CRAN takedown notice

[smjenness]

Perhaps related to #10.

Dear maintainer,
Please see the problems shown on https://cran.r-project.org/web/checks/check_results_networkLite.html.
Please correct before 2025-01-10 to safely retain your package on CRAN.
The CRAN Team

@krivit : could you help me address these issues and get a version of networkLite ready for publication before Jan 10?

Running R CMD Check with different versions of network and statnet.common:

• The Github Remotes in the DESCRIPTION for network can now be removed
• The Github Remotes in the DESCRIPTION for statnet.common is still needed, as the CRAN version of statnet.common does not include the modify_in_place fx now needed by networkLite. Will you be publishing to CRAN statnet.common first?
• With the Github remotes for statnet.common, I am getting the following documentation errors:

✖ matrix_conversions.R:103: S3 method `as.matrix.networkLite.adjacency` needs @export or @exportS3method tag.
✖ matrix_conversions.R:130: S3 method `as.matrix.networkLite.incidence` needs @export or @exportS3method tag.
✖ matrix_conversions.R:146: S3 method `as.matrix.networkLite.edgelist` needs @export or @exportS3method tag.

[smjenness]

@krivit A quick update: I filed PR #12 that removes the Github Remotes to network and fixes the documentation errors. I think all that is needed still is your publication of statnet.common.

[krivit]

@smjenness , does EpiModel have a standard release process? I don't want to break it if you do, but my suggestion is to fork a release branch that gets cleaned up (in particularly by removing Remotes: directives and enabling stricter checks) and keep the development stuff in main.

I haven't gotten around to adding stricter R check scripts; I'll do it tonight.

[krivit]

Oh, and I can push statnet.common shortly.

[smjenness]

@smjenness , does EpiModel have a standard release process? I don't want to break it if you do, but my suggestion is to fork a release branch that gets cleaned up (in particularly by removing Remotes: directives and enabling stricter checks) and keep the development stuff in main.

I haven't gotten around to adding stricter R check scripts; I'll do it tonight.

I don't have that type of branching process set up for networkLite. Once you publish statnet.common, I'll ask @AdrienLeGuillou to just double check this updated version of networkLite against our depending packages (principally, EpiModel). Thanks!

[krivit]

I don't have that type of branching process set up for networkLite.

There is no set up as such. Release branching is a procedure one follows when preparing a release. For example, in this case, the code in the main development branch has different needs from the code to be compiled into a tarball and sent to CRAN. It needs to be able to depend on other development code, but it's not a big problem if it has partially implemented features, incomplete documentation, or temporarily fails some tests. On the other hand, the code sent to CRAN must depend only on other CRAN packages, and it must pass strict tests. Thus, a branch is created to stabilise and clean up the code for release.

[AdrienLeGuillou]

Note for me:

• wait for statnet.comon 4.11 to be on CRAN
• make a 1.1.0 branch with the updated requirements and removed remotes in description
• update on CRAN (see with @smjenness )

[smjenness]

Fixed with #12 and updated statnet.common published on CRAN.