point for the devices would range from $150 to $200 for the cheaper model and $250 to $350 for the better model. Do not under estimate the potential for a low priced Windows 8 (x86) tablet.
Now I have an idea for a better business tablet and the specs could be a lot better, large displays, etc, but with one big difference. Replaceable parts ! Yes, design some kind of standard CPU slide in module, RAM slide in module, SSD slide in module, WIFI slide in module and even a module for a set of ports. Modules could be slid into the device by a non-technical person too. Just slide in and go. Even the batteries would be replaceable. This would solve a big problem with such computers and that is rather than be throw away once they reach their peek, they would be upgradeable easily by end users. This would increase the lifespan of the tablets, but would open up extra markets for addons by OEM’s. You would think by now someone would have come up with the idea. The BIOS on the tablet should be able to automatically recognize the change in parts and repsond accordingly. No technical knowledge should be required.
The software matters.
One of the biggest complaints people have with OEM products is the bloatware they often add to their computers. Yes, OEM developed software does offer some perks to customers which can provide OEM’s options to make their computers different, but if the software is bloated, then it makes no sense and also it makes a lower end, low cost tablet look bad. The solution is to build better apps with a smaller footprint. Herb Sutter (Microsoft C++ expert) in his talk, “Why C++ ?” discusses the value of going back to native coding (see: http://channel9.msdn.com/posts/C-and-Beyond-2011-Herb-Sutter-Why-C ). Native code apps an have an amazingly small footprint.
I will go one better !
If you have checked out my web site, you know that I work with Powerbasic compilers, rather than C++. They are native code compilers too. One of the things that I have noticed about software is that Object Oriented Programming has become the mainstay for most development, but I strongly feel that procedural style native coding produces faster, smaller and more reliable software than does using managed languages. To appreciate this, I was having a conversation with a new friend who years ago used to do software development. He ventured into different work in recent years, but when we started discussing programming he was interested in how I wrote my software. I explained how I use PowerBasic and that I built a GUI engine for it and started listing the many features in this GUI framework of mine. Now when he worked with software he used OOP and I though don’t. I then asked him, how big do you think the runtimes (DLL’s) are for my GUI framework ? I then told him the main runtime DLL is 700 KB. He quickly attempted to correct me and said “you mean 7 meg” right ? No, I said, it was 700 KB and the entire set of runtime DLL’s come to about 1 megabyte total. What was interesting was his response. He basically said “that is impossible” !
The point is that using a native code compiler (like C++ or PowerBasic) and using procedural style coding (a little OOP is OK, but the main code should be procedural to produce amazingly small footprint apps) can produce apps so small and fast then even on a inexpensive tablet, you can build desktop applications which can put past software to shame.
Also for years, Windows software was designed to look like everyone elses and now with Metro all of than is being thrown away for something new. Personally, I find that you can build desktop applications which can look totally different, yet will work on Windows XP to Windows 8. Build your applications to blend well with Metro (Metro style) or simply be totally different.
An OEM with a little effort could easil build a suite of apps which make their tablets different, which can fit easily on less than 1 gigabyte of space and use minimal memory and which runs fast on even a lowly Atom CPU. The software experience is what matters.
Have to mention EZGUI.
For anyone looking for a way to build such amazingly fast and small applications using native coding, I have to mention PowerBasic (see: http://powerbasic.com ) and my own GUI engine, EZGUI 5.0 (see: http://cwsof.com ). Powerbasic produces extremely fast and small native coded apps. EZGUI provides the advanced feature set not found in Powerbasic for build very complex graphic oriented applications. EZGUI has a built in drag and drop engine, so you can build WYSIWYG style applications. It supports the majority of Windows controls, but has many ways to customize such controls so you can build totally unique UI’s. It has a number of its own built custom controls such as Masked Edit control, Shape/Hot Spot/SplitterBar control, Files Listbox control, Property Listbox control, Canvas control, MCI control and glCanvas control. The Canvas control has a strong support graphics engine include a proprietary 2D Sprite engine. The glCanvas control is a hybrid control built upon the Canvas control, but adds an OpenGL based 3D scripting language.
Because EZGUI is a shareable DLL library, you can build a suite of apps which all share the same runtime and which are amazingly small. EZGUI itself was written using Powerbasic which is why it is so small and fast. The nice thing about EZGUI is that you can write apps which will run on Windows 95 to Windows 8 (yes, it has been tested on Windows 8).
So there is something left for OEM’s !
I am just provide some examples of things I would personally do in making a Windows 8 tablet. OEM’s you have many more choices than this. Microsofts Surface looks great, but people will pay a price for these tablets. Some suggest the Windows 8 Pro Surface tablet will be at least $1000. That leaves a lot of room left for OEM’s to build less expensive tablets for specific nitches. I really like the Acer Netbook I for for my wife. For $250 it was a real deal, even with a nice little dual core Atom CPU. I like the small size and low weight. So OEM’s find your nitch and build for it.
Date: June 22, 2012