Is oSCR still actively under development?

[Daniel Barton]

Hi All (especially Chris, Andy, Dan Linden, etc.),

Is oSCR still under active development? I'm making tough decisions about which R package(s) to teach in our curriculum - I used to use oSCR - but I see there hasn't been a repo commit in over a year, and it seems like development has ended. I don't mean this as some kind of dig - I really appreciate all the work that everyone put into developing oSCR and into developing SCR more generally. Just trying to decide whether to use, over the next few years, JAGS & secr as my teaching combo, or stick with JAGS & oSCR. Thank you!

Best,
Dan

--

--
Dan Barton

Note: this is not my official Humboldt State email.

[Jeffrey Royle]

hi Dan,

 Thanks for the inquiry -- right now there is essentially no active development, although we have no intentions to mothball things, but also not much time to engage with it right now (on my end).  

regards

andy

--
You received this message because you are subscribed to the Google Groups "oSCR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oscr_package...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/oscr_package/CAGSCbgqc%3D9nTkxvus8pvfDeL5zsLhrx702OkhhfRT3qc-DfnJQ%40mail.gmail.com.

[Daniel Linden]

Hi Dan, been meaning to reply to this.  FWIW, I still think oSCR is a great teaching tool.  It was initially developed to do some things that secr could not, and then it kept expanding to address various needs (sometimes identified while teaching workshops).  Given the transparency of the R code, anyone could take this and add features if you were willing to dig into the likelihoods.  But we have definitely not added all the functionality one might want.  It was also designed to make the data processing needs straightforward such that you could simply jump to a JAGS model for estimation.

That said, if you are concerned about active development and cutting edge, the same argument could apply to JAGS.  There is more efficiency and capability with NIMBLE now, for fitting complex and comprehensive Bayesian models to SCR data.  But... teaching the basics can still be adequately achieved with JAGS (and oSCR!).