In the strain on scientific publishing, we showed that total publications have grown out of control. A huge part of that was guest edited special issues by groups like #MDPI and #Frontiers.

This ongoing practice is the largest delegation of editorial power academia has ever seen.

2/n

A more hands-on approach could start with patchy, messy data and what we can do (or not) with it. This is what most practitioners face, they are confronted with the problem before having any clue on the solution. Or the solution was never programmed, and go modify someone else's code...

nice, this is similar thinking to @daniel-innerarity.bsky.social

there's snow inside the room. A cup has a flame. Tissue would burn.

"sic transit gloria mundi"

Also worth reading the reviews of the original article. Obviously these reflect past versions of the article. It would be interesting to read the first submission to see how extreme a position the first submission took.

static-content.springer.com/esm/art%3A10...

As part of a conference on the topic I wrote a commentary on why I was quite pessimistic about this approach, echoing some of the concerns lifted here. The proceedings are online: www.nibjournal.ed.ac.uk/article/view...

One reason is that reviewers and editor for journals do a good job. I've seen quite a few preprints (including mine) improve *a lot* and remove mistakes after revision. The same process can catch fraud.

quand en France on arretera de penser que la science est fait *que* par des "Pasteur"