Brandon Invergo

Damn It Mendeley

Mendeley is one of the last couple of non-free programs that I depend on, simply because I have not found a suitable alternative yet. And now here I am taking time out of my extremely busy day to vent about it. God damn it, Mendeley, I can't take it anymore. I'm done with it.

To be honest, I'm not 100% sure what Mendeley is doing behind-the-scenes (part and parcel of its proprietary, closed-source nature) but it appears that, when it comes to filling in the metadata for an article, it not only uses reputable sources such as PubMed but it also refers to what other Mendeley users have filled in for that article. If this is true, it is a horrible idea because, as it turns out, most scientists are idiots when it comes to this. I'm not prone to using internet catch phrases, but every time I import a new article into Mendeley, I'm forced to mutter "Oh, what fresh hell is this?"

Some examples:

  • A period is automatically appended to the article title. Sorry guys, that period is part of citation formatting, not part of the title. Let the citation style add it.

  • Shortening of page numbers (i.e. 517-19 instead of 517-519). Sorry guys, that numeric shortening is part of citation formatting. Let the citation style do it. For the occasion that the citation style calls for the full page numbers, you force me to check visually every single citation to make sure someone didn't shorten it because they think it looks cool.

  • Adding publisher location after a journal name. When I cite a paper from Science, I don't want the citation to read "Science (New York, NY)" because the journal is just called Science. Add your unnecessary metadata somewhere else. I'm tired of having to remove it so that my citations are correct.

  • Adding society information after a journal name. I don't care that the Journal of Neuroscience is the official journal of the Society of Neuroscience. Once again, put your unnecessary metadata somewhere else so I don't have to remove it just to have correct citations

  • Journal names in sentence case. Seriously, this one is a shock to me. Article titles are in sentence case. Journal titles are in title case. It's not difficult. Now I have both "Journal of Neuroscience" and "Journal of neuroscience" in my journal filter and once again I need to manually intervene on the vast majority of files that I import before I cite them in a manuscript.

  • Multiple variations on a journal name. My database has articles from PNAS, Proceedings of the National Academy of Sciences, Proceedings of the National Acadamy of Sciences of the United States of America, Proceedings of the National Acadamy of Sciences of the USA, and any number of variations depending on creative capitalization of these.

  • Multiple variations on a journal name for the same article. (ie Proceedings of the National Acadamy of Sciences : PNAS)

  • Here's a new one that prompted this post: adding authors as "Last name, first name middle initials full initials" (ie Smith, Joe M JM). Great, so now when I cite it in a manuscript, I get Smith, JMJM. That is absolutely wonderful. So now I have a new thing to manually edit every time I import.

  • Author initials when the full name is available. If the article lists the full names of the authors, please use them. Again, let the citation style handle initials

  • Mixed case author initials! Are you kidding me? One of my own articles had middle initials in lower case ("Cock, Peter Ja", "Chapman, Brad a") (note the journal title while you're at it). What an embarrassment. But don't worry about consistency because my name was listed as "Invergo, Brandon M". One out of three correct isn't bad, right? Right? (the fourth author had no middle initial)

  • Mixing initials with and without periods. Of all the points so far, this one's probably down to personal taste; if the full names aren't available and you put the initials, you may choose to put periods between the letters of the initials. But for the love of whatever you find sacred, don't use periods in some of the author's names and then not use them for some authors in the same damn article.

I'm sure I've encountered more annoyances but that's enough for now. What is the point of automatically extracting/fetching metadata if you have to manually edit it every time? I've wasted far too much time just trying to standardize all of the articles in my database.

The most asinine part of this is that the only way to provide such feedback to Mendeley is through some feedback system that depends on votes to get your suggestion seen. Needless to say, back when I still had hope, I suggested that they filter metadata. That was two years ago and it has only garnered 9 votes, 3 of which are my own. This is obviously a huge flaw in their software (everyone around me complains and curses at the software constantly), but for the suggestion to get noticed it has to compete in the arena with more important suggestions like allowing to highlight in colors other than yellow. Who decided that this was a good idea? Why should bugs have to compete with feature requests? Why not use a good, old-fashioned bug tracker?

So like I said, I've had enough. Good riddance to Mendeley. After this current manuscript, it's being removed from my harddrive. It's looking like I can use cb2bib for metadata extraction into bibtex files and then bibus, pybliographer or referencer for working with those files. But whatever I do, I can no longer place faith in the basic citation capabilities of my peers. Sorry guys/gals.