Saving group libraries on a WebDAV server

For security and simplicity purposes, I would like to keep all my articles, and their attachments in one place.

I already save attachments related to My Library on a WebDAV server. Is it possible to do so for Group LIbraries? If so, how?
  • edited January 22, 2014
    I think that Zotero is a great tool. Way, way better than Mendeley (I cannot trust what is going on with that tool, after years of using it).

    However, this (lack of private storage for private group libraries) is a real big problem with Zotero. Almost an utter dealbreaker. I hope that the comments below are taken seriously because they are meant as a genuine criticism from someone who loves Zotero and is actively promoting its use locally.

    And its not just a matter of money. Read on:

    I want to use Zotero as an article sharing platform for my collaborators and my students (using private storage for obvious legal and privacy reasons). As you may imagine, my group libraries will grow both in number and in size as students work on different lines of research in my group.

    The highest amount of storage that Zotero offers is 25 GB. This is utterly puny. I have thousands of articles in my personal Mendeley library that I am slowly transitioning over to Zotero. The size of the watched folder on my disk where they are stored (representing about 14 years worth of doctoral and post-doctoral / scientist work) is about 27 GB. I hope to have a career that lasts another 30-40 years. If my students (lets assume I have an average of 6 students / postdocs in any given year) add articles at the same rate as I did (27GB/14 approx. 2GB per year), I can expect about 12GB of accumulated article data per year. Assume that we get rid of even 10% (even that is a very unlikely figure - who wants to lose access to citations to work that they have previously used?), that still leaves at least 10GB of accumulated articles per year. Assuming that PDF remains the primary mode of article storage (and that is changing, in some fields, quite rapidly), I am looking at an additional 40 x 10 GB = 400 GB by the end of my career.

    I can already not fit my own library into the group libraries (I want my students to have every advantage of my past lit. surveys) into the biggest storage solution you offer. So, even if I had an unlimited budget, your best storage plan would not work for me, and I suspect, for any faculty who really wanted to use Zotero as a heavily used research tool in the spirit it was perhaps intended to be used. I want my students to stop printing out papers, or email PDFs back and forth to me. If one has any social conscience, one needs to stop printing out articles on paper. Emailing PDFs is just idiotic and is no better than the pre-Mendeley/Zotero situation that existed for most of my doctoral work.

    My colleagues in biological sciences (I am a materials/device scientist, if someone is wondering) have even heavier space requirements. I would not be surprised if they use twice or even more the amount I use (you just have to pick a random paper in, say, biochemistry, and just count the number of citations).

    Now, this is where money, and more importantly, legality does come in - your storage prices, compared to zero cost ownCloud implementations that most Universities are offering increasingly, often in response to government mandates, or the pay what you actually use services like S3, are utterly exorbitant. Its a practical matter - I cannot justify this kind of expense in my research funding when I am being given nearly zero cost practically infinite storage right on campus, where I do not have to worry about sharing of copyrighted work (such as University-paid subscriptions to journals) outside the University paywall.

    Our University has commissioned a WebDAV enabled ownCloud server, with practically unlimited storage, which I and many other colleagues are using for our own libraries on zotero. However, if the group library function is crippled the way it is, Zotero basically becomes a toy for collaborative work, not a serious tool.

    Long story short - Zotero does not scale as a collaborative tool. This is a basic problem of functionality.

    One expects better from Zotero (given that Zotero is a tool created, and thankfully, overseen by my fellow academicians, unless the character of Zotero has changed).

    I have followed some of the discussions on this topic:

    In the above thread, some of the reasons offered by the responders are to be frank, in light of the above, specious.

    Crappy user experience, as opposed to what? No collaborative experience at all??

    Further, when I store my library (not group) on our WebDAV enabled ownCloud servers, I still access them through Zotero. I never have to even think about the storage form on the server. The comment by Dan Stillman (Mar 20th 2010) about having to access directly is misleading. Any collaborators of mine, if you allowed group storage, will also use Zotero as their interface. Am I missing something from that comment?

    I find the last comment on the second thread a bit more hopeful but its not clear what has come of it.

    Are there any updates on this heavily requested feature?
  • As Dan says, WebDAV for groups is not on the near-term horizon. Linked files for groups probably are, so that would likely provide a viable group-storage solution outside of Zotero.

    Beyond that, you'll have to accept the fact that Zotero isn't going to spend time and energy developing features that provide massively worse alternatives to functionality it already provides (i.e. group storage). (Storage amounts over 25GB are available on request. As for the costs - money for development has to come from somewhere. Zotero is non-for-profit but it has to remain sustainable)

    And it's, quite frankly, bad style to call arguments "specious" and "misleading" when you're clearly not aware of the technical background: The whole point that Dan is making about user experience is that even when users are all using Zotero, they'd have to deal with accessing, unzipping etc. files like directly when dealing with WebDAV as a group storage. Unless you want to make specific code suggestions, you'll have to take his word for it.

    As for ajlyon's comment in the last thread - there is nothing to "come of it" - Zotero's server code is open and available, people are free to implement a local instance of the Zotero server.

    (Disclaimer: I don't work for Zotero and while I'm pretty confident this is an accurate description of both the technical and the logistical aspect, it is in no way an official statement)
  • The comment by Dan Stillman (Mar 20th 2010) about having to access directly is misleading. Any collaborators of mine, if you allowed group storage, will also use Zotero as their interface. Am I missing something from that comment?
    Yes, you are. I was responding to a very specific suggestion by Gitty K in that thread:
    if I were able to see in the web-interface the actual file-name of my attachment as it is stored on the webdav-server, I could access it there
    That suggestion, and my response, had nothing to do with accessing WebDAV-based group files via the client.
  • And otherwise, what adamsmith says.

    In the not-too-distant future you'll likely be able to use linked files and library-specific attachment base directories in groups. As I say in the thread you linked to, with some help from the OS or third-party tools you could probably use your existing WebDAV storage with that.

    WebDAV support for groups will probably happen eventually, but the UI and sync functionality would have to be made to support separate WebDAV URLs and credentials for each group, and all group members would have to be given those out-of-band and would need to enter them properly (though the URLs themselves could possibly sync). Web-based access wouldn't be possible, since attachments are uploaded compressed (to avoid various other technical issues) and won't have access to the WebDAV servers to download and process the files. (But access from also isn't possible for linked files, of course.)

    Implementation of either of these will depend on either our getting around to them or people submitting patches (which is how attachment base directory support for personal libraries happened). Since there's an existing solution that provides a better user experience than either, these aren't top priorities for us.
  • Thanks for explaining all that.

    It is indeed more complicated than it is for My Library.

    Very often when I use web apps, I find that cloud-based options like Google Drive, ask me for permissions to allow web apps to access certain details related to my Google account.

    I suggest that maybe, when you do get around to doing this, you should look at ownCloud (which is rather popular in University infrastructures) the same way.

    Further, I am afraid that I must take issue with your last statement. I think that the priority of tasks should be driven by actual user-demand, and not your judgment of what users will like.
  • I am afraid that I must take issue with your last statement. I think that the priority of tasks should be driven by actual user-demand, and not your judgment of what users will like.
    Sorry, but that's not how this works.
  • Is there some new possibilities from now ? Post update.
Sign In or Register to comment.