I have done a bit of compare & contrast when it comes to the Code4Lib mailing list. And I can demonstrate how discussions about specific computing issues in libraries have declined, and job postings have increased.
In 2011, there were approximately 3,000 postings to the list for a total of .98 million words. By comparison, in 2021 there were 1700 postings and a total of .5 million words. To put it another way, the mailing list's volume has decreased by about half.
I calculated quite a number of frequencies based on a wide variety of features (ngrams, parts-of-speech, named entities, etc.). In 2011 the names of people dominated the entities, but in 2022 the names of organizations dominated. When I did topic modeling against 2011, themes included: "conference", "library", and "data". On the other hand, themes from 2022 included: "library", "experience", and "digital".
Probably the most telling model was the word collocations. While the attached images may be too small to appreciate all the nuances, the 2011 graph includes the names of many individual people and computing issues like "data", "MARC", and "RDF". The 2021 graph has a much larger emphasis on "experience", and there is a set of related words regarding race & gender, which come from the boiler plate paragraphs of job postings.
Mailing lists aren't what they used to be. Of the mailing lists in which I subscribe, zero discussion happens. There are really only announcements. I suppose the Code4Lib mailing list is no different.
That said, the mailing list's subscription base continues to slowly increase. We are about 3,800 people strong. On the other hand, a robot is the most frequent poster to the list. Signs of the times? :-D
Eric Lease Morgan
Navari Family Center for Digital Scholarship
University of Notre Dame