OpenMoko
Monday, September 17th 2007 | Ismael Ghalimi
Looking for a way to build the Redux Model 1 around the iPhone or iPod Touch allowed us to appreciate the benefits of leveraging an existing platform instead of starting from scratch. Nevertheless, this can only work if the platform is open enough, which the iPhone or iPod Touch are not, making it very difficult to build what we really want. An alternative approach would be to start from an open platform, and the best we could find so far is OpenMoko, and the Neo 1973.
Building around the GTA02 release of the Neo 1973 would give us the following:
- Samsung S3C2442/400 CPU
- SMedia Glamo3362 Graphics Accelerator
- Global Locate Hammerhead PMB 2520 AGPS
- Wolfson Microelectronics WM8753 Audio CODEC
- National Semiconductor LM4857 Stereo Amplifier
- Two 3D Accelerometers
- 128MB SDRAM
- 256MB Flash
- USB 1.1
- Bluetooth 2.0
- 802.11 b/g
- 2.5G GSM (Quad Band, Voice, CSD, GPRS)
- MicroSD Slot
- Mini USB Port
- Vibrator
We would get all this for $300, including a superb 2.8" VGA (640 x 480) tpo TD028TTEC1 touchscreen that could be used as touchpad on our magnetic removable keyboard. This would essentially leave us $200 to pay for a bigger touchscreen and the proper enclosure, which is totally doable. The screen would actually be the easy part, for it would not need to be a multi-touch screen. The hard part would be to replace the video controller, which currently only supports a resolution of 640 x 480 pixels. Worst case scenario, we could use a 6" VGA display, which would give us twice the iPhone’s resolution, but SVGA (800 x 600) would make most websites display a lot better. At this point, we should consider SVGA as the desired option, and VGA as the fallback option.
On the plus side, using the OpenMoko platform would have many benefits beyond openness, such as support for any cellular network, embedded GPS, and total control over screen orientation. On the minus side, we would lose the iPhone’s polished user interface, transparent software updates over iTunes, and possibly the ability to play videos (Cf. OpenMoko Wiki), even though this might change with the GTA02 release, which will increase CPU speed from 200 MHz to 400 MHz. We should also expect OpenMoko’s power management to be sub-optimal for quite some time, even though software updates could improve things significantly in the future.
From a form factor standpoint, using the Neo 1973 as base platform should not have any major impact on our original design. Even though the Neo 1973 is a lot thicker than the iPhone (18.5 mm vs. 11.6 mm), this disassembly page clearly shows that reducing the device’s outside dimensions definitely was not a priority. Our form factor being a lot longer and wider, we can locate the battery next to the logic board, thereby reducing thickness down to 12 mm or 13 mm.
On the software side, the device would run what OpenMoko simply calls Mobile FOSS, Free and Open Source Software for mobile devices. The stack is primarily made of the Linux kernel, the GNU C library, the X window system, and the GTK+ toolkit, to name only a few. Of course, it would give us complete freedom to build whatever we want, and since all we need is a web browser, this should not be too difficult.
So let’s recap. Our Mini Tablet, code-named Redux Model 1, would be made of a standalone tablet, which could be connected to an optional keyboard that would attach to the tablet through a magnetic connection. This keyboard would itself embed additional components, such as a visual touchpad and a secondary battery. Here is a summary of each component’s features:
Redux Model 1 Tablet
- 6" Touchscreen SVGA Display (800 x 600)
- Dynamic Screen Orientation
- 128MB SDRAM
- 256MB Flash
- USB 1.1
- Bluetooth 2.0
- Wi-Fi 802.11 b/g
- 2.5G GSM (Quad Band, Voice, CSD, GPRS)
- Assisted GPS
- SD Card Slot
- 3.5 mm Stereo Headphone Minijack
- Mini USB Connector
- Magnetic Docking Connector (Landscape Mode Only)
- Very High Capacity Battery
- 7" tall, 5" wide and 0.5" thick
- Targeted Retail Price: $499
Redux Model 1 Keyboard
- Full Keyboard
- 2.5" Visual VGA Touchpad (640 x 480)
- Mini USB connector
- 7" tall, 5" wide and 0.5" thick
- Targeted Retail Price: $199
Next step: a conceptual block diagram…
Entry filed under: Office 2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
I was about to write a comment along the lines of “why bend the hardware to meet software? Especially when software is almost infinitely flexible?” when I read this post and see that some of my concerns are being covered.
The main thrust of my suggestion would be why not use Linux as a baseline? You can configure it to look like Mac OS X — Google for several KDE skins and “HOWTO” documents on mouse over enlargements and other eye candy.
The main reason though is not the look and feel but the flexibility offered with Free/Open Source Software (FOSS). I know I’m speaking to a largely sympathetic audience, so I’ll try to keep off my soapbox but…
1) You can configure the kernel, drivers etc. to meet exactly your device’s needs — build what you want and leave out the bloat associated with hardware support for things you’ll never use.
2) There’s a huge amount of software available, allowing you to tailor it to the audience’s needs.
3) You could load up Apache, MySQL, and a scripting language (Perl, Python, PHP or Ruby) and have a real application server. Why would you want to do that? Well, add in a Zoho / Thinkfree / OpenOffice.org and you then have an Office 2.0 server which you can carry around with you. The advantages of taking head office with you and only requiring a browser (well a network as well, obviously!) to give others access wherever you are. Plus you are keeping the files under your control… Quite a good sales vehicle for the concept…
4) Add in Ruby on Rails and you can configure database driven, professional looking, tailor made applications / presentations / demos with relative ease — almost on demand whilst you’re sitting in a train / airport lounge.
5) With ssh or scp, updating and backing up become simple and secure.
One final comment on the hardware… where’s the microphone input? Add that and some VOIP software, and you’ve got phone functionality. Even without the phone you could implement a simple “dictaphone” recorder or (and this would require more storage) voice to text.
Build that for 500 USD / 350 EUR or 250 GBP and I’d buy one…
Andrew,
We’re getting there…
-Ismael
I wrote on the Apple Forum and requested to be provided with and SDK and API to write a Skype-like phone for the iPod Touch. But Apple said it is not releasing this at present time.
What we need is Skype on iPod Touch. Imagine all the educational institutes having wireless access for their students, for whom it is freely accessible. If they had a Skype-like solution on this platform, there would be no need for paying carriers any more.
As it is I cannot understand why I should select a carrier to select a phone. Why is there no anti-trust regulation against this? I will select a device, and then chose a carrier which can service that device best!
I am ready to spend my time and energy to build such a solution on iPod Touch.
Any takers and ideas?
Has anyone actually tried making enquires to FIC about an alternately layed-out neo? Contact Sean Moss-Pultz. I’m sure they would appreciate your business more than Apple.
Trackback this post | Subscribe to the comments via RSS Feed
Leave a Comment