On Jan 19, 4:30 pm, duncan smith <[EMAIL PROTECTED]> wrote: > Carl Banks wrote: > > On Jan 18, 7:01 pm, Paddy <[EMAIL PROTECTED]> wrote: > >> On Jan 18, 9:47 pm, [EMAIL PROTECTED] wrote:> Tim, > > >>> Thanks for the topsort code. It would be useful in a project I'm > >>> working on. Can I use the code for free under public domain? Thanks! > >> When I needed one I didn't know the name. I'm curious, how did you > >> know to look for the topological sort algorithm by name? > > > I spent quite a bit of time looking for this one myself. It was quite > > a stumper. Sometimes Google gets us in the habit of just firing > > random search terms when we ought to be thinking it out. > > > After quite of bit of dead end searching--like a week--I stepped back, > > thought about what I was looking for. I wanted a sort of dependency > > system: A must happen before B. What other programs do that? make, > > of course. I looked at the documents for make and found the term > > "directed acyclic graph", and pretty much instantly knew I had it. > > Searching for "precedence diagram" might throw up some relevant results; > but you've probably already discovered that. I have some basic > precedence diagram code (no warranty etc.) somewhere if anyone is > interested. > > Duncan
I searched for dependancy sort, and later dependency sort (cos I couldn't spell). I had convinced that I was using the right term and was flummoxed by the lack of hits. Even today the term topological sort means far less than what it describes: sorting items based on their interdependencies. Is this a case of something being named after its mathematical/ technical description and so obscuring its wider practical use cases? - Paddy. P.S. we have revived a thread started in 1999! -- http://mail.python.org/mailman/listinfo/python-list
