Alright, let's add another entry to the annals of "seamless integration." I've been trying to get the Zotero sync working with Scholarcy for the better part of an afternoon, and the experience has been… well, let's just say it's given me flashbacks to trying to migrate a legacy on-prem database to the cloud with only "documentation" written in optimistic corporate futurism.
I'm using the Chrome extension, and the "Connect to Zotero" button seems to operate on a philosophy of hopeful thinking. Click it, it opens a new tab to a localhost address that either:
* Spins forever until it times out.
* Briefly flashes a Zotero pop-up that vanishes before I can blink, leaving Scholarcy in a perpetual "connecting" state.
* Does absolutely nothing, as if the button itself has given up.
My environment, for the record:
* Zotero Standalone (v7.0) running locally, with the built-in syncing turned off (I use a WebDAV server).
* Zotero Connector is installed and works fine for saving directly to Zotero.
* No firewall rules blocking localhost traffic (I've checked, and other local apps communicate fine).
The steps I've taken, in classic troubleshooting pilgrimage:
* Restarted both Zotero and Chrome. Repeatedly.
* Ensured Zotero is open before attempting the connection.
* Tried with Zotero syncing paused and enabled.
* Looked for any hidden setting in Zotero to "allow external connections" (found nothing).
This feels like a classic vendor handshake failure. Two pieces of software, each assuming the other will handle the protocol perfectly, leaving the user in the middle staring at a loading animation. Reminds me of the time we tried to auto-sync between two cloud data warehouses based on their "native integration" – spent a week debugging before finding out each vendor blamed the other's API latency.
Has anyone actually gotten this sync to work reliably? If so, what's the secret incantation?
* Is there a specific version pairing that's known stable?
* Does Zotero *need* to be logged into its own sync service for this to work?
* Are there console logs somewhere in either application that might spit out a useful error?
I'm not looking for a "it works on my machine" – I'm looking for the gritty, practical details of the successful deployment. Because right now, the workflow is about as smooth as a data migration that's just discovered referential integrity constraints.
been there
Test your rollback first