Web Snapshots Not Opening

I recently installed Zotero for a curator and changed the location of her Zotero Data Directory from the hard drive to her personal drive on our network, following these directions - https://forums.zotero.org/discussion/70081/moving-zotero-data-directory#latest. I change the directory all the time with no problem on staff PCs. I do this BEFORE adding any items to a person's library.

After doing so, I started to show the curator how to add items to her Zotero library, such as JSTOR citations and PDFs. Any attached PDF file that I added to her Zotero client from JSTOR and other full-text databases shows up without a problem in her library. When I double-click on the PDF file icon (embedded underneath the citation in the library's center column), the file opens.

However, here's what happens when I add a webpage with snapshot:

-first, the webpage capture icon does not seem to be working properly as it only captures the citation, not the snapshot.

-Instead, I use the right-click menu and select either “Save to Zotero (Web page with snapshot)” or “Save to Zotero (Embedded Metadata).

-The icon for the snapshot then appears embedded underneath the citation in the library's center column. But when I double-click the icon to open the snapshot, I get the message that the file cannot be found because it was deleted or its location may have moved.

-This is usually the message one gets when they move they move their Zotero Data Directory without also moving files they have previously added to Zotero to the new location. But this is not the case here since I moved the Data Directory right after installing Zotero and before any files at all were added. Also, why are the PDFs opening and not the snapshots?

It would be a shame if the curator could not get access to her archived webpages, so any help is appreciated. Thanks.
  • I neglected to add that this is the link provided in the error message when the curator tries to open her saved snapshots - https://www.zotero.org/support/kb/files_not_syncing

    To clarify:
    -She only has Zotero installed on one computer
    -We have synched her data, but not her files, because of the 300mg limit on the cloud and because she only intends to access Zotero on this one computer.
  • Could we see a Debug ID from Zotero (not the Connector) for a save attempt that produces a broken snapshot?
    first, the webpage capture icon does not seem to be working properly as it only captures the citation, not the snapshot
    That would be separate, but for pages where this happens, what does it say when you hover over the save icon, and "Automatically take snapshots when creating items from webpages" checked in the General pane of the Zotero preferences?
  • My apologies, I won't have access to the curator's PC until after the holidays - in early January - but I will check then and provide you with the Debug ID and so forth.

    Thanks.
  • Hello, I now have access to the curator's PC and am submitting the Debug ID requested above for the save attempt - i.e. when I click on the Zotero capture icon in Chrome to save a webpage to the curator's Zotero client. In this case, I get a citation only - no snapshot appears.

    D1048260507

    If I use the method of right-clicking on the webpage, I get these options - note, the first option was not appearing until today:

    -Save to Zotero (NYTimes)
    -Save to Zotero (Embedded metadata)
    -Save to Zotero (Webpage with snapshot)
    -Save to Zotero (Webpage without snapshot)

    The only one of the above options that produces a snapshot in her Zotero client is the "Save to Zotero (Webpage with snapshot). However - and this is the problem I describe in my initial post - when I click on the snapshot icon embedded underneath the citation, it produces the error message "The attached file could not be found - it may have been moved or deleted outside of Zotero .... and so on" with a link to https://www.zotero.org/support/kb/files_not_syncing. (I have followed/checked the instructions at this link - see above thread.)

    Why is this happening? As I mention above, I changed her Zotero data directory BEFORE adding any files. While the PDFs open with no problem, the snapshots are not opening.

    I have also checked her Preferences and "automatically take snapshots when creating items from webpages" is checked.

    So there seem to be two problems here:

    1) the capture icon is not downloading web snapshots. This is actually happening on other staff computers as well.

    2) when a snapshot is "downloaded" the file cannot be found.
  • edited January 14, 2019
    www.nytimes.com:443 uses an invalid security certificate.
    This looks like another instance of the same problem you've seen before — secure connections on your network are being intercepted in a way that breaks Zotero. See SSL Certificate Error for details.

    Attachments are downloaded from Zotero, not from the Zotero Connector in the browser, so Zotero needs browser-equivalent access to the web. In a restricted environment, that means whitelisting the custom root certificate authority, as explained on the linked page. That you're getting this for the NYT and not for other zotero.org requests suggests that your IT department may have whitelisted specific Zotero domains instead, but while that will allow things like Zotero updates and syncing to work, it won't allow attachments to be saved or various other functionality (metadata lookup, open-access PDF retrieval, etc.) to work.

    The fact that you're getting an attachment with a missing file rather than no attachment at all (and a red X in the save popup) would be a bug, and we'll try to fix that, but, basically, things shouldn't be configured so that zotero.org requests work but others fail, so it should generally be clear what's happening. (When zotero.org requests fail, you get a proper warning on Zotero startup or sync.)

    However, in this case this is all somewhat irrelevant, because it looks like NYT pages can't really be saved properly at the moment. Zotero uses the page-saving code from Firefox, but if I save that page from either Firefox or Chrome as "Web Page, Complete" and then open the saved file, I get a broken page. Assuming others are seeing the same, we should probably disable NYT snapshots for now. (NYT pages are kind of out of control in general, technically speaking, so this is the sort of thing where simplified snapshots would come in handy.)
  • dtsillman - a very belated thank you for your thorough response. I will try this out with our systems department and report back.
  • A quick update. I am able to save web snapshots and then open the attachment in my Zotero client from both the NYT and other sites, like Art Newspaper. But other staff in the museum - like the curator I was helping above and some others - are not able to get snapshots from any website, not just the NYT. If secure connections on our network are being intercepted in a way that breaks Zotero, how come downloading from the same sites works on some installations of Zotero and not others? I've done all the installations myself with the same settings across the board, so I know we're all working with the same client configuration.
  • edited January 22, 2019
    (We disabled snapshots from the NYT, since they were broken for the unrelated reasons mentioned above, so you wouldn't be getting a snapshot at all from there now. You'll want to test on a different site.)

    If you're getting certificate errors on a given computer, the custom certificate authority just hasn't been properly installed into Zotero there. It's possible you tried to do it from a newer version of Firefox rather than Firefox 52 ESR, as explained on the certificate override page.
  • I'll work with our systems librarian to see if the the custom certificate authority has been properly installed on this curator's PC.

    We generally only use Chrome as our browser in the museum, so we're not using Zotero with Firefox.
  • The browser you use isn't relevant — since Zotero is currently based on Firefox and doesn't have its own UI for whitelisting custom certificate authorities, you have to use the mechanism from Firefox 52 ESR and copy a file over. This is explained in the page I linked to.
  • Thanks for clarifying.
Sign In or Register to comment.