This would mean the same on trademarks, plus:
Restriction of copyright to a very short term (one could think of one year - this is also the difference Peter Deutsch and the Ralph Levien [17] have used for their different ghostscript licenses - and five years (librarians estimate that 50% of books go out of print within 5 years [21])) so that news and entertainment providers as well as software developers are not touched.
It could also be demanded that copyright is not granted per default but only after registering the document (and maybe deposition of its electronic version) with a unique document object identifier (librarians would be happy about this and surely put up a web-interfaced registry).
Patents are far more difficult and proposing optimizations can easily backfire [35, p. 109] quotes an SPI representative that the institution has failed; the same arguably applies to the simplified jurisdiction at the CAFC (Court of Appeal for the Federal Circuit).
Probably it is best to treat different industries differently (make e.g. patent duration variable on patent classification); this may mean to continue non-patentability for software and finance or even extend non-patentability to other areas. One could think about dropping the examination (which consumes time and resources; furthermore granted patents are more lucrative to the offices which of course also corrupts the examination process [1].
Also one of the compromises that Lutterbeck [18, p. 132] proposes such as to allow distribution but not commercial distribution of source code (source code privilege) is problematic in that it hinders commercial usage of open source and commercial usage of non-open source software in general.
Personally, I could imagine the minimal alternative as a long-term option and the moderate alternative as a more realistic political agenda.