Fantastic reference managers and where to find them.
Tl;dr: After years of painful shift between reference managers and months trying to find the holy grail, I decided to take the plunge. I made a choice and decided to write about it, hoping that nobody would have to go through all this ever again.
I am a lover of desktop apps, and I am currently running macOS. For this reason, this blog post will deal entirely with desktop reference managers available also on macOS. If you’re happy with your web based one, or you’re running Windows//Linux, this post is not for you and deep down I envy you!
Four years ago, when I got a placement as a research assistant in my current research lab, I was introduced to the concept of reference managers. At the time, having had no experience whatsoever with this kind of software, I just followed what my boss recommended: Papers 2.
Papers at the time was a great piece of software. Even though it was not free (If I remember correctly I payed £30 taking advantage of an educational discount) it had everything I needed. It was a sleek macOS-style app that worked seamlessly with Word. With a double tap of the [ctrl] key, I was able to call the Papers 2 assistant, type the name of the paper wanted and stick a properly formatted in-text reference. Once I was done with the document, I would compile it and Papers 2 would magically create a Bibliography at the end of the file.
It was just perfect.
Everyone knows that, unfortunately, perfection does not last long. With the advent of Papers 3, a newly updated version of the same app that shifted to a subscription based model, Papers 2 was slowly forgotten and started becoming buggy. So buggy that, in the last weeks of use, underlining a paper would always cause it to scroll back to the beginning of the article. Papers was not serving me well anymore, and thus I had to abandon it.
Not willing to move to a subscription based model for a reference manager software, I started looking at free alternatives. To be completely honest I don’t know how I stumbled upon Mendeley, but that was the reference manager I chose to move to.
Mendeley was not fun from the beginning. While Papers lured me in with a sleek app and failed me over time, Mendeley’s appeal was solely the fact that it was free. The interface was bad, the resolution was horrible on retina displays, it was buggy and not pleasant to use.
But it did the job.
At that time, 2 years ago, I started using LaTeX more and more often. The ability of Mendeley to output a bibtex file and constantly update it was one of the strong points of Mendeley, and the one that got me to stick with it.
Unfortunately this period didn’t last very long at all. Mendeley’s bugs kept getting in the way of my work: UI scaling badly when searching through the library, random crashes on a daily basis, highlighting issues, metadata issued. But nothing disturbed me majorly, I powered through.
This was until the Elsevier scandal. This scandal made me seriously reconsider Mendeley as a tool fighting against the concept of open research. Interestingly enough in the same period of time, two colleagues of mine had major issues with their PhD submission due to Mendeley screwing up their references at the last minute.
That was the moment when I knew that I had to move away from Mendeley.
The Final Move
In that moment, I had just started writing my PhD dissertation, and I was in desperate need of a reliable reference manager. I was determined to spend money again to get a reliable software, but I was not willing to spend it on something that would have been sunset anytime soon. This time, I had to do some research. The alternatives were:
- EndNote – Academia’s favourite. Expensive but renown.
- Bookends – The niche alternative. A Papers 2 throwback.
- ReadCube – The company that bought Papers.
- Zotero – The real Mendeley alternative.
While ReadCube’s web app seemed sleek enough to make me want to try the desktop version and invest some money, the desktop app seemed to be in beta for the past century. The company had been silent on social for months, and the online support wasn’t of much help.
Because I needed a reference manager, and I needed it fast, ReadCube was dropped immediately as a possible alternative.
After ReadCube, I went straight to EndNote. I was willing to pay a premium to get a reliable software but after a quick trial version of the EndNote X9, I was faced with a whole bunch of problems.
First and foremost: an amazingly unintuitive interface. Menus in menus hidden within menus. An overall bad UI, amazing features hidden instead of being highlighted: needless to say I was not impressed.
However, I powered through. I contacted the customer care multiple times to have some features explained, I setup the environment to tailor my personal needs and I was ready to give it a proper try.
Ding ding ding – EndNote wants to update!
Feeling strong about my newly set environment and excited about new features, I update my trial version of EndNote. Everything broke. All the LaTeX features that I needed from EndNote were not working anymore: bibtex customisation – not working, menu items – not working, bib file key editing – broken.
Googling didn’t help either. Turns out, EndNote hadn’t been treating its customers very well. Since version X7 the customers felt that with every update the software became more and more buggy. Moreover, the 32bit nature of the app and the inability of the company to make it 64bit compatible was a looming prediction of future problems.
I couldn’t spend £100 on a software that was trailing on the edge of being obsolete.
Bookends was one bet that I was not willing to make.
Not free, but low price, I saw Bookends as an ‘independent’ alternative. The app was… ok. The menu system was cluttered and at times incomprehensibly complex, but nothing I couldn’t have learned with a bit of application.
The real problem here is that I was not willing to pay for an app that was not recognised in the field. I had no assurance that the app would have kept updating through the newest OS versions, and that it would have listened to customers when it came to bugs, fixes and features. While Bookends seemed like a good idea, I was just not willing to spend the money on it.
I discarded Zotero from the beginning because I had this weird idea in mind that if Mendeley failed me, so would have Zotero. Being both free, there was no chance that Zotero was better.
But Zotero is better.
First of all, it is truly open. Open source, open reseach: open. It is very customisable: the ability to add plugins makes it awesome for people who want to tailor the reference manager to their own needs. It outsources what it doesn’t want to do: and that’s a good thing! Zotero doesn’t care about dealing with your pdfs and how to view them, it lets you use whatever viewer you want outsourcing possible problems and bugs.
Zotero does ONLY reference management and output, and it does it well!
In conclusion, Zotero was the only valid alternative of the ones tried. I am sure that many people are having a great time with some of the tools listed above, but I needed not only something good, but also something reliable and future proof.
The open source nature of Zotero gave me hope: I trust community driven projects, and I trust the openness of Zotero. The ability to develop and install plugins makes it a really powerful tool, and with the Better BibTex addition it becomes an invaluable asset for LaTeX documents.
Thanks to ZotFile, I was also able to take advantage of my Google Drive subscription and sync all my PDFs there, keeping the Zotero repository only for text references and avoid having to pay zotero’s subscription for extra space.
I think I found a good reference manager, and it was in front of me all along.