![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Last post about TouchBook
31 August 2009 07:05 pmI responded thusly in mail to "why did you return it?":
Hi,
You pretty much have the whole story. In brief from either of two positions:
As a consumer, the device is not quite ready for use. Although the software is developing rapidly, the hardware (particularly that revision) has some manufacturing flaws (display, USB, power, touchpad) and is not fully useful yet. So, that's a solid argument for a polite quick return.
As a possible dev or QA tester, the platform has a lot of great potential and is really exciting. The tools are not well developed and it's frustrating to do even simple troubleshooting. Also, as gently as possible, there is a contrast between community development (Mozilla, Debian, SugarLabs) and helping a company develop it's product for free ... and I paid full price for the product, not to enter the developer's program. In that situation, I'd have given you money for access and possibly gotten a discount on hardware (as per Apple or Microsoft).
On a related note I am uncertain of working with a port of Angstrom instead of actually working with a large upstream distribution. You really must have a large active developer community to drive sales and development. You'll get a lot more positive attention and sales once the thing boots Ubuntu or at least Debian, and I think your custom packages and drivers will have to be integrated into a mainstream kernel/distro to be genuinely useful. IRC traffic and forum posts back this up as of the very few people with TouchBooks a few have already bought additional SDcards and are installing other distros to work on the drivers. I am pretty solid on all of this because of my experiences with OLPC. OLPC went with a highly customized RedHat variant because RedHat was paying for it, and now they are trying to port their whole system to (normal) Fedora Core in the midst of a hardware revision. It's a nightmare to do and worse to have to do later. I think you'd be better off shipping BeagleDebian or possibly even BeagleUbuntu . Of course there are performance and engineering reasons that complicate the matter. And Android looms ... somewhat ominously.
Thanks for reading and gook luck in future endeavours,
[adric]
Hi,
You pretty much have the whole story. In brief from either of two positions:
As a consumer, the device is not quite ready for use. Although the software is developing rapidly, the hardware (particularly that revision) has some manufacturing flaws (display, USB, power, touchpad) and is not fully useful yet. So, that's a solid argument for a polite quick return.
As a possible dev or QA tester, the platform has a lot of great potential and is really exciting. The tools are not well developed and it's frustrating to do even simple troubleshooting. Also, as gently as possible, there is a contrast between community development (Mozilla, Debian, SugarLabs) and helping a company develop it's product for free ... and I paid full price for the product, not to enter the developer's program. In that situation, I'd have given you money for access and possibly gotten a discount on hardware (as per Apple or Microsoft).
On a related note I am uncertain of working with a port of Angstrom instead of actually working with a large upstream distribution. You really must have a large active developer community to drive sales and development. You'll get a lot more positive attention and sales once the thing boots Ubuntu or at least Debian, and I think your custom packages and drivers will have to be integrated into a mainstream kernel/distro to be genuinely useful. IRC traffic and forum posts back this up as of the very few people with TouchBooks a few have already bought additional SDcards and are installing other distros to work on the drivers. I am pretty solid on all of this because of my experiences with OLPC. OLPC went with a highly customized RedHat variant because RedHat was paying for it, and now they are trying to port their whole system to (normal) Fedora Core in the midst of a hardware revision. It's a nightmare to do and worse to have to do later. I think you'd be better off shipping BeagleDebian or possibly even BeagleUbuntu . Of course there are performance and engineering reasons that complicate the matter. And Android looms ... somewhat ominously.
Thanks for reading and gook luck in future endeavours,
[adric]