Syncing multilingual sources

o_zelinska
I have two computers, two multilingual Zoteros of the same versions, and one on-line library I sync on regular basis. And still one computer gives me transliterated references, and another does not, as the program there does not see the multilingual component in the online library.
«1
  • sdspieg
    edited October 31, 2011
    To expand on this - we work with a distributed international team that uses (or tries to use) Zotero's multilingual capabilities in a 'group library'. The idea being that anybody can add 'their' references to the shared library, can then enter 'their' Zoteroed footnotes into Word and we can then generate a single nice (automated) bibliography at the end of the book.
    Yet as Olga wrote, we seem to have difficulties to make this work properly. Does somebody have detailed instructions for this? Thanks,

    -Stephan
  • fbennett
    Stephan,

    Sync of multilingual data is not yet supported, unfortunately, and the multilingual client is not supported by the core team. The best we can do in the short term would be to get the new duplicates detection mechanism working in the multilingual client, so you can safely import and merge items from another machine and have the multilingual data slide into place.
  • sdspieg
    edited October 31, 2011
    Frank,

    Nice to hear from you! Too bad the multilingual client is not supported - that is disappointing. The Zotero team has always seemed to me to get so many things right, but in an age of accelerated internationalization of research, it really is surprising that this is not a higher priority. Especially since you DID make a working version. Isn't there anything we can do to change their minds? Because I am pretty sure there are many (and will be more and more) of us who will want this functionality.

    In the meanwhile, yes - if what you describe would help us, we would be grateful if you could make that available. Thanks!

    Cheers,

    -Stephan
  • fbennett
    edited October 31, 2011
    It's not supported at present only because the team hasn't reviewed the code, but that could certainly change down the road. I suspect that it would turn on grounding a round of project funding on multilingual support. I haven't heard anything specific, but would agree that the prospects for that happening must be pretty good.
  • o_zelinska
    Dear Frank! Nice to hear from you :) Thanks a lot for your support! Could you elaborate a bit on the merging. I now face a situation when I need (again) to merge four libraries, two of them multilingual. Is there a way not to do it manually?
    Thanks, Olga
  • fbennett
    It will be like last time, unfortunately, but without the bugs in other places. The latest multilingual client includes the duplicates detection code from version 3.0, built by the core team. It's slicker than what you used before, and also preserves document links for you after the merge, but it doesn't yet support merging of multilingual data. What's your time frame for merging the sources?
  • o_zelinska
    It is rather limited. But the general number of sources is smaller this time. I do regularly update my Zotero, so I assume I have the newest version. So how do I start?
  • fbennett
    In the left panel, do you see a "Duplicate Items" folder?
  • o_zelinska
    I took care of the duplicates. And indeed, this is soo much easier!!! After I receive Chinese sources, do I import RDF file?
  • fbennett
    edited November 3, 2011
    That's how to get them into Zotero. When you do the merge, though, the multilingual data in the imported items will be ignored. I can take a look at making that work this weekend if that will help.
  • o_zelinska
    So you can make possible merging of multilingual sources. That woul be great! But I am not sure about the weekend. Otherwize I will have to enter multilingual data manually. Right? Will sqlite (opposed to .rdf library) help me somehow?
  • fbennett
    If it's a smallish job and needs to be done quickly, moving the fields across manually is your best solution. You don't want to work with zotero.sqlite directly. Copying the database file between systems is a way of moving exact copies of data around, but any modification to the database itself should be done through Zotero.
  • o_zelinska
    I have received Chinese sources in .rdf and I have to merge them. Multilingual data does not reflect. Is manual adding of relevant fields the only solution?
  • fbennett
    When you import the items, do you see the multilingual data in the imported entries? With the latest builds of multilingual Zotero, you should. Please confirm that that much is working.

    I have a gap in my schedule today, and may be able to implement a partial fix that will work for this particular case (one entry without, and one entry with ML data). Will post here again shortly.
  • o_zelinska
    No, there is no multilingual data in the imported items. When I checked with the Chinese expert, he said he has the same multilingual Zotero version as I do.
  • fbennett
    Do you both have the latest version installed (gear menu -> About Zotero should show 3.0m2)?

