I understand Zotero is going to be dropped from LogSeq DB edition as a built-in feature; this makes me pretty sad, especially since there doesn’t appear to be a commitment to even continuing support through an official plugin . Is this because there is already a reliable community-supported plugin that will support DB? Otherwise, it’s difficult to understand why such a useful feature would be left out.
Even after finding the section addressing Zotero in the unofficial DB FAQ, I feel this change is significant enough to warrant its own dedicated discussion rather than being tucked away in an FAQ. The community’s response (or lack thereof) could also provide valuable insight.
Is Zotero integration not widely used by other users? I personally rely on it extensively—how about others? Even with fantastic community developers like benjypng stepping in, it’s a bit disheartening to see what was, in my opinion, a prominently advertised built-in feature like Zotero integration become deprioritized.
I agree that Zotero integration is pretty essential for scholarly work with LogSeq. But I also understand that the dev team is under pressure to simplify the codebase, given the scale of the rewrite of the codebase.
It looks like benjypng hasn’t replied yet to the question of will he continue to develop the local Zotero plugin for the DB version. But @Flan’s message above implies that maybe that is happening? (Edited my original post to add that according to the unofficial DB version FAQ, benjypng will indeed update the plugin. That’s great news!) Any other efforts at implementing Zotero support for the DB version?
The zotero integration as a first party citizen or something close to it was a killer feature why I was able to choose logseq over obsidian. Though everyday I have to use it I envy the obsidian world for their two way sync of notes.
there’s also this repository which seems to be an alternative to the one from benjypng: GitHub - e-zz/logseq-pdf-extract: Import Zotero into Logseq, extract text or area Highlights into block.
I haven’t tried either, but the latter from e-zz looks a bit more planning ahead.But I would love this or something similar to become a plugin with official support (including testing and fixing for every new version).
Hi,
I agree that it would be nice if the Logseq team continued providing the Zotero feature for the DB version. I use it all the time to organize my references, which is why I’m still using the MD version instead of the DB version. Technically, the Zotero feature shouldn’t be too difficult to implement. In the DB version, Logseq still allows PDF annotation so the core part is already there. The other step, importing literature metadata from Zotero desktop or cloud, should in principle always be possible (as long as the DB version supports plugin development).
Some problems may remain, such as the lack of a proper way for plugins to handle PDFs stored outside the Logseq graph folder. We need the {{zotero-linked-file }}or `{{zotero-imported-file}}` macros to access those PDFs, Also, Zotero integration isn’t the only thing missing from the DB version… Without enough plugin APIs, plugin developers can’t really help much. Overall, these wouldn’t be an issue if the Logseq dev team were directly involved. But so far, they’ve been focusing on the DB version itself, which takes time. Fingers crossed!
Now that db version is in beta, anyone knows if anyone has started developing a Zotero plugin? I really didn’t use the notes feature, but it was useful to use `TODO Read [[whatever]]` or `TODO Consider adding [[something]] to the SOTA`
One option could be the following: After the launch of Logseq Pro, some substantial piece of the contributions to the Logseq Open Collective (of which there have been $700,000+ with an estimated annual budget of $250,000+) could be used to officially support popular plugins like this with regular payments to the developers of those plugins.
I also am a regular user of the core Zotero functionality in Logseq, although it is frustratingly slow for me. (This is possibly because my Zotero library is excessively large.) I wonder if issues like this could actually be better addressed with a third-party developer focused just on Zotero functionality, instead of the current setup where the other priorities of Logseq tend to come before the Zotero functionality.
I would love if they could do something like zotflow (obsidian plugin): bidirectional sync to zot and even using the zotero viewer - meaning website snapshots, epub and drawing into pdf is supported as well.
Might even be possible to fork it. The developer said (when I asked him about adjustments for vscode) that most of the backend should be reusable
A huge amount of my scholarly work involves Zotero too. For what it’s worth, I’ve been very happy with logseq-citation-manger (it’s on github, I can’t link it). It uses Zotero’s BibTeX output (but you can configure Zotero to keep an .bib file up to date), and it has a rough edge or two, but setup was straightforward. If you want two-way sync and are using a Markdown database, the Zotero plugin MarkDB-Connect can bring your Logseq notes back into Zotero.
. Is this because there is already a reliable community-supported plugin that will support DB? Otherwise, it’s difficult to understand why such a useful feature would be left out.