I'm sorry I cannot give a repr.ex. for this.
I have observed this regularly but still rarely (I use ESS+R every working day for many hours).
As suggested in the title, here are symptoms / typical cases:
I open some R related files, almost always *.R source files when I have no running R process (inferior, i.e. inside my emacs).
I want to send a line or two from my R script to the R process (i.e., I want to evaluate that R code),
calling one of the several ess-eval-* functions via a key stroke
Emacs/ESS notices that I don't have a running R process and asks me about the starting directory.
I typically press [Enter] as ESS guesses the directory fine,
and an *R* buffer is created and R "started inside it".
Some things work as usual, but quite a few others don't.. and I enter search() in the iESS buffer
or use the C-c C-s i.e., (ess-display-search) which shows my short search path and does (in that case) NOT contain "ESSR"
I kill the R process and restart via M-x R or similar, and then all is fine.
As said initially, most of the time this does not happen. However when it happens ESS behaves slowly and "hangs" until my fingers automatically use C-g (without my main brain even being aware of it ;-) ) and use it again and again .. until after some frustration I give up and then typically do notice ("4." above) that attaching "ESSR" had not happened when R was started and so --- no wonder at all --- ESS can not work appropriately, as all the .ess*() utility functions are provided via "ESSR"
I'm sorry I cannot give a repr.ex. for this.
I have observed this regularly but still rarely (I use ESS+R every working day for many hours). As suggested in the title, here are symptoms / typical cases:
ess-eval-*functions via a key stroke[Enter]as ESS guesses the directory fine, and an*R*buffer is created and R "started inside it".search()in the iESS buffer or use theC-c C-si.e.,(ess-display-search)which shows my short search path and does (in that case) NOT contain"ESSR"M-x Ror similar, and then all is fine.As said initially, most of the time this does not happen. However when it happens ESS behaves slowly and "hangs" until my fingers automatically use
C-g(without my main brain even being aware of it ;-) ) and use it again and again .. until after some frustration I give up and then typically do notice ("4." above) that attaching"ESSR"had not happened when R was started and so --- no wonder at all --- ESS can not work appropriately, as all the.ess*()utility functions are provided via"ESSR"