automatic link to pubmed

Until now, when doubleclicking the reference, I came automaticly to pubmed website. Now I'm automaticly forwarded to the complete reference website if its free available. Is it also possible to disable this?
  • May be its related to the same update of the translator. I am not sure.

    To clarify my question:
    Go to pubmed, search for the folowing title of Smartt and download citation:

    β-catenin negatively regulates expression of the prostaglandin transporter PGT in the normal intestinal epithelium and colorectal tumour cells: a role in the chemopreventive efficacy of aspirin?

    Usually when doubleclicking on a title in zotero you were linked to the abstract in pubmed. This was happen everytime until 1st october. My next citation download was on 8th october, and from this day on I'm linked to original journal page (in this case = BJC), and when no open source is available, nothings gonna happen when doubleclicking on the title.
  • Yes. That's related.
    What adamsmith said in the other thread:
    I can confirm that has changed, but it was done on purpose:
    Since PubMed is a catalog of articles and doesn't actually store any articles, the URL and accessed date shouldn't be part of the citation. The URL is still attached to the item as a link, though.
    You have to double-click on the child item ("PubMed Link") to open the abstract in pubmed.
  • edited October 10, 2012
    Just so it's clear what's happening, since the pubmed URL is not longer stored in the URL field and this field is empty, Zotero uses the DOI to redirect you to the article.

    This should happen for closed or open access publications as long as there is a DOI associated with them. IMO, if there is no DOI, Zotero should redirect to the catalog entry (PubMed).

    @ubauer, from your point of view, are you only concerned about "closed source" articles not opening a pubmed page, or do you have a reason to want pubmed page instead of the publisher's page opening for all articles? In my opinion publisher's webpage is preferred. I understand that PubMed offers certain features like "related articles" that are desirable, which is why the URL to the PubMed page is still attached to the citation.
  • In our business group we use to link to the pdf of the whole document if we have it in stock. So everybody is familar with the plus symbol and is used to think that the relevant pdf is available.
    Now all new entries have this "+" but without having the pdf.... This is not helpful in our case.

    Thats why we disabled this automatic link now. But e.g. for easier reading of the abstract we used to go back to pubmed. This doesn't work now after disabling of the link.
  • Thats why we disabled this automatic link now.
    What do you mean by disabled the link? Did you edit the PubMed translator to no longer attach the link?

    If so, you can also edit the translator to store that same URL in newItem.url
  • With "disable" I mean that we deactivated the common preference to automatically include snapshots and pdf documents. After that the link to pubmed was no longer included.
  • @Simon, Dan

    Is this behavior intentional or just an oversight? IMO it doesn't make sense that disabling file attachments should disable link attachments.
  • edited October 12, 2012
    I'm not really sure what the appropriate behavior is here. We can easily save the links even when snapshots are disabled, but then should we also save snapshots as links if they don't have the same URL as the item?

    It's also worth noting that either of these approaches would interfere with the OP's solution for determining whether a given reference has full text. I suspect the OP is not the only person to do this, although a tag would be a more appropriate, if more tedious solution.
  • @ubauer: You can get the old import behavior back when replacing the new pubmed translator with the old which you will find here: http://www.logies.de/NCBI%20PubMed.js
    Before you copy old over new disable automatic updates of site translators. I think your problems are a good example why it is seldom a good idea to change critical behavior of software without giving a option to keep old behavior.
  • Michael: It's not helpful to advise other users to install an out-of-date translator from an unofficial website and change their Zotero settings such that Zotero's saving functionality will break on many sites within a short amount of time. Given that all the same data is still being imported, the reasons for the change have been explained in great detail, and all Zotero developers are on-board with the change, it'd be much more helpful to focus on ways that the new saving behavior can be made to work with your particular workflow.
  • Dan, it is not my fault, that Zotero has not rolled back the changes to the Pubmed site translator (and has not communicated the problematic changes in advance to every user of PubMed as it should have happened). I think rolling back still should be the way to go till the problems with the new translator have been solved. I have no problem with my workflow, thanks. It`s working again as it should, IMHO. I`m not patient enough for simply waiting for another solution by Zotero, I have to accomplish work in the meantime. It is wrong that stopping automatic updating of site translators will break Zotero. First, many users will not be interested in much more than PubMed. Second, it`s easy to update the translators manually from within Zotero and after that to copy again the old PubMed site translator (which I`m offering on my website) over the new in the directory, where Zotero sits.
  • It is wrong that stopping automatic updating of site translators will break Zotero. First, many users will not be interested in much more than PubMed. Second, it`s easy to update the translators manually from within Zotero
    No, it's not wrong. Translators break all the time, and you have no idea what other sites people use. Yes, it's easy enough to update translators manually, but instructing users to disable translator updates without explaining the consequences is irresponsible.

    Many people use Zotero, and I believe there have now been a total of two reports about this, so you might consider the possibility that this isn't as huge a deal to most people as it is to you.

    As indicated in this and the other threads, we're happy to look for ways to improve access to catalog URLs, but this change was entirely consistent with Zotero's standard metadata practices.
  • I'm not really sure what the appropriate behavior is here. We can easily save the links even when snapshots are disabled, but then should we also save snapshots as links if they don't have the same URL as the item?
    I don't see a general use case for having such a link. IMO determining if a link like that would be necessary should be done on a per-translator basis.

    We could add a general "retrieved from" link to all records (or only where it's not the same as the stored url). This would be useful in a couple cases.
  • Dan, "and I believe there have now been a total of two reports about this", I think, 2 reports are already many more than you got (from actual users) complaining over the old PubMed site translator, right? "but this change was entirely consistent with Zotero's standard metadata practices" In the other thread I already discussed in some length why these practises are wrong, IMHO: Catalogs are original sources and should be treated as that. As some other arguments I brought (for keeping the old PubMed site translator) this was not answered. But I will not discuss further this topic, sorry, no time. I (with your help) have found a solution that will work for me for some time, that`s enough.
  • (Just because this is now being discussed and suggested here, there is a right way to do this:
    1. As Dan says, downloading translators from anywhere but Zotero is a security risk. Zotero uses a version control system for all of its code, so you can go back to any translator at any point in time. E.g. you can download the last version of the Pubmed translator before the URL change here:
    https://raw.github.com/zotero/translators/cb4769e82933875e4e387f5e82d5f4e02c43251d/NCBI%20PubMed.js

    2. The right way to keep using the old translator is to download it into the Translator directory using a different filename, then change the translator ID and lower the priority to 99. Zotero will then use that translator on Pubmed and will never update or change it otherwise, so you can leave that option enabled.
    )
Sign In or Register to comment.