oSCR does likelihood analysis of limited classes of SCR models, much less than 'secr', and so how can oSCR possibly be useful? Here are a few of the things that motivated our development of oSCR:
1. 100% native R code, making it (reasonably) accessible to people who know R and, presumably extensible by ordinary R programmers.
2. Because it's written in R you can look at the code to figure out exactly what's going on.
3. It's quite a bit slower compared to secr but we think it's quite robust to massive sized problems.
4. The data structure is relatively simple, just as ordinary R lists (for the most part).
5. The models accommodate least-cost path models of distance [note: 'secr' has this capability now]
6. oSCR forces you to define the state-space of the point process which we think is important.