…tory. Simplified code by removing the special case of linepos = -1 denoting the end of the line.
…then selecting the previous history item. I might want to look at a better approach for this. Maybe s.bksp() and such should change s.linepos themselves?
… code more understandable.
…ak(), and continued working on Conn.InterpretCommand(). The latter is pretty ugly; I still have to think a bunch more about how I want it to work.
…few tests have been written. Documentation has also been added. Now I just have to figure out if I want to split this into it's own package...
…to add cursor movement and unit tests. Not sure what terminals this will work on.
…le-manager (it's functionality is now in moduled)
…s to what is considered an error), moved wfdr/pipes into wfdr/moduled (since that's really where it belongs)
…to make it more flexible, more robust, more secure, and more testable. Added unit tests to test FSManager (default session manager).
… I was writing it (I guess that's the point). Next is a test for Daemon sync :)
…o wfdr/cache). Currently working on tests for Sync()
… because right now there are way to many child processess for monitoring resource directories
…r automatic building.
… go versions of these c programs (or else somehow otherwise integrate them into our build chain)
…ror checks, and removed wfdr/perms GetPerms() function (it has long been depreciated; perms.Get() is preffered)
…a more intuitive interface, and have more functionality (and thus less functionality duplicated in various modules)
…brary, and removed experimental diff package.
…ize git would do that automatically
…(not wfdr/dlog), changed error variables from e to err, and fixed a few spots where errors were not correctly handled.
… were cleaned up, but rather just assume that they closed when sent SIGTERM. Also fixed shared-daemon script *not* exiting upon receiving a SIGTERM, which was causing the problem to manifest itself.