Revision history [back]

Thanks, these are all good answers (I don't have any points, so I can't give thumbs up, I'm afraid); I will try to work those into my routines. Did I simply miss them in the documentation somewhere?

I agree Mike, that the syncing is the hardest part. How lucky we are that Sage is open source, so such documentation is less essential ;-) I guess that if such a table from algorithm names to Sage functions had been made and was complete, most Sage developers would be aware of having to update that along with some given patch that would implement or change the use of a named algorithm. Because of the review system (it could be explicitly added to the documented review procedure), a reviewer would often catch it if the patch developer forgot. One could also implement a best-effort tool to check if the patch might cause changes to the table: the tool could see if the patch changes stuff in a function that is mapped to or a direct dependency hereof and tell the developer to examine these cases. We still have the problem, of course, whenever one of the external libraries changes or implement new algorithms.