Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

"Cyclic dependency check skipped" in release() #602

Closed
krlmlr opened this issue Sep 29, 2014 · 9 comments · Fixed by #767
Closed

"Cyclic dependency check skipped" in release() #602

krlmlr opened this issue Sep 29, 2014 · 9 comments · Fixed by #767

Comments

@krlmlr
Copy link
Member

krlmlr commented Sep 29, 2014

When I call devtools::release(), at some point during the check process the following message appears:

Package has help file(s) containing install/render-stage \Sexpr{} expresssons but no prebuilt PDF manual.
* checking package namespace information ... OK
* checking package dependencies ... NOTE
  No repository set, so cyclic dependency check skipped
* checking if this is a source package ... OK
* checking if there is a namespace ... OK

I'm not sure if the NOTE about package dependencies is related to the note about the \Sexpr-s. With check(cran = TRUE) these notes are not shown.

@hadley
Copy link
Member

hadley commented Sep 29, 2014

We could resolve this by setting a CRAN options somewhere, but I'm not sure where given that we're running with --vanilla. Is there an env var?

@krlmlr
Copy link
Member Author

krlmlr commented Sep 29, 2014

I have to take a closer look. This might be related to #603.

@hadley
Copy link
Member

hadley commented Sep 29, 2014

I don't think it's related.

@krlmlr
Copy link
Member Author

krlmlr commented Oct 6, 2014

The note also occurs with devtools::check(check_version = TRUE). It is related to an unset repos option.

From the description of R startup, I guess we'll have to:

  • fabricate a one-line .Rprofile
  • set the R_PROFILE_USER environment variable to that file
  • use --no-site-file --no-environ instead of --vanilla

How about adding a new argument with_repos = FALSE to the internal R function?

@hadley
Copy link
Member

hadley commented Apr 21, 2015

How does this work on travis? You can set a CRAN env var there.

@hadley
Copy link
Member

hadley commented Apr 21, 2015

Related to #678

@krlmlr
Copy link
Member Author

krlmlr commented Apr 21, 2015

We seem to need all of the three above -- our own .Rprofile, setting the env var and changing command-line arguments. The .Rprofile would look like this:

options(repos=c(CRAN="http://cran.rstudio.com"))

@hadley
Copy link
Member

hadley commented Apr 21, 2015

Seems like we should modify R() and RCMD() to do that?

It might be possible to just do it via env vars, by setting R_ENVIRON, R_ENVIRON_USER, R_PROFILE, R_PROFILE_USER etc. I think it requires a close reading of ?Startup to figure out what the best process is.

Maybe we can drop --vanilla in favour of ignoring user files, and overriding the location of R_PROFILE to something we create?

@lock
Copy link

lock bot commented Sep 18, 2018

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/

@lock lock bot locked and limited conversation to collaborators Sep 18, 2018
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants