+1 I, too, need this feature! It should work just like a Unix symlink.
Just one example of why I need this (the one that prompted me to finally come search for it here):
I have collections for each class I teach. But often a topic is relevant to more than one class. For example, I was just adding an item to the "Price Controls" subcollection under my "Government and Business" class, but this subcollection also belongs under my "Principles of Microeconomics" class. As it stands, I either have to remember that this topic lives under the first class, or I have to create separate subcollections under each class and then copy every item over each time I add one.
I would be repeating what everyone else has already said about the need to be able "hyperlink" from one sub-collection to another. I have been struggling with the lack of this functionality for quite some time. I know resources are limited, but action to make this happen will be greatly appreciated. Thanks
I'd be curious to know what makes @dstillman think this feature is especially complicated to implement/conceptualize. It (naively) seems to be the exact equivalent of a UNIX symlink and thus could simply have a link icon next to its name or something. This feature would indeed be extremely useful and is complementary to tags IMO.
I would like to add some notes to this thread (I stumbled over this only as well).
zotero:// url in the url field of an item
It was useful to find out that placing a zotero:// url in the url field of an item, properly opens a (sub) collection or item in another collection. I never tried that KungFu before. The only issue is that this (as far as I know or experienced) only works with subcollections, but not with the library / groups root item in the app. Of course it works in the web version.
Since Zotero 6 copying an item from one library/group to another left cross references betwen this intertwingled items in the database. I found this out hwen dealing with relations in BetterBibTex custom export JSON Data and finding relations not visible in the Zotero6 UI.
It turned out that in Zotero 7 these origin-relations were exposed and showed other Libraries containing a copy or ancestor.
These copies are not clones or duplicates.
They are copy on write items as far as I guess. It takes almost no time to copy them to another library because the local storage is copied as a lightweight reference first and only duplicated in the background with delay. Then if the duplicate is modified the duplicate clone stops sharing the data and be modified independently.
Relations of multiple items at the origin copied at once are lost at the destination and need to be recreated.
Further readings
It is useful to understand the different ways how symlinks, hardlinks, references deduplication etc. are managed in filesystems on different operating systems / fliesystems and sync services.
I suggest the article by Howard Oakley [2] on this in the context of macOS and the corresponding objects under Windows (Junctions etc) and how object references are managed in Python.
Interesting is the strategy how Apple Aliases are resolved during lifecycle changes. Oakley has cretaed a tool to find broken aliases and a helper to fix them (Alifix).
Mojave has 5 different types of copy/clone/alias/link, some of which can appear to behave oddly. Here they are, fully explained.
In my Resarch documentation Tool Logseq I make intense use of aliases for tags to join thesaurus like references with different spellings or languages. This is also missing in Zotero. I have created scripts to manipulate tag listst in the clipboard before pasting into different tools, joining spellings and similar stuff. Zotero ´can help renaming tags and join this way, but loses the original misspelled tag.
My usual research workflow goes this way:
Collecting links, snippets and tags by pasting into a Telegram bot on the road.
Catching them in Logseq with a plugin that collects the chat snippets in a journal.
Open the links in the source app (PDF, Websites) and save Zotero Snapshots. Enricht hem with tags, notes an relations. make use of the excellent extraction of Metadata in Zotero.
Then export selected items into clipboard as Markdown for logseq using BetterBibTex Custom eta-template to render the Markdown incliding zotero://.../select urls or zotero://.../open-pdf links
Pasting the enriched Markdown back to Logseq with Citation, urls, Tags etc. and even Notes.
Finding and further note taking takes place in Logseq with later publication preparation. All self created stuff is in Logseq, all the external sources and annotations, notes and clippings are in Zotero with their source snapshots and ability to revisit the origins.
Upgrade Storage
Just one example of why I need this (the one that prompted me to finally come search for it here):
I have collections for each class I teach. But often a topic is relevant to more than one class. For example, I was just adding an item to the "Price Controls" subcollection under my "Government and Business" class, but this subcollection also belongs under my "Principles of Microeconomics" class. As it stands, I either have to remember that this topic lives under the first class, or I have to create separate subcollections under each class and then copy every item over each time I add one.
Thanks
It (naively) seems to be the exact equivalent of a UNIX symlink and thus could simply have a link icon next to its name or something.
This feature would indeed be extremely useful and is complementary to tags IMO.
zotero:// url in the url field of an item
It was useful to find out that placing a zotero:// url in the url field of an item, properly opens a (sub) collection or item in another collection. I never tried that KungFu before. The only issue is that this (as far as I know or experienced) only works with subcollections, but not with the library / groups root item in the app. Of course it works in the web version.
Since Zotero 6 copying an item from one library/group to another left cross references betwen this intertwingled items in the database. I found this out hwen dealing with relations in BetterBibTex custom export JSON Data and finding relations not visible in the Zotero6 UI.
It turned out that in Zotero 7 these origin-relations were exposed and showed other Libraries containing a copy or ancestor.
These copies are not clones or duplicates.
They are copy on write items as far as I guess. It takes almost no time to copy them to another library because the local storage is copied as a lightweight reference first and only duplicated in the background with delay. Then if the duplicate is modified the duplicate clone stops sharing the data and be modified independently.
Relations of multiple items at the origin copied at once are lost at the destination and need to be recreated.
Further readings
It is useful to understand the different ways how symlinks, hardlinks, references deduplication etc. are managed in filesystems on different operating systems / fliesystems and sync services.
I suggest the article by Howard Oakley [2] on this in the context of macOS and the corresponding objects under Windows (Junctions etc) and how object references are managed in Python.
Interesting is the strategy how Apple Aliases are resolved during lifecycle changes. Oakley has cretaed a tool to find broken aliases and a helper to fix them (Alifix).
[2] Aliases, hard links, symlinks, and copies in Mojave’s APFS
creators:: author: hoakley
url:: https://eclecticlight.co/2019/01/05/aliases-hard-links-symlinks-and-copies-in-mojaves-apfs/
#### abstractNote
Mojave has 5 different types of copy/clone/alias/link, some of which can appear to behave oddly. Here they are, fully explained.
In my Resarch documentation Tool Logseq I make intense use of aliases for tags to join thesaurus like references with different spellings or languages. This is also missing in Zotero. I have created scripts to manipulate tag listst in the clipboard before pasting into different tools, joining spellings and similar stuff. Zotero ´can help renaming tags and join this way, but loses the original misspelled tag.
My usual research workflow goes this way:
Collecting links, snippets and tags by pasting into a Telegram bot on the road.
Catching them in Logseq with a plugin that collects the chat snippets in a journal.
Open the links in the source app (PDF, Websites) and save Zotero Snapshots. Enricht hem with tags, notes an relations. make use of the excellent extraction of Metadata in Zotero.
Then export selected items into clipboard as Markdown for logseq using BetterBibTex Custom eta-template to render the Markdown incliding zotero://.../select urls or zotero://.../open-pdf links
Pasting the enriched Markdown back to Logseq with Citation, urls, Tags etc. and even Notes.
Finding and further note taking takes place in Logseq with later publication preparation. All self created stuff is in Logseq, all the external sources and annotations, notes and clippings are in Zotero with their source snapshots and ability to revisit the origins.