If such reuse can be accomplished, it becomes much easier for other researchers to build on existing work rather than having to “reinvent the wheel”. As well as the obvious ease of reusing existing code that (hopefully!) already works and has been thoroughly debugged and tested, there is another – less obvious – benefit of code sharing: It greatly eases the replication of past work, which is essential as a foundation for new development. That is, without access to another researcher’s code, it can be surprisingly difficult to replicate her results because the success or failure of a numerical algorithm frequently depends on subtle implementation details not described in even the most complete of published papers.
Event and apparent horizon finders are excellent candidates for software reuse: Many numerical-relativity researchers can benefit from using them, and they have a relatively simple interface to an underlying numerical-relativity simulation. Even if a standard computational toolkit is not used, this relatively simple interface makes it fairly easy to port an event or apparent horizon finder to a different code.
Table 2 lists event and apparent horizon finders which are freely available to any researcher.
|
http://www.livingreviews.org/lrr-2007-3 | ![]() This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 2.0 Germany License. Problems/comments to |