Zotero is one of those pieces of open source software that has really improved my life. I first discovered it as a teenager and it revolutionised writing papers and managing sources. Sadly I don’t find myself using it that much in my job.
I always got a kick out of introducing it to my friends in the humanities and blowing their minds. I have no idea how anyone writes papers without zotero or another similar tool.
Keep up the good work.
What really blows my mind is how many of my academically inclined friends manage (and cloud sync) a big library of pirated papers using a rival piece of software owned by a major rights holder of said pirated papers. I mean... there's a good chance nothing will come of it, but it just seems like tempting fate. Zotero isn't just good enough, it's so good that the choice is completely trivial IMO!
I've converted all my colleagues to Zotero, even though the University has contracted subscriptions to EndNote et al. Zotero is just better. I wish that the UC System would contract with Zotero for their mutual benefit.
It really is the superior product/project. I switched to Zotero from Mendeley this year. The thing that was keeping me back before was the lack of internal PDF reader and annotator. I'm not sure when that came into Zotero, but now it is... /chefkiss
I had to use a trial version of Endnote 9 because a collaborator was using it. It was slow. Adding paper by DOI would freeze the interface for ten or more seconds on a database of less than 100 papers. Meanwhile adding a paper by DOI to my Zotero DB (curently 2000+ papers) is a snap.
I assume you're talking about Papers? The main feature is that there's an iPad/Android app that syncs really cleanly.
With Zotero, syncing reliably across multiple devices was always hit or miss for me. Saving a few papers throughout my day, going home and reading/annotating them on my iPad, and then going back to work and pulling up a paper from the previous day with its annotations intact is the dream workflow for consuming academic lit that has always just worked with Papers but always required a lot of effort with Zotero. Maybe this is more of a "me" problem than a Zotero problem, but if I can't configure Zotero to sync reliably then many academics won't be able to.
There are some other neat features too. The built-in reader is good enough to not bother going with a third party. The handling of in-text references is better than any alternative (the actual reference is listed when hovering or clicking the reference number in text–a seemingly basic feature that lots of citation managers and PDF readers don't have that is essential for maintaining concentration). The handling of figures, metrics, and smart suggestion of related articles is also top-notch. I was sad to make the conversion from FOSS but haven't looked back yet. Hopefully this new Zotero release will make it possible to transition back–it looks like lots of killer features in Papers are offered by Zotero 6, which is great to see.
As for the rights-holding thing, most journals give their authors the right to distribute their work for free upon request, and there are some other loopholes for getting papers afaik, so I don't think it would be worth the time or effort to go after people with a few sci-hubbed papers in their Papers account. I am afraid that DRM will make this much harder though, especially for papers that have been downloaded on a university network for later offline reading.
Does Elsevier really own Papers, Endnode, and Mendeley?
EDIT: no, it looks like Elsevier just owns Mendeley, which is what I was getting at. Uploading a bunch of papers to a company that's in a position to turn around and say "Those are some nice papers you uploaded to our servers, unfortunately they are evidence that you have fallen victim to piracy, now fork over $15k or we will be forced to proceed with a $15M lawsuit" just seems a bit reckless. Shrug.
I prefer Zotero, even though I would never even think of pirating a paper. Perish the thought!
To be fair Endnote license was given out for free and we didn't have to pirate articles as the university paid for all of these academic journal subscriptions.
That is exactly how I use it and it works great for that.
Among other things, I love that I can subscribe to RSS feeds and then save individual items I want to be able to reference later on.
Before that I tried several things, local bookmarks, bookmark services (actually just delicious), and an org file for references, but Zotero is the most seamless system for knowledge repository for my use case.
The web plugin actually falls back to 'web page with snapshot' if it can't detect a journal paper, which in many ways is better than trying to drag around big PDF binaries.
Thanks! I've used zotero only minimally, mostly to support users of my software who use zotero, rather than using it myself for my own purposes.
Is the "web plugin" something built into zotero, or an extra plugin to install? Calling it "plugin" makes it sound like something extra to install, but not finding it easily googling. Help me out?
I think OP talks about the Zotero Connector Browser Plugin. So its not a plugin for zotero but your browser (Firefox, Chrome, Safari and Edge are supported)
The plugin ads a button to the Browser where you can add the currently viewed page to your database.
