miniblog.

TIL Tcl has a notion of 'safe interpreters', a mode where you can run untrusted code in a sandbox: https://www.tcl.tk/man/tcl8.4/TclCmd/safe.htm Not many programming languages have this, but it's way safer to include in the implementation than try to build as a userland library.

Has anyone built a great solution to 'run all my unit tests automatically'? It's straightforward to write a while loop in bash, but handling timeouts, syntax errors etc well is hard. Running on save would be good, although I wonder if you could run fast tests on each keystroke.

It's a small thing, but I'm much happier with the output of --version in the latest version of difftastic. It shows the release version number, the commit hash, and the commit date. This gives you a sense of the age of release, but you still have a reproducible build (unlike build time). It also shows OS, arch and compiler, because those are common requirements in bug reports.