On Thu, Apr 9, 2009 at 5:19 AM, Robert Bradshaw < rober...@math.washington.edu> wrote:
> > On Apr 9, 2009, at 1:49 AM, John Cremona wrote: > > > Can we at the same time convert more files to ReST and hence get them > > included in the Reference manual? Although that manual is long, there > > is a whole lot of stuff not mentioned in it, which is bad PR. > > > > What I mean in practice is, if someone is going through a source file > > whose doctests are incomplete and also old-style, then they convert > > each to new-style as they go, as well as (at the same time as) adding > > the new doctests. > > That seems like a good idea to me. I've seen the guide to ReST, but is there a document somewhere summarizing what usually needs to be done to Sage functions? Is changing EXAMPLES: to EXAMPLES:: and adding a newline after it usually enough? > > 2009/4/9 William Stein <wst...@gmail.com>: > >> > >> Hi, > >> > >> In order to get Sage's doctest coverage to 75%, which is one of the > >> main goals for Sage-4.0, it is necessary to doctest about 1900 > >> functions in the Sage library. I just did a test (see > >> http://trac.sagemath.org/sage_trac/ticket/5724 and > >> http://trac.sagemath.org/sage_trac/ticket/5725) and it took me on > >> average about 2-3 minutes per doctest, including fixing a couple of > >> bugs I found. Estimating 3 minutes each, that is about 95 hours of > >> nonstop work. > >> > >> It is very easy to find files that are full of code that isn't > >> doctested and *must* be doctested. It was trivial for me to find the > >> sequences and quotient_ring_element had very bad coverage. In both > >> cases, I found *several* small bugs in both files. Probably, any > >> random file with low coverage has several bugs in it. That's why > >> getting the doctest coverage in Sage up is critical to Sage > >> becoming a > >> truly high quality system. > >> > >> I hope those of you who feel qualified to write doctests, will help > >> out. It's nearly impossible for one person to do all 1900 of those > >> doctests in the next month. Writing doctests is not easy and it > >> absolutely requires experience experience with the Sage library, and > >> good understanding of the code you are testing. > > +1. I often end up writing a lot of doctests for surrounding code > when converting things to the new coercion model. > > How about another Sage doc day in the coming month? I bet with enough > of us we could at least make a sizeable dent in the 1900 tests needed > in one big shot. Sounds like a good idea to me. I'll try to show up and work on p-adics. :-) How about this Saturday? David --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