It even tries to detect references to papers or other content and choose the matching entry type. Eg on Google Scholar you can choose which search result you want to add. It its a PDF the file gets saved to your zotero database.
I've been using this in beta for quite a while. The PDF reader is really very good, and I love the fact that my highlights and notes are not added to the actual PDF file. (I don't want to mess up the original PDF file, because I like to keep those files in pristine form so that e.g. I can share them with students.)
Some people here are asking about paying for storage for the PDFs. My solution to this is to store my PDFs using Dropbox, and to use links in Zotero. That way, even if I give up on Zotero, I still have the files. And I already pay for Dropbox, and just don't want one more bill to pay.
Zotero also offers out it the box WebDAV Support for file sync. I use that to sync files using my Nextcloud. Once it is set up, everything works behind the scenes.
If you sync your library to at least one device, you'll still have all the files anyway if you stop paying - you just won't be able to sync them anymore. The metadata will all still by synced.
The dropbox/onedrive/etc. approach works, but honestly I find it easier to share files from a group library (they don't need a paid account) and keeps structure etc. there also....
Zotero is also free for all intents and purposes. There's a paid tier with unlimited cloud storage for your pdf collection, but besides that everything is free.
Not just for scholarly reference management: Zotero is a great tool to keep track of the books you have if your personal library exceeds one or two Billies in volume. For books printed after 1970 the ISBN alone is enough to have it fetch the most relevant metadata for any book published anywhere from online catalogues. I find Zotero particularly useful when going to a book-fair; it is easy to export a list of titles and authors of books you own so you can easily reference it to see which books from a particular author or series you already have.
Well-designed, proven, and sturdy though. It's not made of MDF, it's particle board. Although a cheaper material, this is actually a good thing here, because the same bookcase made of MDF would be much heavier and the planks would sag.
Prices are certainly on the high side, but that phenomenon does not seem limited to Ikea.
The only thing holding me back from using Zotero is that, afaik, you need to use a Zotero account to reliably sync metadata between devices (the Zotero Data Server isn't supported). I wish it was easier to self-host; there's no reason for me to interact with their servers.
It's always a disappointment to find otherwise great open source projects that have key components that deny privacy.
Hint: You can sync the library for free by making an account. You need the "Zotero Storage" only if you want to sync the files as well. However, you can simply sync the folder ~/Zotero/storage using SyncThing. You could also use WebDAV, but SyncThing is just so much easier (set up & forget). The IDs/Metadata is consistent among syncs. :)
Of course, that doesn't resolve the privacy concerns, but personally idc about that.
edit: It seems like you might be able to sync everything by just putting ~/Zotero/ into something like SyncThing, avoiding the Zotero servers. But you'd probably have to be careful only opening the app after syncing has been completed. I haven't tested this though.
This is completely missing the point of GP. You can't (unless something has changed) sync metadata without using their servers. webdav is for the data, not the metadata. If you don't personally care about privacy, fine, but that's irrelevant here.
Replicating the local Zotero folder across devices won't properly sync the metadata in the face of concurrent changes.
If I didn't care about privacy and freedom I would have just used a proprietary product to begin with. It's an incredibly disappointing situation to have dragged on like this for so many years.
"Just fork it" is dismissive and misses the point. The criticism remains valid even if the maintainers refuse to act on it for whatever reason. Even if I were to invest the time to fork the project, the criticism would _still_ remain valid as long as upstream didn't take steps to fix the problem.
I‘m on windows, and here it is still possible. Try checking your settings again, it is a bit hidden: goto settings/sync, then file-sync -> via WebDAV is available in a Dropdown menu.
Zotero supports WebDAV and I was using it with box.com without any troubles for years. What lured me to switch to Zotero's sync and cloud storage in the end, though, is the group library function.
Yes, I wish they supported the server instance. Still, I don't mind subscribing to their syncing because it is probably the project's main source of revenue; although I am not for certain about that.
As someone who writes web apps used by academics... I find it really hard to find documentationon on how to make my web site/app work well with zotero! They seem so used to more-or-less adversarial web development, that when we're like "We want to make this work as well as possible with zotero, like I'm going to spend development time doing that, can you give me some clear docs on my options?" -- I have trouble finding it. Like how can I make sure zotero can properly find an attached PDF, or citation info (including for list pages with multiple things cited) or whatever. Without requiring a custom zotero connector for my (tiny) site, how do I make my site just work? Many of my users definitely use zotero.
thank you! I maybe have seen this doc page before... I'm afraid I don't find it totally clear!
