Share on Local Servers

We are involved in a grant project where we are trying to locate and preserve digital copies of fragile online records of news events. We will have 20-30 librarians locating these files and saving them locally using Zotero.

The challenge we’re running in to is the difficulty associated with sharing Zotero databases across several users. I have been closely following the forums/roadmap and am interested in the new features included in version 2.0 that will enable users to share Zotero databases with other users.

It appears that the shared Zotero databases will be stored on Zotero servers.

Due to security concerns, we’re curious if there will also be a way to share on a local network in version 2.0. Or, has a 3rd party plugin been developed along those lines? Does anyone have any ideas and/or guidance for our situation?

Thanks!
Matt
  • Zotero will soon offer the ability to store and sync attachments in user-defined storage (i.e. WebDAV) while keeping item metadata synchronized via the Zotero server. Would this configuration meet your needs? If not, could you perhaps elaborate on your concerns? We're committed to protecting the security and privacy of all of our users and would like to learn more about your needs.
  • I assume the "user-defined storage" options will also include storage on local or network drives?
  • Sean,

    Thanks for your reply. User-defined storage (WebDAV) would be nice... I'll echo Rintze's question, might that also include local or network drives?

    Our concerns deal with the subject matter we are collecting, not the security and privacy offered by Zotero (we trust you). This news release of the grant describes the project in a bit more detail - http://www.lib.utexas.edu/about/news/bridgeway_grant.html.

    Out of curiosity... any plans to also allow us to store and sync metadata in user-defined storage?
  • I assume the "user-defined storage" options will also include storage on local or network drives?
    Could you clarify the request here? Zotero already stores files via the filesystem—whether those files are stored locally or on a network is up to you.
    Out of curiosity... any plans to also allow us to store and sync metadata in user-defined storage?
    We've discussed the possibility of offering the syncing of notes to user-defined storage in addition to attachments, since notes might contain information that users or organizations might want/need to keep on their own networks.
  • My request was about sharing/syncing files, instead of the storage of files within a local copy Zotero. Within my research group, files are mostly shared via network drives, instead of via any WebDAV solutions. In this setting, it would be easiest, if I wanted to share the files I collected in Zotero with other Zotero users, if I could just appoint a network drive for sharing/syncing purposes.

    In my turn, I would like to echo mlisle's remark. Due to the confidential nature of the work in my research group (most research is performed in collaboration with industry, and we want to avoid giving away research leads), we would like to keep the information we collect as private as possible. This also extends to metadata, as even the specific collections of papers we collect could reveal our research focuses.
  • I just thought of another case were the ability to sync files via local storage would be very handy. Imagine that you have prepared a Zotero collection at home with 500 MB worth of files. If you then want to sync that collection with your Zotero installation at work, and you don't have a (speedy) internet connection, it would be very convenient if you could sync your files (and perhaps also the metadata) via a USB thumb drive or other portable storage.
  • In addition to WebDAV, we're also now moving toward allowing for the specification of a local or networked storage location to synchronize attached files. Thanks for the suggestion. We do not have any immediate plans to develop and distribute a local Zotero server for the local synchronization of metadata.
  • Sean,

    That's great! Any chance that will be included in 2.0?

    re: Metadata - I'm guessing we'll still have the option to export the metadata from the zotero servers to a local location. That true?

    Thanks again...
  • I would like to echo Rintze's comment about the need for confidentiality. I am in a similar situation where we do research in dug development. I can imagine other people from the medical and legal fields also will have concerns about privacy and confidentiality with the data (even the metadata) stored on remote servers controlled by third party. I hope the zotero developers will consider allowing local installation of the Zotero server.
  • While I trust you Zotero guys, I would like to echo the need for syncing libraries (and not just attachments/notes) to a local location (be that webDav or a local server). Some areas just won't be able to use Zotero for legal or research policy related reasons. (I know I would have a hard time convincing my boss to let the library be stored outside the University network, although I might convince him in the end)
  • I'll echo this as well. We may end up using Zotero on a classified network that's not connected to the public Internet - hence synchronization has to be done on a local server.
  • I am from a company research department and we have two mayor problems with using zotero.

    One is a handy duplicate detection and merge that we miss extremely in our group.
    We have all different bibtex files with small differences and its pain in the ass to eliminate duplicates!

    The second is local sync server. We can't use internet server. No way :-(

    Without duplicate elimination and sync, we can't handle our research with zotero.

    However Zotero is really nice, great job!

