Talk:Problem of Apollonius: Difference between revisions
→Meta error propagation
Line 26:
::::::: My comparison was not between "J solution" and "Incorrect solution implemented in language X". My comparison was between "J solution" and "languages such as Javascript which yield NaN for 0/0". If the J solution is incorrect for giving a non-answer when there are an infinity of answers, then the Javascript language must also be incorrect. --[[User:Rdm|Rdm]] 19:50, 16 September 2011 (UTC)
:::::::: <code>[(0, -3), 2], [(0, 0), 1], [(0, 3), 2]</code> doesn't have infinite number of answers. Nor does it have an answer involving an inf: there are 8 solutions, all are finite circles. --[[User:Ledrug|Ledrug]] 20:12, 16 September 2011 (UTC)
::::::::: Oh, I misunderstood then. Ok, yes, looking at the J solution, it explicitly checks for the cases where all circles are interior tangent and logically, something like <code>;(_1^#:i.8) <@apollonius 0 _3 2, 0 0 1,: 0 3 2</code> should treat all the cases, but that's not working for me, and I am going to have to do some debugging to figure out why. (Note, by the way, that I did not write the J implementation here.) --[[User:Rdm|Rdm]] 20:27, 16 September 2011 (UTC)
| |||