Charles Basenga Kiyanda

classifying papers

Every few months I ask myself the question again: “What to do with my scientific library?” I’ve run across solutions here and there, bu none which convinced me: jabref, zotero, papers, etc.

I actually tried jabref, but it struck me as not having many functions beyond holding a list of all my papers and generating a bibtex file. Zotero is hosted on servers far away somewhere and my work doesn’t allow this. (Well, they might allow it, but to be prudent I host all work related files locally.) Papers seems awesome, but is unfortunately mac only. The situation seemed desperate until I saw gPapers! I thought I finally had found THE program of my dreams. Alas, gPapers is rather unfinished. The person who developed it originally doesn’t seem to have contributed in a while. What to do? Well, contribute of course! Though I have been known to file bug reports and test alphas, I’ve never contributed actual code to an open-source project, so this is my first experience. It seems like the timing and circumstances are perfect for me to get involved:

  • gPapers is written in python, which I’ve taken the time to learn lately;
  • I’m tired of looking for this perfect software and have gotten frustrated enough that I’m willing to write it;
  • right now, the feature-set of gPapers seems contained, so the initial deciphering of the code shouldn’t be impossible (it’s not like I just decided to rewrite the whole linux kernel);
  • there are few developers and the pace is slow, so I should have time to get up to speed without having the software go through three major revisions.

The only problem is: the active fork of gPapers is hosted on github. I have no idea how git works, let alone github. (I DO use version control, but I’m a svn user right now and have no experience with distributed version control.) I managed to create an account on github, create keypairs and upload the public key to github and…that’s pretty much where I was as of this morning. After a few days of looking, I came across a great tutorial containing the information I had been looking for so far. It basically answers the question I’ve been stuck with for a while: “Ok, I’ve created an account on github and now I want to contribute to a project. What do I do?” (spoiler: the answer is basically that you fork the project, as surprising as it may sound to svn users.)

Getting started with GitHub from Charles Max Wood on Vimeo.

As for gPapers itself, I got the software to run a few days back and now I’m off to see if I can actually make it work. Let’s see how much has to be done to make gPapers better than papers.

[Update 11/17/2010: I was corrected by a commenter. You can use zotero entirely locally. I’ll give zotero a whirl again sometime in the next few days then. I’ll probably still try to contribute to gpapers, be it only to learn how to use git properly.]

2 comments to classifying papers

  • Sebastian

    you’re mistaken about Zotero. By default, Zotero stores all of your data locally (the default is in your FF profile, but that can easily be changed to any location on your harddisk).

    Except for features like data sharing etc. that by definition require an online component, you do not need to store your data on any Zotero or other server.

    You do not even need a Zotero account to have access to all of its functionality (except, again, any features that by definition rely on data exchange).

  • Sebastian:
    You’re right. I looked at zotero a while back and I remember looking at a video which made a big deal about the syncing ability. I must have been mistaken as the zotero website is pretty clear now that you can work entirely locally.

    Thanks for the comment.

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>