On Jul 18, 2008, at 1:02 PM, Kurt wrote: > > Robert -- > > I'm afraid that I did not take detailed notes of your talk at Dev > Days 1. > Is there documentation that shows how one ought to now implement > coercion for brand new code? > Perhaps a good example that we can all look to for guidance? > > -- Kurt
http://wiki.sagemath.org/days7/coercion/ is the most up to date, which is to say not very. Documentation is certainly a high priority for me now, though one *very* good thing that came out of Sage Dev Days is that a lot more people understand the coercion model now. The major push of this coercion rewrite (at least as it started) was to make a better, more understandable API for people to use. In terms of status, I got the new parent under the old parent (and all doctests passing) and swapped out the coercion models (in the process of writing a compatibility layer for old parents). When this goes in, we can start writing against the new model. QQ, RR, CC, etc. will probably be the best examples. > > > On Jul 17, 11:25 pm, Robert Bradshaw <[EMAIL PROTECTED]> > wrote: >> Some of you may be wondering how the new coercion model is coming >> along. After a heroic effort by many people at Dev Days 1, and a >> little followup work, we managed to merge the coercion branch into >> 3.0.3 and got all doctests to pass with the exception of /modular >> and /schemes/generic. It has become apparent however that the >> "coercion" branch contains much, much more than just coercion. We >> also want to have the highest assurance of the integrity of the >> resulting code which is hard to do with such a large set of >> interdependent changes all lumped together as a single patch bomb. >> Therefore we have decided to take the work we have so far and move it >> over piece by piece. Specifically, we are inserting a new Parent >> underneath the the old Parent object so we can migrate one thing at >> at time. We will also be merging in the many not-directly related >> improvements individually as separate tickets. The most significant >> impact this will have is that it may take longer for the "new >> coercion" to be 100% done, but the part that is done will be in Sage >> itself, as well as being sanely refereed. >> >> - Robert > > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
