Upgrade StorageCollaboration over a local area network
Hello,
Zotero would be the perfect tool for managing bibliographic information if there was a possibility to share a database over a local area network. At the moment, my group can't adopt it as we cannot rely on the zotero.org server for many reasons, including:
- confidentiality: your privacy policy is fine but I guess you will not take any responsibility in case of information leaking from your server for any reason.
- continuity of the service: if zotero.org was temporarily or permanently unavailable to us for any reason (technical issue on your side, disruption of communications, lack of funding to maintain your server), this would have a major impact on our work.
In addition, I don't really understand why you insist on attracting an awful load on your server.
I also think that the demand for a collaboration tool has a limited scope (e.g. a research group, a department, an R&D project), which does not require a central server.
It would greatly help if the synchronisation/collaboration was more open. Here are a few ideas which could make the job (I don't know to which extend they are technically feasible):
- open/close the connection to the SQLite database each time an item is added or modified, instead of having the connection permanently open. This way, I guess a single file could be shared over a local network.
- release the server side of zotero's sync feature (a PERL/PHP/Python extension for a web server with an SQL database, perhaps?) so that users can implement it on their own server.
Of course any other solution would be welcome.
Many thanks.
Zotero would be the perfect tool for managing bibliographic information if there was a possibility to share a database over a local area network. At the moment, my group can't adopt it as we cannot rely on the zotero.org server for many reasons, including:
- confidentiality: your privacy policy is fine but I guess you will not take any responsibility in case of information leaking from your server for any reason.
- continuity of the service: if zotero.org was temporarily or permanently unavailable to us for any reason (technical issue on your side, disruption of communications, lack of funding to maintain your server), this would have a major impact on our work.
In addition, I don't really understand why you insist on attracting an awful load on your server.
I also think that the demand for a collaboration tool has a limited scope (e.g. a research group, a department, an R&D project), which does not require a central server.
It would greatly help if the synchronisation/collaboration was more open. Here are a few ideas which could make the job (I don't know to which extend they are technically feasible):
- open/close the connection to the SQLite database each time an item is added or modified, instead of having the connection permanently open. This way, I guess a single file could be shared over a local network.
- release the server side of zotero's sync feature (a PERL/PHP/Python extension for a web server with an SQL database, perhaps?) so that users can implement it on their own server.
Of course any other solution would be welcome.
Many thanks.
It seems to me that this should allow a single file to be used by several zotero instances.
You can share the Zotero data directory on a network drive (though it's unsupported), but you can only have it open on one machine at once. Asynchronous statement execution doesn't have anything to do with locking. It's just a performance feature. (It's unlikely we'll switch much over to asynchronous statement execution, because it requires a completely different coding pattern from what's used throughout Zotero. But 2.0b7 will bring some major improvements in UI responsiveness during long operations using a different method.)