search

llrs / blogR

My blogdown site with R related posts
https://llrs.dev
0 stars 0 forks source link

In depth analysis of the Bioconductor submission #34

Closed llrs closed 4 years ago

llrs commented 4 years ago

I doubt this will be long, but check the comments of Bioconductor issues,

some issues get closed after they got assigned a reviewer and before the reviewer actually gets a chance to start the review.

I see two paths, check the labels or download all the comments of the issues and see what happens:

idea make a plot where horizontal line is # comments and vertical line issues colored by who post (bot, reviewer, other, submitter), and try to cluster them to see patterns.

The same could be done but with time since submission. That could make post 2 and #32 would be part 3 of the serie

This repo might help, also it would be nice if based on gh I set up a package to retrieve "social" information from repositories to simplify this queries.

llrs commented 4 years ago

Created https://github.com/llrs/socialGH to help with that

llrs commented 4 years ago

Check when issues get on hold like https://github.com/Bioconductor/Contributions/issues/1481 Ok on June 2nd and still waiting for reviewer official acceptance a month later.

Suggesting assigning a reviewer only when the package has been built on Linux, Mac, and Windows without errors or warnings on all platforms. (This would avoid wasting reviewers time, at the expenses of perhaps not give suggestions early on, but I think that is covered fairly well by Martin ).

llrs commented 4 years ago

Also consider doing this for rOpenSci and present it at rstudio::conf(2021) and then/maybe write a paper on R journal or alike

llrs commented 4 years ago

Submitted for rstudio::conf(2021), also posted online

llrs commented 4 years ago

Done for Bioconductor (that's why I closed the issue), now working on rOpenSci analysis (See rOpenScie blog for previous analysis), for CRAN see this issue: https://github.com/lockedata/cransays/issues/8

Could try to do that manually/just once for the talk (if I get to give it)