This is no longer developed with last commits in 2011.

PapersDB is an elegant scientific paper management webapp similar to Mendeley or Papers but operating completely in browser.

This was my entry to MIT's 2011 iCampus prize. The submission was well-received but did not win since it was more geared towards researchers rather than addressing living and learning topics at MIT. I meant to further develop this app for my personal usage, but my doctoral research projects took over any free time. I moved to Mendeley instead, and it works pretty well.

Home	Login	Sign up
Dashboard (blank)	Add URL automatically	Retriving paper information...
Automatically filled information	Add a paper manually	Successfully added paper
Paper entry view	Edit paper entry

1. After cloning this repository, copy local_settings.py.sample to local_settings.py.

2. Build the Dockerfile. Note, this Dockerfile uses SQLite and Django's development server:

   docker build -t mikexstudios/papersdb .

3. Run it like:

   docker run -d -p 80:80 mikexstudios/papersdb

   If you want to develop while running the script, mount the current directory by:

   docker run -d -p 80:80 -v `pwd`:/usr/src/app mikexstudios/papersdb

As scientists and engineers, reading scientific literature is part of our daily routine. However, these papers start piling up and keeping track of all of them is a chore.

One solution is to keep a digital library of papers. This concept is not new, and there are many existing software and websites to address the paper management problem. However, none of them do it well. The software that gets it mostly "right" is "Papers", a mac software that provides a minimalistic but powerful interface for a papers library. However, it is platform dependent (only mac), does not have the ability to sync libraries across computers, and suffers from weak paper searching ability.

Some of the mentioned deficiencies can be addressed by putting the software "in the cloud" (in other words, making a web application). However, current web-based solutions lack the usability and feel of a well-designed web application. For example, zotero is a "web application" that tries very hard to mimic a desktop application! Because it attempts to bridge two disparate design principles, zotero fails to be good at either. CiteULike is a very functional web application, but lacks usability. Mendeley lacks focus and sacrifices its online interface for social features and a kludgy desktop client.

Therefore, there is a need for a better paper management web application.

PapersDB's focus is building an excellent and elegant paper management web application. It attempts to reach this goal by:

1. Having less unnecessary features. Components such as "social networks" and desktop clients is unimportant. The most important feature of PapersDB is to manage papers well.

2. Designing a good user interface. The software should be minimalistic and get out of the user's way. The papers library should be the most prominent part of the application. Users should be able to immediately use the application because it is intuitive.

3. Designing from a web application approach. A big mistake that destroys the usability of a web application is to design it like desktop software. An example would be Microsoft's Web Outlook vs. Google's Gmail. The desktop version of Outlook is a usable program. However, Microsoft tried to apply the same desktop design principles to their web version of Outlook. The result is a confusing and unresponsive application that lacks elegance. In contrast, Google's approach to building an online email client is to start from the browser and design the application to look and feel like a website. Similarly, PapersDB approaches the user interface from a web point of view, avoiding browser limitations while capitalizing on browser strengths.

As a web application, PapersDB can be accessed on any computer with a modern web browser. Users can work from any computer/platform, and all library changes are synced automatically between computers. Users can search for papers like they naturally do on Google or on the journal websites. Importing papers into the website is envisioned as a single click from the browser.

The following major features needs to be completed before the application is considered stable:

1. Fulltext searching - the ability to search through all citation and document text in the papers library.

2. Bookmarklet paper adding - the ability to click on a small button/link on the browser when visiting a publisher's website to import a paper into the database.

3. More robust automatic citation grabber - only certain popular journals are currently supported for automatic paper adding. This list needs to be further expanded and tested.

4. Better UI for viewing individual papers - when individual papers are clicked, the resulting page is not very useful and look crude. Improvements to the UI is necessary.