Active to-dos are on top.
-
Create build process - perhaps even resulting in an executable? (Though I personally prefer to use this in a REPL).
-
Add intervals feature
-
"Show negative implications" (only for graphs with under 10 nodes.)
-
Generalise jeffrey to any database (with area for own databases on the web interface)
This seems very complicated (i.e., manual work).
Instead of using the labelmaker in bulk. Not sure if this is a good idea.
Reduced complexity of graph-implies-not-p
by switching from a nested (some ... (some ...))
to two separate loops. Measured improvement of about 7% on SBCL and 25% on CCL (notes available).
PDF output does not render fancy labels properly. Why? For the moment I use png, which is the current default.
The external-programs (bash scripts, calls to the CL-library external-program
) only work with Linux. I am withdrawing claims to future MS/iOS support, as there are no interested users, as far as I know. The occassional MS/iOS users seem happy with the web interface. If you are an MS/iOS user and want to use this program in a CL REPL please let me know.