Bad interpretation of metadata (author's first name)?

Ulrich Windl
When viewing the URI http://it-fixes.povolotsky.info/2016/06/windows-10-date-time-frozen.html I noticed that Zotero (4.0.29.16 for Firefox 51) adds an extra "author" to the author's first name when adding the entry from metadata.
However I'm unsure whether it's Zotero's fault, or the web page's. But the experts will find out, I'm sure.
  • adamsmith
    well, "fault" isn't necessarily the right category here. The site has minimal metadata, and in a last attempt for an author name we're getting the author from
    <span class='post-author vcard'> We could try to clean that up (e.g. in this case the name is in a separate span with itemprop="author") , but given that this isn't something that's standardized much, we could just as likely be breaking other things if we try. If someone wants to take a look, I'd be happy to review this as a pull request, but I'm not going to do it myself.
  • Ulrich Windl
    My suggestion is to make the metadata detection more strict, i.e.: If the metadata does not conform to a robust standard, don't offer to add a page using the metadata. Maybe the page given is just a nice playground ;-)
  • adamsmith
    it's a trade-off, for sure, but I prefer Zotero to at least try and grab the best available metadata to giving up on all sites that don't have something super standardized (which is actually pretty hard to tell). If anything, we'll may be moving in the direction of removing the distinction between recognized and not-recognized pages entirely (because that improves, e.g., saving from mobile devices).
Sign In or Register to comment.