a bit more detail would help -- which PDFs, did this happen after retrieving data for lots of PDFs already, does it also fail for PDFs that have previously worked...
I've just starting using Zotero. I've managed to retrieve it for a couple of PDFs but the other ones just say that the meta data can not be retrieved. When I tried the same ones with Mendeley there was no problem.
Plus when I try to add PDFs from my computer it comes up saying that the format doesn't match.
Zotero uses different methodology from Mendeley to retrieve PDF metadata, so it's certainly possible that it wouldn't find something where Mendeley does. Do you have sample PDFs that aren't working? (link to a copy on the web or on your Dropbox).
If you getting the format error, though, you're likely trying to "Import" -- which is for data formats (e.g. exports from Endnote or Mendeley) not files. You can simply drag files to Zotero.
Could you provide the exact message you're getting? You can save every file type in Zotero, so something is off when you get a message referring to a problem in formatting when adding any file -- let alone a PDF -- to Zotero.
Zotero isn't able to find metadata for either of those files -- since it relies on Google Scholar mainly for that purpose, it does best on academic articles. Reports work pretty irregularly (though you can obviously add the metadata manually: right-click --> create parten item).
Right -- that's almost certainly the message you get when you use Import and then select the file. I'd be very surprised if you got that when dragging a file straight to Zotero.
we don't know exactly (it being closed source) but I think they rely a lot more heavily on a) their internal catalog of PDFs including some sort of hash-sums (which is interesting but _much_ harder to do than it may sound). b) trying to guess the title of the PDF (which Zotero doesn't do at all)
I have no idea how the two tools currently stack up. The last systematic test I've seen was in 2010 or so (with Zotero having a tiny edge) and since then I've seen people claim better success with either tool -- generally I'd kind of expect Mendeley to do better, since they put much more effort into that and much less into web importing, but one would have to test systematically to actually know.
I don't know about a) or b) but I know that Mandeley also uses user library documents as its resource. I have found that the wrong document is often found and matched. Even when the correct document is identified in the combined user libraries, the metadata has obvious major errors. I learned this by performing an experiment in September 2016. The incorrect metadata was clearly hand-entered by the owner of the Mandeley library record that contained the document. The hand-entered fields very often contain all manner of inappropriate or incorrect text. In some cases, I suspect that the individual library owner intentionally and maliciously entered incorrect metadata.
edit: Some metadata fields contained curses and vulgarities; some contained social / political statements. I found what were clearly intentional errors in citations of Andrew Wakefield's articles. Reports on torture and ptsd among refugees were vandalized by Mendeley users with accounts that only contained records with errors and these were only records on controversial topics.
In fairness, much of the matched metadata is okay but enough is wrong so that I can't trust any of it.
I always want reliably correct data over metadata with errors. I believe that no data is far better that east-to-get metadata containing errors.
Upgrade Storage
I've just starting using Zotero. I've managed to retrieve it for a couple of PDFs but the other ones just say that the meta data can not be retrieved. When I tried the same ones with Mendeley there was no problem.
Plus when I try to add PDFs from my computer it comes up saying that the format doesn't match.
I may be doing something completely wrong...
If you getting the format error, though, you're likely trying to "Import" -- which is for data formats (e.g. exports from Endnote or Mendeley) not files. You can simply drag files to Zotero.
So here are a couple of examples:
https://www.unisdr.org/2015/docs/climatechange/COP21_WeatherDisastersReport_2015_FINAL.pdf
https://www.ucl.ac.uk/global-health/project-pages/lancet1/ucl-lancet-climate-change.pdf
Thanks for looking into this for me.
Zotero isn't able to find metadata for either of those files -- since it relies on Google Scholar mainly for that purpose, it does best on academic articles. Reports work pretty irregularly (though you can obviously add the metadata manually: right-click --> create parten item).
When I try to import it reads:
Confirm
The selected file is not in a supported format.
(Options)
View supported formats (which brings you to zotero website) or
OK
a) their internal catalog of PDFs including some sort of hash-sums (which is interesting but _much_ harder to do than it may sound).
b) trying to guess the title of the PDF (which Zotero doesn't do at all)
I have no idea how the two tools currently stack up. The last systematic test I've seen was in 2010 or so (with Zotero having a tiny edge) and since then I've seen people claim better success with either tool -- generally I'd kind of expect Mendeley to do better, since they put much more effort into that and much less into web importing, but one would have to test systematically to actually know.
edit: Some metadata fields contained curses and vulgarities; some contained social / political statements. I found what were clearly intentional errors in citations of Andrew Wakefield's articles. Reports on torture and ptsd among refugees were vandalized by Mendeley users with accounts that only contained records with errors and these were only records on controversial topics.
In fairness, much of the matched metadata is okay but enough is wrong so that I can't trust any of it.
I always want reliably correct data over metadata with errors. I believe that no data is far better that east-to-get metadata containing errors.