Software paper for report about report

[bwiernik]

Behavior Research Methods is also super excited about software package papers lately.

[IndrajeetPatil]

I would vote for AMPPS.

[IndrajeetPatil]

Should we start working on this?

I think it's about time. The API is more or less stable; it's now just a matter of supporting more and more models.

[DominiqueMakowski]

I think before that issues to be fixed are:

• Solve the effectsize/parameters issue (on which one to rely)
• Communicate about how it currently works (so that all maintainers have some idea of the code logic)
• Have the code logic double (triple) checked because currently it might not be the most efficient one (it's slow, because there might be some duplicated operations)

I'll get to that at some point

[mattansb]

I also agree with @DominiqueMakowski that those issues should be resolved first.

@IndrajeetPatil If indeed the API is stable (yay!) that probably deserves a vignette - currently the docs on the pkgdown site are lots of standalone examples. Would be great to see a more cohesive look at the package's functionality.

I suggest focusing on htest objects, g/lm models and data frames at first (before moving on to mixed models, Bayes, etc...)

[DominiqueMakowski]

I suggest focusing on htest objects, g/lm models and data frames at first (before moving on to mixed models, Bayes, etc...)

agreed, having those working well will already be a big thing