On Monday 15 June 2009, Xiaofan Chen wrote:
> >> Yet, it must be noted that libftd2xx is dynamically linked on Linux.
> >> If the Windows version is also a DLL then it is arguable whether or not
> >> they are already linked when the OpenOCD binary is distributed.
> >
> > One might argue that, yes.  Safer IMO not to though ... especially
> > if a D2XX.dll (or whatever) is distributed with the OpenOCD binary.
> >
> 
> I am not so keen in the discussion of licensing issues.

Few of us are lawyers.  Some of us have spent time with them
though, talking about such issues.


> But I think 
> the FTDI D2XX library is a DLL (ftd2xx.dll). It is also distributed
> through the Windows D2XX driver package and installed when the
> driver is installed. So there is no need to distribute this DLL for the
> OpenOCD binary under Windows.

That's irrelevant to whether or not distributing a version
of OpenOCD which *expects* that library violates the license.
Go back and read that FAQ entry I referenced before:

  http://www.fsf.org/licensing/licenses/gpl-faq.html#GPLIncompatibleLibs

Especially:

        If you want your program to link against a library not
        covered by the system library exception, you need to
        provide permission to do that. 

The FTDI code is such a library, and there is currently no
such permission.  End of (non-pointless) argument.


> http://www.ftdichip.com/Drivers/D2XX.htm
> http://www.ftdichip.com/Drivers/CDM/CDM%202.04.16%20WHQL%20Certified.zip
> 
> So in this case, there is no issue to use D2XX for Windows, right?

Wrong.


> Under Linux, I am not so sure.

I'm sure.  The legal issues are the same in either case.


> Anyway, libftdi and libusb both 
> work under Linux, so there is no real need to use libftd2xx
> under Linux now that Nicolas says that there is no real performance
> gain to use libftd2xx under Linux.

The serial engine is better-documented than the bitbanging
support.  If OpenOCD needed to use e.g. an ft232 (no serial
engine) instead of an ft2232 (with SPI/JTAG/... serial engine),
that could make a difference.  But ... we don't.



_______________________________________________
Openocd-development mailing list
Openocd-development@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to