Displaying 2 results from an estimated 2 matches for "f7e52a3".
2017 Apr 03
3
Very hard to reproduce bug (?) in R-devel
...is called (self$protect, at levels
4 and 7 on the call stack), and in the inner call (7), a variable,
`task`, does not have the correct value. It should be a function, but
it isn't; it's either NULL or some other non-function value.
That code is here:
https://github.com/rstudio/pool/blob/f7e52a3/R/scheduler.R#L79-L81
Note that there's a `force(task)` a couple lines above, which is
commented out. If that is uncommented, the test passes, even though
calling force() there shouldn't make a difference. Also, calling
`format(1)` at that location also causes the test to pass.
I've w...
2017 Apr 03
0
Very hard to reproduce bug (?) in R-devel
...evels
> 4 and 7 on the call stack), and in the inner call (7), a variable,
> `task`, does not have the correct value. It should be a function, but
> it isn't; it's either NULL or some other non-function value.
>
> That code is here:
> https://github.com/rstudio/pool/blob/f7e52a3/R/scheduler.R#L79-L81
> Note that there's a `force(task)` a couple lines above, which is
> commented out. If that is uncommented, the test passes, even though
> calling force() there shouldn't make a difference. Also, calling
> `format(1)` at that location also causes the test t...