Zotero not responding even for small tasks

cappellettig
Dear All,

I have decided to migrate from Qiqqa to another reference manager, and I am currently trying out both Zotero and Mendeley. While I overall prefer Zotero, I have encountered a problem that is making me lean heavily towards Mendeley.
Namely, everything I do in Zotero freezed the application for a varying amount of time. For example, even importing as little as 5 small pdf references will freeze the application for as long as 5 minutes, and using Zotfile to rename/organize my pdf files through Zotero freezes the application for even an hour.
By "freeze" I mean that it is marked as "Not Responding" by Windows, and if I try to close the program it tells me that "Xulrunner is not responding". Now, I understand that leaving the program idle and letting him do its work eventually get the task done, but is it normal that it freezes in this fashion for so long?

Thank you for your time and consideration, best regards,
GC
  • aurimas
    but is it normal that it freezes in this fashion for so long?
    No, this is not normal. How many items do you have in your database?

    Could you submit a Debug ID for some action that takes longer than it reasonably should, but try to find something that's on the order of minutes (5 should be ok), not an hour.
  • cappellettig
    edited January 10, 2014
    Dear aurimas,

    Thank you for your kind response.

    Here is a DebugID: D917870075.

    The operation - renaming 9 references through Zotfile - took more than 5 minutes.
    I will post another DebugID for adding references shortly.

    Thank you again for your time and support,
    GC
  • cappellettig
    Another Debug ID: D779242173.

    It took more than 5 minutes of "Not responding" to add 10 references.
    Mendeley required like 30 seconds, and the software was still 100% responsive in the meantime :-(
  • aurimas
    As I said, this is not normal. Adding 10 references (from where btw?) should take from 1 second (from a file) to about 15 seconds (from a web page depending on your internet connection and website).

    Dan will have to take a look at the Debug logs.
  • cappellettig
    The references were from pdf files stored on my computer. I have noticed that large-ish files take much, much longer. Specifically, I have tried to add one 18mb pdf, and it takes way longer than adding ten 1-2mb articles (which still takes almost 5 minutes, and that is definitely not normal according to what you say).

    Thank you again for your help,
    GC
  • aurimas
    So you mean you're dragging and dropping a PDF file into Zotero and it freezes up? (I'm not really clear on what you mean by adding references _from_ pdf files). Can you describe in detail what you did for that last debug ID?

    In either case, renaming the files using ZotFile for that first Debug ID should not have taken this long.
  • cappellettig
    I mean that I click "Add", "Link to File" and select the pdfs I want to add to my library.
    I know that Zotero is more web-centric (and I have no problems adding stuff directly from websites such as Jstor), but I have many pdfs already, and this seems like the way I am supposed to add them to Zotero. Am I doing it right?

    Thanks again,
    GC
  • aurimas
    If you're set against having Zotero manage your PDFs in its "internal" folder structure, then yes. Otherwise, you could just drag and drop the PDFs into Zotero or select Store Copy of File. I'm assuming you're using Retrieve Metadata after adding the links?

    Where are the actual PDFs stored? Are they on your local hard drive or perhaps a network share?
  • cappellettig
    Actually, I tried using the "Store Copy of File" option, but the time required to import the pdf does not change. Yes, I do retrieve the metadata for the pdfs once they are in my library.

    They are stored on my local hard drive.
    And by the way, my computer is a decent one (Lenovo X200 Tablet with P600 and 3GB of RAM, for the record), so I doubt it's a problem of the computer. Besides, Mendeley is importing everything in a matter of seconds.

    Thanks again,
    GC
  • cappellettig
    Ok, I think I figured out the problem.
    I tried to install Zotero on a different computer and everything seemed fine - it imported pdfs really fast and seamlessly. I then activated PDF indexing and the problems started...up to 5 minutes of "Not responding" for a handful of articles.
    Deactivating PDF indexing solves the problem, but of course it drastically reduces the usefulness of the entire software - especially as opposed to Mendeley and Qiqqa, which have full text search. Is there a way to tell Zotero to index pdfs in the background after importing them?

    I really love this software, but honestly this would be a dealbreaker for me :-(
  • aurimas
    You can disable indexing during import (by setting either maximum pages or maximum characters to index per file to 0) and then re-enable it when you're done organizing your library and leave it overnight. Unfortunately, indexing does not currently run in the background. I suspected that this was due to indexing, but I still think that the times you report are way too long.
  • cappellettig
    I just found this plugin: https://github.com/friflaj/zotero-auto-index

    After activating it it seems that the indexing is a lot faster. Maybe it's just psychological because I really want to love Zotero and use it, but it does seem a lot faster. Just as an information, I simply have to install the xpi file from Zotero itself, right?

    It does not solve the problem, but definitely makes it more usable. Background sync should definitely be on top of the development list, imho.

    Thank you again for your kind help!
  • aurimas
    Just as an information, I simply have to install the xpi file from Zotero itself, right?
    Not sure what you mean.
  • cappellettig
    I mean that I just select the "Install Add-on from file" function in the Zotero Add-on manager and open the xpi file of the Add-on. There is nothing else I should do right?
  • adamsmith
    yes, that's all.
  • cappellettig
    Thank you all very much for your help, I really appreciate it!

    All the best,
    GC
  • dstillman Zotero Team
    edited January 10, 2014
    This may not be relevant for you, but note that part of the reason the Zotero Auto Index plugin is faster is that it doesn't work properly for various languages (e.g., Asian languages), at least based on my reading of the code.

    Zotero's built-in PDF indexing is currently pretty slow, and we have plans to improve that in a future version, but it also looks like you increased the number of pages indexed from the default. So if you did that and then indexed some large PDFs, that could easily slow down anything that uses the full-text index going forward.

    Unless you have a huge library, just indexing normal-sized PDFs with the default settings wouldn't produce the kind of slow-downs you're seeing.

This is an old discussion that has not been active in a long time. Before commenting here, you should strongly consider starting a new discussion instead. If you think the content of this discussion is still relevant, you can link to it from your new discussion.

Sign In or Register to comment.