Arxiv paper with lots of revisions and changes made to subject matter; shouldn't Arxiv clean this up?

Question

In the domain of "Computer Science > Programming Languages" American Computer Scientist Carl Hewitt has put up a paper that changes completely from the first revision to the current one:

From

• v1, July 2009: ActorScript(TM): Industrial strength integration of local and nonlocal concurrency for Client-cloud Computing

to

• v46, February 2015: Inconsistency Robustness in Foundations: Mathematics self proves its own Consistency and Other Matters

... which seems to be outside the subject of "Computer Science > Programming Languages".

The entry also goes through several revisions along the way, notably:

• v35, July 2010: ActorScript(TM): discretionary concurrency for privacy-friendly, client-cloud Computing
• v37, August 2010: iScript(TM) extension of Objective C(R): discretionary concurrency for privacy-friendly, client-cloud computing
• v40, September 2010: A 0-sized withdrawn entry superseded by another paper entirely: ActorScript(TM) extension of C sharp (TM), Java(TM), and Objective C(TM): iAdaptive(TM) concurrency for antiCloud(TM) privacy and security which itself has 60 versions.

I will not discuss the merits of v46, as I am interested in the contents of v1, which is also meant to be found when going through links in other web-based publications discussing Actor approaches to concurrent computing.

But it is confusing and looks irregular. Shouldn't arxIv clean this up?

Answer 1

It does indeed look confusing and irregular. Oh, and it only gets worse once you start looking into how interrelated it is with different versions of some of his other arXiv postings... In an ideal world, perhaps arXiv should clean this up somehow. However, the moderation team is relatively small, works on a voluntary basis, and has to handle a lot of submissions every day. One really can't expect them to sort out messes like this one, nor do I think it'd be good for them to have an extensive editorial function. Assuming the initial submission meets the criteria, the onus is on the author(s) to communicate clearly - not on arXiv.

It's also worth noting that only the first five versions are included in the daily listings, so individual authors making lots of revisions don't necessarily impact the community too much. Given all this, I think it makes sense for arXiv to focus on rapidly processing new submissions instead of worrying about excessive changes to a (presumably) small number of postings.

Finally, I'll note that in cases like this it's a very good thing that arXiv stores all versions. You can cite a specific version and leave it at that.