Talk:Percolation/Site percolation
Tcl percolation?[edit]
Hi, it seems as if the Tcl output does not show a successful percolation through a 15x15 grid? --Paddy3118 (talk) 16:12, 8 September 2013 (UTC)
- Yes. It's random, and around the threshold level where it can go either way. Requiring a successful percolation is lame. I'll remove that word from the task description. –Donal Fellows (talk) 23:26, 8 September 2013 (UTC)
- Why remove that feature Donal? Seeing a successful percolation is much more involving for the viewer than not. Also the other examples have that feature, making Tcl the odd one out? --Paddy3118 (talk) 01:27, 9 September 2013 (UTC)
- Because it's a foolish requirement. –Donal Fellows (talk) 09:44, 9 September 2013 (UTC)
- Why remove that feature Donal? Seeing a successful percolation is much more involving for the viewer than not. Also the other examples have that feature, making Tcl the odd one out? --Paddy3118 (talk) 01:27, 9 September 2013 (UTC)
I can live with your amended wording "Optionally depict a percolation through a cell grid graphically." Although I would have preferred to emphasize that if there is no through path then there is no percolation. --Paddy3118 (talk) 02:59, 11 September 2013 (UTC)
Granularity and tries number[edit]
The increase of granularity and tries number in the D and C entries was just to make this task a little more interesting. Requiring a fixed number of simulations in the task description is silly. The simulation outputs are similar, just with more significant digits. Different languages as Python and D have different advantages, and the D entry is designed to be faster. If you don't allow a higher number of simulations, those speed optimizations become useless, and using D instead of Python to run a simulation becomes not much useful. One of the points in using a language as C/D to perform this task is to run a higher number of simulations compared to a Python implementation. So please allow a higher number of tries in the task description. -bearophile (talk)