Do you have any advice for the simplest way to do this from these docs, if starting from scratch? Or what most people do? If you were writing a new website yourself, just an ordinary, say, Rails app, that was not based on library formats like MARC... which method would you use to expose to zotero?
COinS is pretty limited and crufty, and doens't allow you fo fully express all the fields that might be in a Zotero citation/CSL, only a few basic ones. UnAPI... most of the formats listed for UnAPI, the list tells you not to use that format unless you have to! It seems to recommend MODS, but MODS is pretty confusing if you don't already have data in it.
Dublin Core is not expressive enough to cover everything in a citation.
I just want to specify every in a zotero citation, you know? Is there a way to provide data in CSL? I don't think so?
`<meta>` tags "using the Google/Highwire key-value system" sounds prommising -- but I'm having trouble finding a list of valid keys at the documentation linked to! Those Google docs feel like they're sending me on another goose chase.
I feel like I have a choice of over-simplified (DC) or baroque/archaic/hard-to-use metadata systems (MARC/MODS), and in either case having to kind of guess how it will be mapped to Zotero citation/CSL.
I wish there was a way I could provide the cite in CSL!
Also... I know zotero sometimes can handle a list of citations on a page, but the page you link to has no guidance for that.
If other website devs have no problem providing good metadata to zotero using these docs... I feel like there's some kind of secret code here everyone else knows that I haven't been able to crack!
> `<meta>` tags "using the Google/Highwire key-value system" sounds prommising -- but I'm having trouble finding a list of valid keys at the documentation linked to!
They're in the Indexing section of the linked Google page, just formatted in a particularly unreadable manner:
> I wish there was a way I could provide the cite in CSL!
We're planning to support `<link>` elements to any supported format — basically a modern replacement for unAPI, which went defunct before CSL could be listed as an option — but CSL wouldn't let you specify a PDF anyway. And most sites want the search engine indexing they get from embedded metadata.
Thanks, your feedback is super helpful! Answering some things I haven't managed to figure out confidently on my own beofre now. Sounds like Google/highwire might be the simplest at this point. Still curious if you have any other recommendations!
Actually, one thing I'm still unclear on how is to supply a book/monograph citation, since google/highwire isn't focused on that (not sure if it's possible with google/highwire). Or do I have to resort to a different format for that? (One like "coins" that might not let me supply all relevant fields? :( )
> We're planning to support `<link>` elements to any supported format — basically a modern replacement for unAPI, which went defunct before CSL could be listed as an option — but CSL wouldn't let you specify a PDF anyway.
That sounds great! The key is still what formats are available -- somehow there aren't any great ones. I wonder if it would make sense for zotero to actually develop one that lets you simply describe what zotero means, and everything.
But sounds like maybe most people wouldn't want to use that anyway, as you say, they want something that will be used by Google and others. (The lack of a suitable standard used by multiple parties, to me reminds me of my thoughts on living in an era of death of internet standardization).
Maybe embedded google/highwire is the best option -- I wonder if zotero docs page could beneficially highlight this, suggest that if you're coming to this from a clean slate and just want the simplest/easiest thing that will work, the google/highwire meta tags are probably a good one. Perhaps with guidance on finding the actual keys in the Google docs, and even linking to the zotero mappings as you've done here!
And the other part that's still not clear to me is what to do on a `list` page -- I understand zotero sometimes can handle being activated on a page listing several articles/citations, and export them all. But I am not sure how to handle that.
(I suspect there are still other things I'd have to spend time reverse engineering, that aren't clear from the pretty sketchy google docs, once I get to the details.. but only one way to find out, I guess! When I've invested time in this before, I've ended up blocked and unable to figure things out. Is there a zotero discussion forum where I'd have a chance of getting questions like this answered from experts?)
Wow. I opened the Linux tarball and had a look around... what a blast from the past. It looks like this is built on XULRunner and friends. I had the impression that platform had slipped into such disrepair that it was no longer practical to ship products based on it.
It looks tremendously useful, though. If it had been around when I was in school, I suspect I'd have found it indispensable.
