On Saturday, February 22, 2020 at 8:13:17 AM UTC-8, Snark wrote: > > Hi, > > I wanted to package sagemath's optional package > "database_jones_numfield" for Debian, but it appears to be merely a > binary object. > > The article describing (a version of?) the database is here:
Jones, J., & Roberts, D. (2014). A database of number fields. *LMS Journal of Computation and Mathematics,* *17*(1), 595-618. doi:10.1112/S1461157014000424 That's not a source in the software engineering sense (and in fact tells very little about how the database was compiled anyway) One way to get a source would be let an SQL query write a text version of the database to a file. Your "compiling step" would then consist of going in the other direction again. Depending on how big the text representation ends up being, this might even be valuable in general, because it would be a form that a little less technology-dependent in its representation (given that sqlite etc. aren't that great at compressing mathematical data, I suspect it won't be much worse than the binary representation) For more extensive verification: checking that the entries in the database are correct would be straightforward -- the invariants listed there are not super-complicated to recompute and check. The special value of the database is in its completeness claims in various aspects. That's not easy to verify. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/a7a71675-e575-42d5-b4b6-89a0999ec6ce%40googlegroups.com.
