(2) Follow the instructions for installing the “Citation” plugin in Obsidian.
(3) In the settings for the citation plugin, set the “citation database path” to your logseq “assets” folder, and the “literature notes folder” to your logseq “pages” folder.
(4) Configure the template as desired. Here is what I have:
This was the easiest workflow I found for now, but it has some disadvantages.
The first is that the template doesn’t have some entries I’d like, such as {{type}}. Ideally I’d like to have a Logseq property at the top level which specified the type (book, journal article, etc.) Also, you currently have to manually enter the date after “date added” as there is no template entry for that.
The second is that it is not a “live” connection to Zotero. You have to first export a bibtex file with all the citations you want to use, and you have to do this each time you add a new citation in Zotero. This is really terrible to be honest, but the payoff is that it makes it very easy to access and manipulate the data in Obsidian
The alternatives right now are mdnotes and zotero roam export. Both sort of work with Logseq, and mdnotes does have a some basic template system that lets you edit the output (zotero roam export does not). However, I couldn’t get it set up how I liked. Mdnotes is still in early development, so maybe it will get better.
How are you accessing the assets/pdfs within Logseq? Is it via the open in zotero link? So essentially you have in Logseq pages just metadata for the titles? Or I got that wrong?
How are you generating your bibliography citation pages? Will you do that in Zotero or Obsidian Citation Plugin?
I’m not in academia, but over the years I have amassed assets/pdfs related to public transportation totalling 5GB in about 1500 files. Early on I tried managing them in Mendeley but that turned sour when Elsevier took over them. So now I’m just managing those files/pdfs via keywords in their filenames and freeform metadata per file in descript.ion files.
I look forward to see how more experienced academics do it in Logseq for me to follow suit.
Yes, my goal is just to be import my notes and metadata and link that to my writing and note taking about other issues. I still use Zotero for everything else. At least until someone writes a more powerful plugin!
Actually my current database of 6000 PDFs (about 12GB) is in Paperpile, but they don’t yet have an API so I’m looking at Zotero for this.
So the Obsidian Citation plugin seems to work with any Bibtex file. I was happy to see that Paperpile has a new beta function to generate a perpetually updated Bibtex. The only downside is that the Citations plugin is optimized for zotero, and the Bibtex file from Paperpile doesn’t generate the unique links to items in Paperpile.
I can confirm that this works beautifully with Paperpile, even with a database of nearly 6000 citations. I have Paperpile set to auto-update the BibTex file to Google Drive, which is synced to my desktop. I then use Hazel to copy that file to my logseq assets folder whenever it detects that the file has been modified, replacing the original file. (The plugin could probably just be given the path to the original file in Google Drive, but this way it could also potentially work on a mobile device, even though I haven’t tested that yet.) Unfortunately the Paperpile bibtex doesn’t include all one’s notes and annotations, so one would have to manually add those to the note separately. That is fine for me, since I usually pull these in via Readwise anyway…
None of this is ideal, but at least there is some way to pull your metadata from either Zotero or Paperpile into Logseq without too much friction…
If you have the add-on Better BibTex for Zotero, you can export citations, and there is a box to check ‘keep updated’. If you check this box, it will automatically update the bibtex export file every time you add a new citation in Zotero. This is mentioned on the Better Bibtex preferences page, Preferences / Automatic Export.
