3

What tools do you use to manage co-authored papers? This is similar to questions active 4 years ago and 5 years ago. I assume tech has changed since then.

Co-authored projects I have been on have used Dropbox to share files or shared files through many emails. This does not seem particularly efficient.

Update: In my field, almost everyone uses MS Word.

I use Trello for solo-authored work. It has the option of inviting collaborators but I have not tried it. I am also interested in reproducible workflow and have some of my code on Github. I am exploring the Open Science framework, too, but have not used it to manage an entire co-authored project.

Improve this question
  • A shared network drive and Word's "track changes" features works very well in the majority of the cases. Not every task requires new technology to be efficient. – Cape Code Oct 16 '18 at 18:59
  • 1
    Wait, are we speaking about actually writing the paper, or maintaining a to-do list? Because Trello doesn't help you write the paper, from what I understand. – Federico Poloni Oct 16 '18 at 18:59
  • 2
    I and my students use Overleaf. – Prof. Santa Claus Oct 16 '18 at 19:01
  • @FedericoPoloni I agre Trello works just like a time management system or to-do list. But I think it is possible to connect the Trello to a Github repository to track your changes. – Alone Programmer Oct 16 '18 at 19:06
  • 1
    On Stackexchange the usual way to ask for an update of an existing question is to comment on the question, and perhaps leave a bounty (the latter isn't really an option for new users, though somebody else might choose to help out). Asking another question the same is considered a duplicate. I'm not sure I personally agree with this approach, but that's how it is at the moment. – Flyto Oct 17 '18 at 14:58
3

The best option is Github I believe. But it pretty much depends on what do you use for writing papers? Do you use LaTeX or other typing softwares like Microsoft Word or LibreOffice which is the free and open source cousin of Microsoft Word. If you use LaTeX, which I recommend it personally and probably many other people here too, you could create a repository in Github and put your tex files as well as bibliography files and then simply you could share that repository among your collaborators. The reason why I like LaTeX instead of using Microsoft Word is that tex files by themselves are human readable files in comparison to binary .doc files. As a result, you could track and maintain your version control for your paper like a code. Otherwise it's not easy and trivial to track binary files (e.g. .doc files) by Github or other version control softwares (e.g. Data Version Control (DVC)) Also, it is possible to link that repository to OverLeaf and that makes sharing/writing much easier. But for more accurate solution in order to fulfill your needs you need to mention what is your main writing software for preparing papers.

Improve this answer
  • Can someone explain why downvote?! – Alone Programmer Oct 16 '18 at 19:09
  • I didn't downvote, but to me it seems very wall-of-text-y and hard to read. I don't blame the downvoter. – Federico Poloni Oct 16 '18 at 19:44
1

Our lab uses Box, largely because our institution has purchased a license for it.

I wouldn't necessarily recommend it as an ideal solution, but one advantage is that it saves a history of past versions, so if two people make edits at the same time at least you can go and get the prior version to merge them, though it doesn't make this obvious or 'automatic' the way a git-based solution would.

You can also lock files while you are editing them, but it's not particularly obvious and it requires people to actually manually lock the files, which I find people don't usually follow through with.

Outside the lab, I mostly do collaborative efforts (that are documents rather than code) with shared Google docs.

Improve this answer
  • 1
    But let's say you want to share your work with a collaborator which not necessarily affiliated with your university or institution? I don't think someone outside of your institutional network could access your Box folder. But the main advantage of Github is that you could share it with people even outside of institutional network. – Alone Programmer Oct 16 '18 at 18:52
  • Besides the accessing difficulties of outsider people, the version control that Box use for binary files is not useful for papers cause the difference between current and previous versions are not comparable generically like Github when you maintain the version control for tex files. I think the Box is a good option to share the final constructed PDF file with other people not while you are actively changing the paper, when maintaining the version control is a crucial point. – Alone Programmer Oct 16 '18 at 18:58
  • @AloneProgrammer I agree it isn't a perfect solution, but neither is git. Mostly the collaborations are internal to our institution. The question asked what tools we use, not necessarily which tools are best :) The version control with box works fine with binary files, it's really no different than Git would be for binary files: you can access an older version, like one saved by another user, and manually merge them using the merge tool in Word. I've had enough trouble convincing people to use git for code, converting everyone to LaTeX from Word is unlikely to be a success. – Bryan Krause Oct 16 '18 at 19:00
  • Well..., I agree to some extent cause people, who are not familiar with coding stuffs, have some fear that LaTeX looks like a coding software. – Alone Programmer Oct 16 '18 at 19:02
1

In my experience, the important part of writing things jointly is to give everyone the time and space to write at their own pace. This means that one of the most awkward things to do is to send around a single word document to which everyone then adds a version number. That's because this mode means that only one person can write at any given time -- independent of whether that person has the time to do that right now, or whether someone else actually just had a good idea worth putting into the document. It also inevitably leads to situations where someone edits an older version and then some poor soul has to figure out what the changes were and port them forward to the current version -- further blocking everyone else's progress.

So whatever mode you adopt, make sure you have a way to ensure that everyone can write whenever they want. MS Word documents just don't allow for that, but that's not a problem: There are many other ways you can use, and in some of these cases the final version can be converted either into MS Word, or into another format that can then be submitted to whomever needs to get the document in the end. Examples that I have used in the past are:

  • For things that don't require extensive formulas, just use a Google Document. Everyone can add and edit whenever they want to add and edit, including concurrently while someone else is working on another section. The final version can then be exported in MS Word, OpenOffice document, or PDF format.

  • If you need formulas but don't know latex, just use Overleaf.com. Again, everyone can add and edit whenever they want and concurrently if necessary. The end result can be exported as PDF.

  • If you know latex, you can also just work on the plain latex files and share them via github. Github has mechanisms to resolve conflicts if several people started writing at the same time, starting from the same version, but commit them at different times.

Of these options, github is probably the most awkward mechanism because your changes are not recorded immediately, but only when you upload them (possibly causing conflicts with changes others have made). The advantages are that you can work offline, and that you retain a perfect version history if you ever need to find out who wrote which line. The latter seems to rarely be a concern when writing papers, proposals, presentations. The former is a concern, but is becoming less of one as everyone tends to be online all the time these days.

Improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.