by RogueWarrior65 » Mon Aug 03, 2009 3:17 am
by RogueWarrior65
Mon Aug 03, 2009 3:17 am
Here's a systemic problem I see with, shall we say, the business of Linux embedded development. You're often on your own in uncharted waters. Now before I get flamed for this, hear me out. My background is not insignificant. I've got a Masters in Software Engineering and I've developed and sold products from the ground up. For the most part up until this past year I've been living in the software world. Now I need to do hardware. Back before the Earth's crust cooled, I did basic hardware development on 8048's with old-fashioned EEPROMs and the like. One of my Master's projects was a computer-controlled laser light show. But I digress. Nowadays, hardware is pretty damn sophisticated yet it seems like a chicken & egg situation in that unless you've done it before, there isn't a great one-stop-shopping way to learn it. A big exception to this is Microchip. IMHO, they made a very smart business decision to make learning how to use their products a priority. After all, if you know how to use their stuff, you're more likely to buy it in the future. The Linux SBC world doesn't seem to have this. There are some half-decent books but they generalize a lot. Many of the boards I see out there have the most basic documentation and little example code which often doesn't work right. For my current project, I've rejected a number of otherwise good SBCs because there was no support for them.
So, a modest proposal to any SBC manufacturer: IMHO, if you want people to buy your products and become an industry standard, make a major investment in training and support.
Let the flaming begin.
Here's a systemic problem I see with, shall we say, the business of Linux embedded development. You're often on your own in uncharted waters. Now before I get flamed for this, hear me out. My background is not insignificant. I've got a Masters in Software Engineering and I've developed and sold products from the ground up. For the most part up until this past year I've been living in the software world. Now I need to do hardware. Back before the Earth's crust cooled, I did basic hardware development on 8048's with old-fashioned EEPROMs and the like. One of my Master's projects was a computer-controlled laser light show. But I digress. Nowadays, hardware is pretty damn sophisticated yet it seems like a chicken & egg situation in that unless you've done it before, there isn't a great one-stop-shopping way to learn it. A big exception to this is Microchip. IMHO, they made a very smart business decision to make learning how to use their products a priority. After all, if you know how to use their stuff, you're more likely to buy it in the future. The Linux SBC world doesn't seem to have this. There are some half-decent books but they generalize a lot. Many of the boards I see out there have the most basic documentation and little example code which often doesn't work right. For my current project, I've rejected a number of otherwise good SBCs because there was no support for them.
So, a modest proposal to any SBC manufacturer: IMHO, if you want people to buy your products and become an industry standard, make a major investment in training and support.
Let the flaming begin.