On 1 December 2010 15:47, pang <pablo.ang...@uam.es> wrote: > On 1 dic, 14:56, David Kirkby <david.kir...@onetel.net> wrote: > >> What has notation got to do with my analogy.? It's the correctness >> that matters. > > For a paper, it is. I would never argue with an author about the > particular notation she uses in a paper, provided it is consistent, > but for a ticket, notation (syntax) is very important, since that is > the syntax we'll all be using from then on.
Yes, but that's not the point I was making. I was making the point that one is expected to check ones proof before expecting others to check it. > A journal is different from Sage in that the journal consists of > independent pieces of text, while Sage is a cohesive piece of > software. If Alice "publishes" any kind of correct code in Sage, this > might conflict with code by Bob, something you don't see in a journal, > where they can use incompatible notation. Yes, I agree things have to fit together - and that includes they work on other systems too. A bit of code that conflicts with other code is bad. > And anyway, reading submitted code is important for many other > reasons. So an author writes correct code. It passes all tests. But > maybe the I'm not disputing that. > In software, correctness is very low fence. In a project such as Sage, > with lots of packages written in different languages, you're asking > for trouble if you accept any code that passes all tests. I agree! I've never suggested a positive review should be based solely on something passing all tests on all platforms. In fact, I've just commented on http://trac.sagemath.org/sage_trac/ticket/10187 that I felt one of the tests was inadequate - also see my email "When is a test not a valid test?" on sage-devel. It should also be noted that the best time to decide on what algorithm is used, is not once the code is written. That should be decided at a design stage, which takes place before coding begins. For new contributors the situation is different. As a reviewer I would not mind spending some time testing code a bit more than usual. But for someone that regularly submits tickets, if they can't be bothered to test them, then I'm personally not going to spend much time on a ticket. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org