summary.mle2 is sometimes not accessible

[andrechalom]

The summary.mle2 is back to haunt us. As of sads v0.3.1, the summary of the fitsad and fitrad classes are done with

setClass("summary.sads", representation("summary.mle2", fixed="numeric"))

The problem here is that summary.mle2 is not exported by bbmle. However, the code with the line above loads without errors, passes CRAN checks and works in a way that's completely compatible with the expectations. Unless you try to load the package again in the same R session:

> load_all()
Loading sads
### Loads with no errors
> load_all()
Loading sads
Error in reconcilePropertiesAndPrototype(name, slots, prototype, superClasses,  (from sads-classes.R#32) : 
  no definition was found for superclass “summary.mle2” in the specification of class “summary.sads”

[piklprado]

Yeah, I noticed that. But because this discussion

r-lib/devtools#623

I thought it was a devtools issue. So I am surviving detaching sads before calling load_all:

> load_all("../pkg/sads")
Loading sads
Loading required package: bbmle
Creating a generic function for ‘points’ from package ‘graphics’ in package ‘sads’
Creating a generic function for ‘lines’ from package ‘graphics’ in package ‘sads’
Creating a generic function for ‘coefficients’ from package ‘stats’ in package ‘sads’
Creating a generic function for ‘fitted.values’ from package ‘stats’ in package ‘sads’
Creating a generic function for ‘fitted’ from package ‘stats’ in package ‘sads’
> search()
 [1] ".GlobalEnv"        "package:sads"      "package:bbmle"    
 [4] "devtools_shims"    "package:stats4"    "package:devtools" 
 [7] "ESSR"              "package:stats"     "package:graphics" 
[10] "package:grDevices" "package:utils"     "package:datasets" 
[13] "package:methods"   "Autoloads"         "package:base"     
> detach(pos=2, unload=TRUE)
> load_all("../pkg/sads")
Loading sads
Creating a generic function for ‘points’ from package ‘graphics’ in package ‘sads’
Creating a generic function for ‘lines’ from package ‘graphics’ in package ‘sads’
Creating a generic function for ‘coefficients’ from package ‘stats’ in package ‘sads’
Creating a generic function for ‘fitted.values’ from package ‘stats’ in package ‘sads’
Creating a generic function for ‘fitted’ from package ‘stats’ in package ‘sads’

[andrechalom]

OK, if this is the case, then it should be safe to ignore this problem