Could you check the Downloads list to see whether Firefox auto-saved the PDF whose tab closed?

If you call up a pdf from history (usually typing .pdf in the address bar will find one), will it load into a tab that way?

If yes:

While I try to resist the cliche about turning it off and on again, could you switch the PDF handling action among different options to see whether that helps?

If no:

The viewer might be disabled. Could you check:

(1) In a new tab, type or paste about:config in the address bar and press Enter/Return. Click the button accepting the risk.

More info on about:config: Configuration Editor for Firefox. The moderators would like us to remind you that changes made through this back door aren't fully supported and aren't guaranteed to continue working in the future. Since we are returning to default, I feel comfortable mentioning it.

(2) In the search box in the page, type or paste pdfjs.disabled and pause while the list is filtered

(3) If the preference is bolded and has a value of true, double-click it to restore the default value of false

If this preference can't be modified and has a lock icon, your IT department may have disabled it using Enterprise Policy.

No, the PDF has not been auto-saved in Downloads.

Yes, I can call up a .pdf from history, and it loads into the tab.

Tried switching the PDF handling action to ‘save as file’, restarted Firefox, that didn’t work, still opened a blank tab that disappeared. Switched to ‘always ask’, same again blank tab. And same again with ‘use Adobe Acrobat DC’.

pdfjs.disabled was set to false.

Did the Downloads list show an error for that file, or is the download just completely missing like Firefox never tried?

On the link context menu, if Save Link As... is available, does it save the PDF or an HTML document?

Do you have any add-ons that block popup windows? Some of them have proven overly aggressive over the years, but I don't know of a new issue in Firefox 99 (too new).

The download is completely missing, doesn’t show up in the Download list.

It saves as an HTML document, but a blank one at that.

I have refreshed Firefox, so there are now no add-ons, still doesn’t work.

Hmm, how about Tracking Protection -- if you click the shield icon next to the lock in the address back and turn off Tracking Protection for the page where you trigger the download, does that make any difference?

If not:

Could you try to track the download using the Browser Console -- it's a separate window you can open with Ctrl+Shift+J. After a few moments, it may fill with messages related to other tabs. Click the trash can icon at the upper left of the list to clear that. At the upper right, click "XHR" and "Requests" to display all requests made by Firefox.

Leaving that window open, switch back to the page where you trigger the download and try it again.

Now back in the console, Firefox should have logged a GET or POST or XHR request for the file. Does it show an unusual status or are there any error/explanatory messages?

Tracking Protection did not make a difference.

One GET status in red, but I don't know what that means.

Hmm, I looked at that first URL, and it seems they run a PDF plugin detection script from 2016 (source available at http://www.pinlady.net/PluginDetect/).

Firefox 99 changed how the browser responds to probes for installed PDF plugins, to conform to new standards. My old test page at https://www.jeffersonscher.com/res/plugins.html now shows some rather odd text in Firefox 99:

• Name (Version): Chrome PDF Viewer
  Description: Portable Document Format
  File name: internal-pdf-viewer
• Name (Version): Chromium PDF Viewer
  Description: Portable Document Format
  File name: internal-pdf-viewer
• Name (Version): Microsoft Edge PDF Viewer
  Description: Portable Document Format
  File name: internal-pdf-viewer
• Name (Version): PDF Viewer
  Description: Portable Document Format
  File name: internal-pdf-viewer
• Name (Version): WebKit built-in PDF
  Description: Portable Document Format
  File name: internal-pdf-viewer

There is a Q&A about the change in comment 19 in this bug: https://bugzilla.mozilla.org/show_bug.cgi?id=1720353#c19 (note: https://bugzilla.mozilla.org/page.cgi?id=etiquette.html )

I don't know how to bypass this problem if the old script self-cancels the request when it doesn't detect support using the methods from 2016. Can you bring it to the site's attention? (I'm not sure why they even need to make it this complicated...?)

By the way, random thought, if you reopen the closed tab using Ctrl+Shift+T, does it stay open long enough to display any message?

Thank you for all your help, much appreciated!

I have sent an email to the site about the issue.

Reopening the tab does not work, it doesn't open the link tab at all, just older tabs that I've visited previously.

Thanks again :)