possible error (relates to stderr() not being a pipe)

[dankelley]

Hello. I am trying to do revdep_check on the oce package, in preparation for a CRAN submission (which is necessitated because oce relies on Rcpp, and the latter is due for a CRAN update next week, which will break oce).

I see as below the line of dashes. I'm not too sure whether this means that there is a problem with the argoFloats package, the oce package, or with revdep_check itself. I wonder whether anybody could advise me on what to try next.

Both oce and argoFloats are big packages, and I'm not asking anyone to look into the details of the problem, if it indeed rests in those packages. What I am hoping, however, is that someone might give me a hint as to how I can find out what's wrong in what I'm doing.

Thanks in advance, for anyone who can provide some guidance.

PS. a git grep stderr within argoFloats produces no results. The same is true for oce.

---

> revdepcheck::revdep_check()
── INIT ───────────────────────────────────────────── Computing revdeps ──
── INSTALL ───────────────────────────────────────────────── 2 versions ──
Installing CRAN version of oce
Installing DEV version of oce
── CHECK ────────────────────────────────────────────────── 10 packages ──
[0/10] 00:00:25 | ETA:  ?s | (1) argoFloats [C_]Error in process_get_error_connection(self, private) : 
  stderr is not a pipe.
Type .Last.error.trace to see where the error occurred
> traceback()
10: stop(cond)
9: throw(new_error("stderr is not a pipe."))
8: process_get_error_connection(self, private)
7: process_read_error(self, private, n)
6: proc$read_error(n = 10000)
5: handle_event(state, i)
4: handle_events(state, events)
3: run_event_loop(state)
2: revdep_run(pkg, quiet = quiet, timeout = timeout, num_workers = num_workers, 
       env = env)
1: revdepcheck::revdep_check()

[gaborcsardi]

Try updating all dependencies of revdepcheck:

pak::pkg_install("r-lib/revdepcheck", upgrade = TRUE)

[dankelley]

Thanks very much. All works fine now, so I'll close the issue.