JOSS paper?

[bwheelz36]

Hey @till-m and @fmfn

Do you guys have any interest in writing up a ~2 page paper for the journal of open source software (JOSS)?
https://github.com/openjournals/joss

I think this package would be a perfect candidate, and now we have all the testing and packaging looking very modern, the paper itself should be a relatively low lift.

One issue is determining authorship. I would have to do some work through the commit history to figure out who has made substantial contributions over the years (also if you're reading this and this sounds like you please chime in!)

let me know what you think. I'd be able to set up the initial draft and submission.

[till-m]

I really, really like this idea!

From my side, I believe #447 would improve the package significantly and I would like to have it finished and merged before submission, though that shouldn't prevent us from starting a draft. Authorship does indeed seem very tricky and a bit of a touchy subject. I'm not sure how to approach this. Are there some guidelines from JOSS maybe?

It's also a bit hard to define a "substantial contribution":

• is adding a minor feature a substantial contribution?
• is implementing a high-quality bug-fix a substantial contribution?
• is providing a lot of code reviews a substantial contribution?

I'm guessing that most software published in JOSS isn't 10 years old and has probably seen fewer contributors...

[bwheelz36]

I guess we could just tag anyone who contributed and say "It's very hard for us to know who contributed what at this stage - if you feel that you made a substantial contribution and should be an author please explain why". I'd tend towards being inclusive as well.

[fmfn]

I love this idea. Both of you have been much more active and closer to development than me for quite sometime now, so I'll defer to you who should be included. But I'm happy to assist however I can with writing / reviewing.

Quickly going through the contributors list reminded me that @Erotemic introduced the concept of target space way back when. Other than that, I think most other contributions were relatively small, but please correct me if I'm wrong.

On a similar topic, both of you should add yourselves as authors in the citation blurb at the bottom of the readme. This package is cited fairly frequently.

[till-m]

Very happy to hear that!

I spied on @bwheelz36' google scholar and it seems he submitted to JOSS before. Presumably he's familiar with the format so I'd suggest he takes charge of the process and lets us know how we can help. I can definitely contribute to writing, editing, making plots or whatever else is needed.

On a similar topic, both of you should add yourselves as authors in the citation blurb at the bottom of the readme. This package is cited fairly frequently.

Thank you for the offer! I would argue for replacing the current bibtex with the JOSS paper reference once that's out. What do you think?

[Erotemic]

I forgot that I did TargetSpace. I suppose it probably counts as substantial. In any case, having my name on a new publication is helpful and appreciated. Here are links to my ORCID and Google scholar profiles, or do you want me to make a PR with the edit to the README?

Also, I'm glad to learn about JOSS. I have a few repos that might be good candidates.

[bwheelz36]

Hi all,
all comments noted.
I am happy to take it forward, but it won't be super quick - will have to wait till I have some spare time or weekend motivation again :-P

[till-m]

In today's academia it is unfortunately true that citations matter, and it is also unfortunately true that they matter more for papers where one is listed in a more prominent position and/or if they have fewer authors. I assume this package gets used heavily in academia (considering how many issues we get here from people with institutional accounts) and, correspondingly, cited frequently. In this sense, the stakes are quite high.

However, I personally still think we should ping everyone that contributed. If think offering authorship shows that we value and appreciate the work people have done and I think making decisions about what does and does not constitute authorship automatically devalues some work. In the end, pretty much every single LoC written contributes to the project in some way.

That being said, @osullivryan should definitely be mentioned. He added the bounds transformer, which is a core feature of this library.

do you want me to make a PR with the edit to the README?

I would prefer if we add the bibtex of the paper once it's accepted, this way we only have to sort out the authorship situation once.

[till-m]

Hi everyone,

given that it's been a month without activity, I guess there's not much to add to the discussion. I think in the end, someone will need to make a judgement call about the authorship situation (probably @fmfn -- after all it's his package).

@bwheelz36 feel free to let me know if I can help in any way with the draft.

[fmfn]

In my opinion the people mentioned in this thread is broad enough of a list of authors. So I would suggest the following:

1. @bwheelz36 (since you're writing the paper)
2. @fmfn
3. @till-m
4. @Erotemic
5. @osullivryan

I'm not particularly attached to the ordering to be honest. In fact, in my graduate field, the standard is to list people alphabetically.