> had the impression that [XULRunner] had slipped into such disrepair that it was no longer practical to ship products based on it
It's not! Zotero has a novel build system where instead of using XULRunner, which is no longer available, the build scripts download a release version of Firefox and then replace the browser-specific modules and other assets with Zotero's bundle.
And as bad as it sounds, it's not worse than most other build systems[3][4] (including Firefox's own build system[5], or anything based on Electron[6]).
…holy cow. I also thought you were joking. When I read the headline and that Z6 is the "biggest upgrade in Zotero's history," I was absolutely looking forward to it ditching XULRunner and friends.
This isn't a very nice comment. It reads like the equivalent of someone trying to use high school bully tactics to make someone else feel bad by convincing them that if they don't stop being weird then no one will like them. No actual criticism—just straight up "look at the weirdo"-tier sniping.
Looks like you failed to absorb the point of the previous comment: No actual criticism—just straight up "look at the weirdo". Instead of addressing that, you just embraced it:
> I mean, that is weird
So? The purpose of my comment wasn't to dispute that it was weird. Want to stipulate that it's weird? Fine, granted. Pointing out that something is weird, however, is not a sufficient substitute for substance.
Trying to appeal to the argument that everyone _knows_ you're a weirdo—as, once again, a substitute for articulating any sort of actual criticism—is for losers.
You don't want everyone to think you're a loser, do you?
I just tested the PDF reader out for about 5 minutes and I can already say it's substantially better for me on MacOS than any other PDF reader besides Preview. No glitching like Acrobat and the annotations features are more inline with what I want than Preview's are.
i agree it’s great. some minor ux issues. for example if i select text then click highlight it should highlight the selected text. keyboard shortcuts for annotation would be amazing.
I cannot tell from the changelog for this or the previous version whether Zotero has expanded its metadata and formatting capabilities to better support legal citations, so I guess I will stick to the Jurism[0] fork for now. If anybody knows of a better way to make this work in Zotero directly, I'm all ears.
It'd be so fantastic if the Zotero team could consider integrating their changes. I know that many probably won't necessarily need its advanced futures, but I see no risk of confusion as it's easy to use and mostly hidden in the "case" category.
I used Zotero quite a bit in the past, the thing that ultimately drove me away is that when synchronizing you couldn't easily access the PDFs just via the filesystem. I use a nextcloud and also wanted to access PDFs from devices without Zotero installed and this turned out to be a pain at the time. Is this possible with more recent Zotero versions? Is the sync still cumbersome with a zotero account + self-hosted webdav?
I have a similar setup (Zotero + dropbox). Zotero syncs the entries and any notes, but dropbox takes care of the pdf storage. You can swap out dropbox for your storage of choice.
Files are all stored in a single folder (which can be in different locations depending on computer) which I point zotfile to. Inside that each paper is put into a folder based on author names. And each pdf is then renamed based on authors, year, and title.
I'm glad it includes some, but perhaps not the feature of storing pdfs in a different place so they can be easily synced via Dropbox? While I'd like to have as few extensions as possible, maybe I'll neat to keep Zotfile around regardless.
Zotfile is an extension that automatically converts internally stored pdfs to links in a filesystem directory -- the latter can be synchronized and accessed however you please.
It's been working great for me, although recent versions have a problem that it no longer recognizes the root directory and pollutes the zotero database with machine-specific absolute paths ((
After messing with the desktop version for a few minutes the note editor is weird. It doesn't modify the original pdf or the copy of the pdf it makes, the only way to get the highlights is to export it as a pdf. I expected something similar to Okular where edits would happen to the original file and update on ctrl-s.
Truly amazing! I wonder if the iOS app in development will have all these note-taking features and work on iPad. If so, Zotero completely blows all of the competition out of the water at a price you can't really beat.
I love zotero, and this isn’t the web product, but I really wish I could share a link to a bibliography to someone. I have to export and send the text. I’d like to be able to share my Foo citation library with a colleague without making my whole account public.
I recently found my mendeley app doesn't work anymore (they've discontinued it) so I decided to try Zotero. To my surprise they have 3 Android apps.. none of which work and a bookmarklet which also doesn't work. Has anyone found a way to easily save papers on Android?
The annotation features will definitely help Zotero to keep relevant, but Zotero still desperately needs a better option for saving websites. The ability to save sites, via browser extension, as a single file (MHTML or even HTML-only) has been an open request since 2007.
