Is cite while you write functionality planned?
i.e., the ability to pull in-text citations into a document (Open Office preferably) from Paperpile, and generate a bibliography. Must also be able to choose from an extensive list of standard reference formats.

I currently manage my references with citulike. I have been brought to paperpile by gjuggler and his citeUsync application. Would it be possible to sync paperpile with citeulike. In this manner I would have the advantage paperpile (local copy and control of all my references, enhanced functionalities) and citeulike (access online from anyplace, sharing with others,...).

PS: I am a windows user and look forward to be able to try out your application (thanks for you work!)

Provide a way to enter notes about a paper within the embedded PDF viewer.

an automatic background task, e.g. whenever the program is started, so I can drop my pdfs in that folder and have them appear in paperpile automatically

I'd like to send references directly to Paperpile while I'm browsing (similar how CiteULike works)

A lot of pdfs come with messed up metadata which lead to confusing situations when you open these pdfs in a pdf viewer or load them to an ebook reader, etc. Currently, I can edit metadata using some tools such as pdftk and calibre but being able to do this from within Paperpile would be great. Can't all the fields I fill manually (author, title, year, publisher, etc.) be written into the pdf file's metadata?

XMP is the standard how PDFs store information like author, title, year, place, journal and any decent PDF reference manager now extracts this metadata from the files. No software but JabRef, however, supports synchronizing its database information with this information residing inside the files. Writing metadata as XMP to the files is standard procedure for image management database software. Paperpile would gain a significant competitive edge if they fully supported XMP. PDFs and their metadata can then much more easily be shared between different databases!

I would like to print PDFs directly from the internal PDF viewer.

originally i just wanted to create an wiki entry for ubuntu users - as there are some for the other available programs (mendeley, jabref, zotero etc.) - but none for paperpile.
But as i got an injury i passed some more time with playing around.

So - i created a small webpage/wiki using a software called dokuwiki (which is nice as it needs very little system resources).
You can have a look here:
http://kaktux.bplaced.net/wiki/doku.php

No pressure in using it - it´s simply a plaything. If you like it fine - if not it´s fine to.

I would like to use the RSS feed option but already have a reader with many subscriptions, is there a way to import OPML files so I don't have to resubscribe to each one. Maybe I missed something.

Allow the citation database to be backed-up and synced to a WebDAV store. This is a really useful feature in Zotero that allows you to share your citation library across devices without going through the (limited) zotero account.

I think it would be interesting to have some documentation aimed at prospective hackers to help them find their way around the source code.

Allow people to plugin additional search services like IEEE Explorer

I label a pdf "X" and also "Y". I want to be able to search for pdfs that only have the combined labels X and Y (both X and Y). Currently, I can only search for X and it will display all the X labeled pdfs, and then I can eyeball and see if anyone of them have a Y label as well. But with a bigger set of pdfs this is not feasible.

I add an article, and it somehow even has its DOI, but paperpile still does not know what the year, volume, journal name, etc. of the article are. Make it easier and more intuitive to extract metadata from the web and fill the fields of the article's info. Mendeley can search the title in the article in google scholar and 80% of the time, it will yield the proper info for the article. How do I do that in paperpile? Paperpile is beautiful, I will use it forever...

Importing bibtex and a folder full of PDFs tends to create duplicates. The bibtex-related entries have great metadata. The PDF-related entries have PDFs. The duplicate feature detects the overlap, but there doesn't seem to be a nice way to choose the bibtex entry as the "good" one and have it acquire the PDF attachment.

It would be desirable to support the DJVU format.

Create an Android application for tablets and the such so that users can read, search, and mark up their research papers on the tablet device.