"An error occurred while saving this link"

dlcamp07 · January 29, 2017

I'm having a problem saving to Zotero in Firefox. When I try to save a PDF obtained from a database, I click the "download" option from then menu and then select "save to Zotero". Then the "an error occurred while saving this link" appears. I can add items using the DOI, and I can drag them in and save them. I have this problem with many PDF files. Zotero is the only extension in Firefox that I have. My debug ID is D1052532489

adamsmith · January 29, 2017

are you saying this happens for every PDF or just for a specific database? If the latter, do you have the URL (exactly as you see it) as well as the DOI?

dlcamp07 · January 29, 2017

It seems to be happening no matter what database I use for access, but I am able to save pdfs from web pages. For example the first two below came from the Proquest Research Library and the ERIC database. Neither saved. The latter came from a web search, and that was added to the library. The final address is the original article I was trying to save that was from the South African Journal of Education. It also won't save, and I can't find a DOI for it.

http://media.proquest.com.proxy1.cl.msu.edu.proxy2.cl.msu.edu/media/pq/classic/doc/1257716741/fmt/pi/rep/NONE?cit:auth=Garcia,+Penny;Rose,+Stephen&cit:title=The+Influence+of+Technocentric+Collaboration+on+Preservice+Teachers'+...&cit:pub=Journal+of+Technology+and+Teacher+Education&cit:vol=15&cit:iss=2&cit:pg=247&cit:date=2007&ic=true&cit:prod=ERIC&_a=[...]&_s=[...]

http://media.proquest.com.proxy1.cl.msu.edu.proxy2.cl.msu.edu/media/pq/classic/doc/3055205821/fmt/pi/rep/NONE?cit:auth=Thomas,+Tommye;Herring,+Mary;Redmond,+Pamela;Smaldino,+Sharon&cit:title=Leading+Change+and+Innovation+in+Teacher+Preparation:+A+Blueprint+for+...&cit:pub=TechTrends&cit:vol=57&cit:iss=5&cit:pg=55&cit:date=Sep+2013&ic=true&cit:prod=Research+Library&_a=[...]&_s=[...]

https://www.myskillstutor.com/guides/classroomguide/mffWorksheets.pdf

http://journals.co.za.proxy1.cl.msu.edu/docserver/fulltext/high/22/1/high_v22_n5_a10.pdf?expires=1485656252&id=id&accname=58209&checksum=[...]

adamsmith · January 29, 2017

So, first thing is that I'd recommend, when you're looking at databases like Proquest or Eric, to tweak your workflow: don't try to download the PDF and run retrieve metadata, but just click the Save to Zotero icon when you're looking at the abstract page/landing page. You'll get better data and more reliable import. In many cases (depending on your setup) you'll also get the PDF auto-attached.

That said, what you're doing should work. What happens if, instead of the Download button followed by Save to Zotero, you just click the Save to Zotero icon (i.e. the -- probably grey on those pages -- page icon next to the Z). Does that work?

@Dan%20Stillman -- can you see something in the debug posted above, or do we need something different?

dstillman · January 29, 2017

(@adamsmith: Just dstillman now. Wasn't working with the space.)

(3)(+0000006): Importing attachment from URL http://journals.co.za.proxy1.cl.msu.edu/docserver/fulltext/high/22/1/high_v22_n5_a10.pdf?expires=1485656252&id=id&accname=58209&checksum=[...]

[...]

(2)(+0000939): Downloaded PDF did not have MIME type 'application/pdf' in Attachments.importFromURL()

adamsmith · January 29, 2017

So what I think is happening is that your institution is trying to re-authenticate you when you're trying to save the links -- that brings up the login screen, which of course is no PDF but a simple website, so Zotero doesn't save the file.

The two most likely scenarios I can think of are:
1) You had these pages/PDFs open for a long time before trying to save them. Sessions with your institution time out.
2) Proxies use cookies for authentication. If you have very restrictive cookie settings or browse in an incognito profile, this may break things.

Could either of these be the case?

dlcamp07 · January 29, 2017

I've checked cookie settings, and that shouldn't be an issue. I'm going to experiment more and see if things are working. Thanks for the help!

dlcamp07 · January 29, 2017

I just tried saving something new in a freshly-opened tab. When I use the DOI to add, I get the information but not the PDF. When I click the Zotero button to add as a webpage, the PDF comes in but the metadata doesn't come with it, and when I try to open that PDF, it doesn't open. When I try to add the webpage as suggested,("but just click the Save to Zotero icon when you're looking at the abstract page/landing page. You'll get better data and more reliable import. In many cases (depending on your setup) you'll also get the PDF auto-attached."), it doesn't save into Zotero.

dlcamp07 · January 29, 2017

To be more specific, if I log into the Taylor & Francis database via my university library, and then click the Save to Zotero icon, nothing goes into my library. If I am not logged in via my library, I get the webpage image to save, but the PDF does not attach since I don't have access to it because I'm not logged into the site.

Logged In URL: http://www.tandfonline.com.proxy1.cl.msu.edu/doi/abs/10.1080/14759390802704097

Not logged in URL: http://tandfonline.com/doi/full/10.1080/14759390802704097

adamsmith · January 29, 2017

but on http://tandfonline.com/doi/full/10.1080/14759390802704097 you get all the other fields imported?

You could try using Zotero Standalone with a different browser? As I say, this is almost certainly due to something in MSU's proxy settings (you could check if the library knows more about this) and/or your browser. Zotero's requests are hitting the library's authentication wall, even though you're authenticated.