PeerJ EC2

[kyleniemeyer]

Another concern is support for readers. This seems a weakness and should be discussed, with plans articulated as to how it will be addressed. It would be nice to know who reads JOSS works, what they do after, if there are statistics on cites resulting, if there are repeat submissions, if newer submissions cite older ones, etc.

[kyleniemeyer]

Similar to #25 and #28

Some of us have discussed this; one idea was to tag the papers based on language and topic area. Plus, we need to add some sort of searching/discoverability tools to the website.

Perhaps putting some concrete plans in the paper will help ensure that we pursue them!

[karthik]

Perhaps putting some concrete plans in the paper will help ensure that we pursue them!

I also feel like it's ok to note that it's early days for us. I can't imagine that a large number of people will just read JOSS cover to cover but that they use a library, see the JOSS citation in the README and cite the JOSS article in their research paper. That is the primary purpose of JOSS - an advertisement for the software itself.

[danielskatz]

I'm not so sure this is a real issue. If the purpose of JOSS is to enable credit for software, we should focus on enabling credit for the software that researchers use, not on how researchers find software.

I don't mind the idea of adding search/tags/etc., but don't feel like this is a high priority.

[kyleniemeyer]

I agree—adding some basic tagging and searching functionality would help, but since we serve the broad research community rather than a single domain, I struggle to picture the use case where people would go to the JOSS website looking for software. Perhaps that is a possible future direction, but that isn't really our main goal.

[kyleniemeyer]

(in other words, I agree with the use case @karthik described as the most common one)

[arfon]

Examples of multiple submissions:

• http://joss.theoj.org/papers/10.21105/joss.00026
• http://joss.theoj.org/papers/10.21105/joss.00365

(And yes, they did cite between)

[kyleniemeyer]

I think it's helpful to mention these occurrences of new versions citing earlier ones, so I've added this to the paragraph on citations in the conclusions:

We have had at least two `repeat'' submissions, i.e., submissions of a new version with major changes from a prior version. Clementi et al. [41] publishedPyGBe-LSPR, a new version that added substantially new features over the originalPyGBe` of Cooper et al. [42]. Similarly, the software published by Sandersen and Curtin [43] extended on (and cited) their earlier article [44].

[kyleniemeyer]

Response to this editor comment:

We have clarified that the journal is mainly designed to serve authors\slash software developers. However, we do plan to add better search features, including tagging based on language and domain, which would make finding software\slash articles easier. However, JOSS is not designed to be read as a typical journal, as the primary focus is to create citable entities for software. We do not have any information on readership for this reason.

We have added some preliminary (and incomplete) information on citation statistics to the conclusions, and also evidence of two articles that cite older versions.

[arfon]

👍 LGTM