[sophacles]

Meh. That is a surprisingly subjective and arguable statement, when you consider there are useful purposes for code besides running on your current dev system. For instance I frequently find code on the net that is rotten from the POV of "it runs on the latest git head of the dependencies" but completely and utterly fresh and relevant in the "how would I approach a solution in this language with this general design paradigm" POV. Other forms of "rotten" include software written to $HEAD, only to have those features never incorporated in an official release. The list goes on and the differences between "rotten" and "doesn't work for my special case" grow very indistinct.

[Someone]

"does not run on my current system" is not a sign of rot; "I cannot even make a guess whether it will be useful for me without trying" is. To make such a guess, the following info can be helpful:

- for what platform(s) it is written (e.g. C code where tagged as 'Unix', that use Linux-isms)

- what the license is (home page claims 'free' or 'open source', source files mention license A, read me mentions license B)

- what the library does and, almost more importantly, what it's limitations are.

- what the dependencies of the library are.

- what systems the code has run on (it can for instance be useful to know that code has run in big-endian machines if your machine is big-endian)

The more of this info is present, the higher the probability that I will try to use your software.

There is lots of code out there where too little of this metadata is present. If such software is not solving some unique problem, I would prefer it if its writer did not make it available.