Can't get PDFs at all with Chrome (not behind proxy)

jbaart

First of all, I'm not behind a proxy, I checked.

I'm new to Zotero and trying to add items to the standalone version through the Chrome connector. I can see the icon next to the URL and can use it, but it only adds an item with the metadata and a snapshot, but never PDFs. The setting "automatically attach associated pdfs..." is active.

Can someone please tell me whether this example should work or not, thanks:

http://www.eludamos.org/index.php/eludamos/article/view/vol1no1-2

If it should work and it's only me, what could be causing this, especially as I'm not behind a proxy which seems to be the most common root of problems like this.

Thanks in advance.

adamsmith

Try http://www.advancesindifferenceequations.com/content/2014/1/318

the pdf from the page you link to wouldn't download ever (if you follow the link to the PDF, which is also what's in the metadata on the page that Zotero uses it's displayed within a webpage, so Zotero can't download it as a PDF. We'd need a dedicated translator to work around that issue, don't think anyone here will have time to write that any time soon)

jbaart

Thanks, your example does indeed work on my end. Good to know this is not a general problem on my end. However, I can't help but wonder why there is no dedicated translator in existence for this type of site. It's one of the possible configurations of Open Journal Systems (https://pkp.sfu.ca/ojs/) which is used by plenty of journals out there. At least three of the major game studies journals alone use exactly this kind of implementation displaying the PDF as a webpage.

adamsmith

We do have an OJS translator, but it doesn't work on that page and it wouldn't be a super-quick fix to make it. Maybe aurimas who wrote it wants to take a look. It's a bit trickier because of the frames (ick).

aurimas

I don't see a way to identify that website as using OJS. We usually do this by looking for a link with href "http://pkp.sfu.ca/ojs/" and id "developedBy", but some other/additional reliable way would be acceptable too. If we could identify it, then our current OJS translator would already account for the PDF view page (we'd need to make a small tweak to work with the frames).

jbaart

If it helps, here's another journal with the same problem:

http://journals.sfu.ca/loading/index.php/loading/article/view/122

You are quite right, there's no OJS or Open Journal Systems tag or similar in the source code which one could analyse. It seems the mention of Open Journal Systems might be mandatory on the Home and the About page, so possibly one would have to open these and look for the text there.

Edit: Forget that, here's another one which does not mention OJS at all. However, the PDFs get downloaded on this website. Weird.

http://press-start.gla.ac.uk/index.php/press-start/issue/view/1

aurimas

I'm not sure to what extent you have the ability to customize this, but if you include <meta name="generator" content="Open Journal Systems X.X.X.X" /> (with whatever version) we can pick up on that. I see that the generator is included on the "issue view" page on that last link, so there might be an easy way to include it on all pages. Might be worthwhile to request this upstream though.

jbaart

Unfortunately I'm just a consumer of these journals, but of course it might be worth contacting their administrative staff to let them know about these problems.

aurimas

Ah, ok. Sorry for the misunderstanding. We'll take a look and report this upstream if necessary. No immediate solution for you unfortunately.

jbaart

That's okay, thanks for taking the time to investigate.