Am 10.02.2017 um 11:47 schrieb oleg_komashko(a)ukr.net:
Dear all,
examples of different uses of '@' in R foreign
different S4-slots alongside with different '@string'
Yes that's expected I'd say. But what if in the gretl context in
addition there is a defined string named 't' or 'test', then the part
"urdf(a)test.name" or "urdf@teststat" will break I think.
I would say as a script writer you don't always have full control over
the outer gretl context.
In any case it would be difficult to debug, because if R fails you don't
get informative error messages in gretl.
So my preference would be to turn it off by default, and perhaps let the
script writer switch it on if he wants to bear the risk.
thanks,
sven