## Packages retired from CRAN [R for BE/BA]

Hi Yung-jin,

» I am totally doomed and the party is over now.

Don’t Give Up

» » ... Since the original GUI-packages are retired, …
» It's not retired. It was removed from CRAN since the author did not respond,…

That’s the same thing.
Retired = archived = only old sources available on CRAN = bad luck for Users requiring binaries for their OS.

» … though it still works well under R v4.0.x.

I believe it. I have no problems with bear 2.8.8 under R4.0.2. However, there are errors in functions (or arguments) you don’t call in you packages.

» Hopefully it will come back again soon. It happened before as far as I can remember.

Only if the maintainer would do what the word implies. Fix the error(s) and re-submit. Since in this case the maintainer of gWidgets* and gWidgets*2 is the same person, the last version of the former was of January 2019 and the first one the latter submitted at the same time, chances are close to nil that he will exhume the former and take care about both in parallel.

» gWidgets2 is a disaster to me. It's not upgraded version of gWidgets at all. It's like an add-on to gWidgets. I have tried it and the GUI just looked messed up.

My knowledge about GUIs is close to zero. I think you have still some options:
1. Fix the errors and submit the packages as the new maintainer. Then it would mean back to the stressful life which was a reason why you don’t host your packages at CRAN for years.
2. As above but host binaries (for Windows and MacOS) and the source (for Linuxes) at your site.
3. Jump in at the deep end and adapt your code-base to gWidgets*2. Nasty first though not that bad on the long run.

PS: Cheer up! Two stories of this year:
• There was an error in package pbkrtest. CRAN’s Kurt Hornik repeatedly asked the maintainer for a fix with no reply. On Feb 19th maintainers of dependent packages were notified that ours will be archived on Mar 4th as well. Panic. E-mails to and fro, fixed on Feb 20th. What if the maintainer would have been on vacation?
• In the transition for R3.6.3 to R4.0.0 some tests of our package replicateBE suddenly failed and I was asked by CRAN for a fix. Why the heck? It turned out that the default singularity tolerance in package lme4 changed. Was a kind of scavenger hunt to figure that out.

Dif-tor heh smusma 🖖
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes