DIYEFI.org Forum

Just post a link and instructions on how to integrate it with your code in order to build it. I'd interpret "use" in the broadest sense, including the code. If you place it up there with GPL, that restriction about it being on their processors only is in conflict with the GPL. Is USB really that hard? :-p Can't you roll your own? Or find a FOSS lib to use?

Or, here is a thought, build the object file from the USB stack and include that with the GPL code, I think that would be acceptable.

Fred.

I would say it sounds like Keil or NXP are not going to come after you for this, and if they do they'll let you change it later. My guess is that no one will fuss about it if you post it or not. They certainly left it open for interpretation.

Fred wrote:Just post a link and instructions on how to integrate it with your code in order to build it. I'd interpret "use" in the broadest sense, including the code. If you place it up there with GPL, that restriction about it being on their processors only is in conflict with the GPL. Is USB really that hard? :-p Can't you roll your own? Or find a FOSS lib to use?

Or, here is a thought, build the object file from the USB stack and include that with the GPL code, I think that would be acceptable.

Fred.

I never thought about the other way round (thanks for pointing that out). By the stack only being allowed on nxp lpc stuff, it's automatically incompatible w/gpl.

The closest foss stack is LPCUSB, and I'm not impressed by it at all (it's polling paradigm disgusts me). It would still need adaptation for my hardware (17xx CM3 series vs 2xxx ARM7 series). What I'm using now is nxp's stack that is found here Anyone can get the source for any reason. It took me about 2hrs total to config it and get it running - laying out the board will take me longer.

I do not want to roll my own stack, I want to finish my nextgen flowbench. If I have to roll my own stack to be 100% gpl compliant, then it's not worth the hassle. I do not want to spend the next few months 'polishing' up the lpcusb stack to my standards in order to have something 100% foss. My time and resources are limited, and I just want to make stuff instead of tinker with it.

Sometimes foss frustrates me because I like to do the right thing and follow the rules. I guess I can ask the factory for clarification since anyone can grab the source from the above link and still do whatever they want with it.

firmware + foss = rare stuff --- which sucks, because everything needs to be 're-grown' from the ground up in the embedded world. pre-existing hardware stacks and frameworks are huge timesavers and big money makers which is probably why there aren't many open alternatives out there.

FOSS code can use and link against non-FOSS libraries (and vice versa). Do as I suggested and include a binary or a link to a binary that can be linked in at build time. It can sit on my server, no worries. They do give you permission to distribute binary files for use on their processors. Prelinked object files must be included in that.

Fred.

interesting project. I machined up a fuel rail and a plate that could hold 4 injectors, and had planned on making a rig where I could squirt 4 at a time and observe the flow pattern and measure the amount of flow, but... still to be finished

Instead of a fuel pump on mine, I had planned on a small pressure vessel that I could use an air regulator and air compressor to regulate the fuel pressure, ie, a short tube for air in, and then a dip tube to get the fuel out.

Nice idea re the air pressure, what about fuel rail hammer and the damping that the air would provide, i imagine it would affect the results somehow?