    If not, you can wait for a short while before updating; I'll soon make a release with basic support for multilingual field merging (I have ordinary fields working already, creators will take a few more minutes).
  • o_zelinska
    edited November 2, 2011
    Dear Frank, this new tool sounds exciting. I will be able to provide you with the feedback in 8 hours?:) I am out of the office by now. Thanks again for your support!
  • fbennett
    edited November 2, 2011
    The updated client (version 3.0m3) is now up. After backing up zotero.sqlite for safety, you can install it from the tools page at CitationStylist. Automatic updates should work fine after that.

    It should be sufficient to just update your own client for an initial trial. If I recall correctly, Zotero RDF exports have been working correctly all along, but subsidiary language data was being lost on import.
  • fbennett
    edited November 2, 2011
    Make that version 3.0m4. I've reworked things so that each language represented in a merge set will be represented, for each field and each creator, in the merged item. This should do what you want.
  • o_zelinska
    Dear Frank, this is working perfectly! Thanks a million!
  • fbennett
    That's great to hear!
  • o_zelinska
    Dear Frank, some feedback on the operating the new version. Nice and muuuch quicker than the previous one. Yet, if running docs like I do ( 80 pages, 200 sources) it is slow. And freeses often ( I restart Firefox in this case). Moreover, the biblio in the doc does not reflect the miltilingual data, giving me the sources in original languages. I have tried to generate biblio without language preferences, get the original list, and then update it - does not work. I also tried to set the correct translit, translate, and sorting preferences first. In the latter case the program responds that my format is wrong (!) and does not allow biblio generation. Is there any way to tackle this biblio question? Thanks!
  • fbennett
    Taking the failure of translation and transliteration first, one question to start. Are you setting the language preferences via the Document Preferences (not in Zotero directly, which only sets the defaults, but through the word processor)?
  • fbennett
    (Meanwhile, I'll look at it this morning. Possibly it's just not updating entries when refreshing.)
  • fbennett
    Olga,

    I've been able to reproduce some of the errors you are experiencing, but not all. When I change the language settings in Document Preferences (through the word processor plugin, not in Zotero itself), a refresh does change the text of the citations.

    There are a few glitches, though. The option to include the text of names in the original script does not seem to take effect, and sorting by multilingual variants does appear to be broken. Also, the language settings in the document do not seem to persist across sessions; when Firefox is restarted, the settings revert to the default values set in Zotero Preferences.

    I'll look into the sorting and persistence issues this week.
  • fbennett
    I've made a new release of multilingual Zotero that fixes some of the things I was able to confirm as broken. If you update, sorting by multilingual fields should work, and settings should persist in the document. Creator names in the original script are still not being rendered.

    You may see an improvement in stability, I'm not sure. As for the speed problem, well, it's not fast, but let's hope that will be tolerable if it does the job reliably.
  • o_zelinska
    Dear Frank,

    thank you for your support. With regards of your first question. The document generates biblio (with no multilingual data) ONLY if the settings are left blank. If I enter the needed language preferences, it DOES NOT generate ANY biblio, and provides me with the message "The current biblio style does not define a bibliography...".

    With regards to updates. Could I ask you to give me a hand and provide a link to the newest version, as my Zotero seems to experience and error while finding updates.

    Thanks again!
  • fbennett
    Right, that's the error that's fixed by the update, so we're seeing the same behavior. Here's the installer link.
  • fbennett
    Glad to report that after a little correspondence off list, the source of the problems raised by Olga have been identified and fixed. All's well.
  • sdspieg
    Frank,

    Thanks much again for all of your help. We have now just finished our latest book with - again - a beautiful multilingual bibliography. This would have been impossible without your kind assistance.

    But could you please specify where things now stand with respect to the Zotero core team's position on multilinguality? We certainly love this combination and I remain convinced that the increasingly global and networked research community will require a consistent approach to multilingual citation. Whereby footnotes and bibliographies can be in the original language, with consistent methods to add language-specific transliteration and translation for every cite. You have already made huge steps in that direction. But having to export and import rdfs constantly across an international team remains cumbersome. So the built-in collaborative features (including sync) of the main Zotero trunk would certainly come in handy here.

    So anyway - maybe you could enlighten all of us out here on what your thoughts are about the future of mulitingual Zotero (and/or of Mulzero)? Thanks!

    -Stephan
Sign In or Register to comment.