It seems to me that there is a glaring hole in the GHI Electronics product line. Over the years I’ve used quite a few chipsets (EMX, it’s predecessor, usbizi 100 and 144, etc…). But if you want to produce something in reasonable quantities, your only choices are EMX and USBizi.
And, unfortunately, it’s a choice between a chip that has a ton of RAM, built-in ethernet, expensive and another one that barely has any RAM at all, no ethernet and is very, very reasonably priced.
There is a ton of space in between. Over the last couple of years, when looking for a solution, I had to take GHI out of the running because EMX was too expensive for my purposes and USBizi was too incapable.
Why do I say USBizi is incapable? Because with the available RAM (30-40k after everything is loaded) you can’t really do anything substantial (loading a 10k file from flash results in running out of memory), you spend all your time optimizing the living crap out of the code, thus defeating the purpose of rapid development with the .NET Microframework.
I think GHI needs a chip that has at least 250k of RAM (better yet 1mb), built-in ethernet support and be priced from $20 to $30 in quantities and $30 to $40 standalone.
What do you think?
I like the way GHI approach most aspects of their new product “ideas” like the way the Panda originally came about and the way the Fez Ant discussion went on (as well as Hydra and Fez Hacker etc etc). Ultimately they are the ones that need to set a business model in place so I can totally accept that they might seek input but it’s their destiny so I’d be totally fine if they developed cool stuff in isolation. It’s obvious they’ve identified a potential area they can explore here that people will buy, so tally ho, bring it on ! 