If I still was doing firmware dev, I could imagine this being amazing for dealing with datasheets, especially when shuffling between multiple parts all with their own descriptions of intended functionality, errata, and so-on.
I’ve been using the better of this for, gosh, must be about two years now. It has been so much fun to use. I read papers and annotate them on my iPad, and it syncs up to my laptop without any problems. I recently considered turning on WebDAV to sync my growing collection of documents, but Zotero has been so useful to my workflow, that I decided to buy some storage from them. I am more than happy to give them a few bucks to keep working on this awesome product. It has significantly simplified my research workflow, and it is so valuable to me for that.
Whoops, yeah, mean to say "I've been using Zotero for the better part of two years now." A better Zotero there is not. :) Sheesh… must have been late or something—don't know why I typed that.
I believe the author intended to say that they've been using this tool "for the better part of 2 years", meaning roughly 1-2 years, not that they were using something better than it.
Anybody have any tips on self-hosted Zotero? Org doesn't like the metadata syncing. Other option would be Mendeley Enterprise which is supposedly more secure but not a great experience overall.
I haven’t used their new iOS app, but it’s a welcome addition. In the past you had to sync your Zotero library with another app for PDF annotation that was synced (papership). It was never something I’d call robust or seamless and the papership app has felt like abandonware.
This looks really nice. Zotero has never stuck with me for some reason but I last used it a few years ago. I'll try again.
One problem with this type of service is that there's usually a huge jump from unpaid to paid, but the jump to the lowest paid tier of Zotero lets you pay $20/year for 2 GB of storage. I can handle that. I have more difficulty with the $15/month and $90/year and such paid tiers. Even at $20/year, that comes to $100 after five years for what is largely a donation to support development.
One probably non-obvious thing is that the $100/yr ish unlimited storage tier includes everything you do as a user including shared libraries.
So if you have a group of people working together in a group library, one shared library subscription covers all of them. It doesn't take many people to make this worthwhile.
And as you say, it's a good way to support a great project. I do it even when I'm not really using the software.
> I do it even when I'm not really using the software.
I'll drop a one-time donation of $10 or $18 or something largely independent of my use - or even if I don't use it at all. I don't treat it as a purchase, though I can understand that others do.
I'm with Zotero for about 2 weeks and I'm very much in love with it. I use it mainly to store and categorize downloaded Websites (SingelFile) and PDFs.
While a PDF viewer (and every functionality that comes with it) is greatly appreciated, I'd wish for same for html and other formats. That'd be overkill :)
One thing that I really miss and actual slows my process, is a proper tag-system. Something like Anki has: being able to make tag hierarchies by higher-tag::lower-tag. I'm a heavy tag-user and the overall experience with tags in Zotero is very average. But maybe I'm using this application for stuff that's not intended for.
Overall great project and surely something I'll donate a few bucks to the next time I'm on that spree. You guys doing gods work. Love it!
I suggest to store them on de-duplicated filesystem or manually run fdupes and co to hardlink any duplicate in your Zotero lib because most modern websites are full of big js and zotero copy them countless of time, if you have a non marginal set of mirrored pages you probably waste more than 60% of occupied storage just in duplicates...
I have to check my folders for that. Since I am using SingleFile for downloading websites, I get single HTML files (as the name of the program says) so there shouldn't be any JS included...
As good as zotero is, it reminds me of the horrors of grad school and how countless livelihoods of really smart and intelligent are destroyed every year in the name of science and research, who don't have any employable skills to find a decent job.
The notes functionality along with the new iOS beta is amazing. Makes marking up papers on an iPad very convenient. I would encourage every who uses Zotero to sign up for a paid storage plan to help support development.
Is anyone else getting horrible stuttering when scrolling that page? (I happen to be on mobile with a bad connection right now, but it's so bad it makes the page almost unreadable)
I see there’s an iOS app in beta. Will that require using Zotero Storage, or will there be integration with Files, or directly with other storage providers (Dropbox etc.)?
I hoped they've fixed Google Docs stability (deactivating GDocs' Work Offline is supposed to help with this, we'll see..) as well as reducing the time for a bibliography update when the number of citations is over 100 - it's currently taking 20+ seconds (happening in the background as I write)
And even still encountering the dreaded "these citations have become unlinked from Zotero" - this is the equivalent of the blue-screen-of-death for OS's . . . maybe Zotero should focus on improving this stability rather than new feature build-out and I would be come a paying customer/member?
