This is a case of too old version (.24.5, current is .28.0) which only supports tls1.1 on Windows (at least win7, not sure about 10). The linux and likely Mac versions already supported tls1.2 and thus still work. Building on Linux is a breeze just do a make and away it goes. On Windows the standard instructions either don’t work or build a 32 bit version which Win64 won’t link. I eventually stumbled across the appropriate cmake incantation to produce a Win64 binary, at which point I hit the Qt assert issue. After being able to create a dll, I tried just substituting the dll, but it still trys to use tls1.1 and fails.Compiling from source with the new libgit2 works and then hits the Qt assert. If you can pull up a dev environment (there are instructions here for linux which should be similar to a Mac as it is FreeBsd on top of a micro kernel:
and are familiar with Qt, perhaps you can see what it wants to avoid the assert. As far as I could see Fritzing calls something to render an svg and I think (but the Qt folks in their forum didn’t) that it is a Qt bug. For me changing the parts db (as tspans don’t appear to have any valid use) is easier (if not very fast.) Hopefully I will soon be ready to make the parts update which will at least enable development to proceed (until the next roadblock). I to the point of testing the parts (at least the first set, there may yet be more) that need to change. Once I’m sure they are correct I’ll submit the parts pull request and see if it is accepted.
Peter