if a published article describes an algorithm for the first time, where is the scientific contribution "attached" to, if the algorithm was used in practice before the article, but it was not disclosed explicitly? To the algorithm, or to the article itself, that describes it, or there are two contributions - one for the invention of the algorithm by itself, the second for the article that makes it public and describes its advantages, disadvantages and compares it to the other algorithms in the field?
Update: This events did not yet occur, and I'm trying to understand the situation in advance. The software is available publicly, but its unknown where and how much it is used. The article cites the software as containing the algorithm, and the software is used for tests inside the article. One of the 3 authors of the article invented the algorithm and wrote the software.