I tried both zotero and mendelay as a grad student. They both had pretty poor bibtex support, which made them more annoying than worth while. I ended up much happier with helm-bibtex for emacs. If you're already writing papers in latex with emacs (which is _really_ nice), it's well worth a look.
Looks good, from whatever I have seen and read. I am going to give this a try.
A question to those who have used Zotero:
Can I opt to store everything on the local device and have zero data transferred to the server? Is there any kind of control over what is stored locally and what is sent to the servers?
By default, you can use Zotero without logging in, which will cause it to keep everything locally. It's not obvious how you might go about partially syncing to the cloud, however.
In practice, though you shouldn't store the SQLite in a cloud sync service, setting up a symbolic link for the backing PDFs in e.g. Dropbox works quite well and gives you more control over your data syncing. Then, you only need to rely on their cloud for the paper titles and metadata.
My recollection is that it is all local by default. If you don't specifically connect to their web service for syncing data, none of the data is stored outside your local device.
Does Zotero have an article suggestion feature (or discoverability) yet? It is the main reason I am still using Mendeley, I receive suggestions on articles related to the ones in my library. Zotero would make my work easier and faster, but right now Mendeley is making it better.
I've been using this for years. Zotero is great. It's not perfect, like you still need add-ons to properly support bibtex for latex users, like myself. It would be alot better if this was a fully integrated part of the base zotero platform. But it works and I like it.
I’m still using qiqqa for my open source pdf search engine but the qiqqa team has moved on. Qiqqa has been a secret weapon for me in pulling up relevant paragraphs from research papers, ebooks, and standards that I have previously saved but not fully read.
How good is Zotero as a bookmark manager similar to Pinboard.in or Raindrop.io? Both have a paid feature that stores a permanent copy of a site, but these are stored only on their servers, and Pinboard's archive download feature never worked for me.
Zotero uses SingleFile. (Side note, this is, as far as I know, a destructive process—you can't recover the original bits. It would be interesting if Zotero switched to SingleFileZ and there were a sidecar file embedded in the ZIP container that would let you reverse the transformations to get the originals back out—or just store them directly.)
Zotero's snapshots work well enough on most pages you'd care to throw at it in an academic context. Your mileage may vary if you're using it as a general bookmarks manager, owing to the amount of client-side scripting-driven chicanery on the Web today. Notable example that I've found that Zotero didn't handle well at one point: Medium (although the data was there―the saved copy just wouldn't render correctly in-browser when opened; this doesn't appear to be an issue anymore for new snapshots, at least for now).
For most stuff that I want to have a durable URL for but that I feel doesn't merit being in my Zotero library, I just use ordinary browser bookmarks and make sure that the Wayback Machine and/or archive.is has a copy.
I've used Zotero for over a decade to save webpages. It works fine in my experience, though I'd prefer if Zotero saved all the assets in a single file.
I’ve been using zotero since 2010 … it’s amazing to see it’s still developing. Dying to try the ios app but I never received an invite,but when I request one it says it was sent. If anyone would like to sell me one let me know.
I love zotero. Amazing software, indispensable. Want an easy way to integrate with pandoc and Zim-Wiki, I'm sure there's a way but I've not found it yet.
You can use the BetterBibtex plugin to export the references to a bib or json file which can be ingested by Pandoc. There's an option to have it re-export the file every time you make a change to the library.
Citations work pretty well. You can customize the output with different CSL files [1].
That is a great new release! I have been using Mendeley because it is easier to take notes while reading a paper, but note-taking lacks rich functionality.
The cherry in the pie would be a Zotero version for the Remarkable2 :)
I'm probably stuck using Adobe for now, just because for anything really important you can't rely on anything aside from the official Adobe PDF editor to not screw things up. Up. But I absolutely love your project, I can imagine this would work great for academics.
You can automate much of your research between machine learning and directly reading PDFs. Many older publications don't have searchable text at all, and searchable text is a lifesaver when you're doing research.
I always got a kick out of introducing it to my friends in the humanities and blowing their minds. I have no idea how anyone writes papers without zotero or another similar tool. Keep up the good work.