This is definitely worth discussing further and it is something CASPER's been wrestling for a long time.
I think it's also important to note that Xilinx and Mathworks like to support their latest versions. Vendor support for older stuff's always problematic. This also motivates for sticking with the latest versions whenever possible. We have worked-around this problem at SKA-SA by using virtual machines for older toolflows when those are needed. Instrument bitstreams and model files are tagged with all version and revision control information required to rebuild it. I think tagged releases of the git libraries for given versions are appropriate. But it's certainly our intent that the SKA-SA repo will always reflect our latest developments, including support for the latest tools whenever possible. We're currently on Matlab 2012b and Xilinx 14.3. Jason On 24 Jan 2013, at 14:20, John Ford wrote: >> Hi >> >>> I did not commit this change because it's not backwards compatible. >> >> I have noticed more and more instances of backward compatibility issues >> lately as I am using older versions of the tools. At some point soon we >> are going to be forced to make everyone upgrade to the latest tools if >> they are to use the latest libraries. >> > > It seems that's always a problem. The changes have been less serious the > last few revisions, I think, but still, it's not really possible for the > group to maintain libraries for multiple versions of the tools with the > amount of effort available. > > We have talked in the past about having regular releases based on the > timing of the Workshop, but it seems maybe we should try to tie them to > the release of Xilinx tool versions? And patch them as we go between > Xilinx versions. The collaboration relies on large part of a few big > players for much of the work on the libraries, and they all have their own > forks, versions, and patches to handle their own needs. Not pretty, but > it is the way it is due to the volunteer nature of the collaboration. > > I'm mot sure of the right way to go, but it merits discussion. > > Should we start a poll of who is using what versions? > > John > > >> Regards >> Andrew >> >> > > >