    But Please, Please add these two important things.
  • I concur. While I like the idea of creating a Social Bookmarking sort of thing out of Zotero data, it would really be nice to be able to sync both files and metadata to our own server.
  • One last thing, I think Zotero is great, and am actively trying to get my department at the University to adopt it. Still the two features mentioned above would really be great. Especially within research groups.
  • i think the same way there is a need for some kind of local server or at least a way for implementing one , we want to use our liferay portal webdav features for our users to be able to sync their data to their accounts and presses their data for them , this is perfectly possible for attachments but not for the metadata part which is actually the important part of the data , is there any document on zotero data syncing or some one which can help us to sync zotero metadata , so far i understood that zotero metadata are also synced with a secure webdav account but the url cant be changed from inside the interface ,ill be needing some guidance to be able to develop my way through .
  • I'd like to cast my vote for this too. We would like to use Zotero within the research group, but having the data send to another server that we don't control is unfortunately not an option. It's not that we wouldn't trust the Zotero developers - after all, we install their software without really looking at any source code - but data can always leak in some ways.
  • In my research group we use aigaion to have a shared library. I'd love to use zotero instead because a lot of things are so much quicker. But we definitely would need a share on a local server to use it in our group.
  • You should be able to use quick copy/export to create files to import into aigaion. For refbase, we have experimented with an extension that allows you to select records in zotero to upload & something could presumably be done for aigaion. Finally, note that there have been announced plans to make embedding zotero server content in your local website.
  • I strongly agree with all those who ask for a possibility to have a local server. We too would only switch to a solution where we control the server. That's not a question of trust. We simply need the flexibility and security that comes with a dedicated server (e.g. we would need the information even if we loose our internet connection).
    Until now the lack of a server prevents us to switch to zotero (even if we really would like to).
  • Until now the lack of a server prevents us to switch to zotero
    What program do you use now? There aren't too many reference managers that are client/server apps where you can install the server locally...
  • At the moment we are working with Bibus that connects to sqlite and mysql databases. We use both versions depending on what we need and how many users have to work together. This solution may is not as convenient as Zotero in some aspects but gives us what need most: An easy solution to share references and work together with the same references.
    We had a discussion here about Refworks too. But our conclusion was that we will not hand over our precious data and research to a server we do not control. And as far as I can see this will remain so for a few years to come.
    Because of many conversations about this topic and 2 big threads in the Zotero forums I think that many users share this point of view.
  • Just a thought, which may not align with what the Zotero team is thinking ...

    The Zotero team hasn't expressed much enthusiasm for dedicating programmer time to making this easy. This is I think in part b/c the Zotero 2.0 vision is essentially based on a centralized hub (though needn't be, I might add).

    For those research companies and institutes (as well as universities) that really consider this critical, can you imagine contributing to filling this gap?

    Of course, the code has to be available first.
  • I won't be of much help as I am an administrator and user but not a programmer. I also do not have a big budget to be of any help. Of course I could and would advocate this at my university if the code is available.

    I don't think that things must be easy. I appreciate that a server is a complex piece of software and that it takes some time to get things like that up and running. And with time people share their knowledge, help improving routines and write wikis.

    Yet I may share some thoughts. I did not realize that the vision of Zotero has shifted that far. As I first get to know Zotero I had high hopes (2-3 years ago). At that time a server was an essential part of the vision and I had not the impression that it should be closed source or centralized. I very well realized that it took far more time to develop the teamwork features but that is understandable and I am really looking forward to the new possibilities. But if I now hear that the only thing that would make me and a lot of colleagues switch to Zotero may will never come I feel as I did bet on the wrong horse all the time.
    BTW: A local server would be something that would make Zotero unique. Because if we are serious: There are a few research tools which are OS and also quite good. Yet a real client/server system is missing until now (as you stated quite correctly above).
  • On the vision thing, don't read too much into what I wrote. I'm just noting that the online sharing and recommendation functionality works best with more users. So unless they explicitly design for it, having many disparate servers limits the amount of data available for these features, and so also theoretically limits the usefulness to zotero.org users. That doesn't mean that local hosting isn't possible.
  • Is it possible to cut out the server portion all-together for group work? Using local metadata files? or linking between users local libraries?

    The reason i ask is that there are a load of other locally installable collaboration tools - wiki's for example - that could be used to augment/replace some of the zotero.org online functionality without affecting the zotero.org vision or other goals. It would seem to be something that could be solved with some extra documentation, and It could even serve to suggest new zotero.org features.

    That is assuming some or all of the metadata syncing is redundant with local access to the user libraries, or could be done via the plugin with read access to the metadata portion local users zotero libraries?
  • First, let me make clear that I appreciate Zotero for all its services.

    Although, if I may specify our clients' own concerns about sharing data in the cloud, I would say that:

    - storing in the cloud (which can be in another country than our client is in) disqualifies most client from coverage by their professional responsibility insurance;
    - raises concerns when you cannot certify that all data is transferred over HTTPS.

    Yet again, the cloud works well to motivate cooperation between organizational units.
  • I'm just going to repeat my earlier question (though also repeat: I don't speak for zotero):
    For those research companies and institutes (as well as universities) that really consider this critical, can you imagine contributing to filling this gap?
    By "contributing" I mean in concrete, tangible, ways (funding, programmer time, etc.).

    Has the server code even been released yet?
  • I am also interested in running my own local version of the zotero sync server within my network. I am currently using the sync server running on zotero.org and am getting frequent timeouts or sync errors.
    Couldn't you just release the source of the server like bdarcus said? This way we could continue using your great tool!
  • We're planning to release the code once 2.0 is final. That said, if you're getting timeouts and sync errors now, running the code locally is not going to help, since network issues and server load are not likely the problem. If you can help us to debug those problems, all other Zotero users will benefit from your assistance.
  • Hello Sean,

    This is great news indeed.
    However, for collaboration purposes, a ZeroConf (Bonjour) feature, which would allow to see other people's shared libraries over a local area network, would probably do the job. I think it would match most of the needs expressed here and would involve less hassle than maintaining a proper server locally.
    I understand that this feature was experimented some time ago for Mac OS X users, and there are still its embryos in Zotero's source. Does anyone know what the issues were, and what needs to be done to solve them? I might be able to contribute some effort but I need to understand the situation first.
Sign In or Register to comment.