Upgrade StorageErrors when I attempt synchronisation using Zotero 2.0b6.1
I have just tried uploading my library to the online Zotero server, using Zotero 2.0b6.1 and have failed (Report ID: 498962374). There was no repeat of the earlier problems with Script stack overflow errors, but the error report shows a large number of Javascript errors. Could the fact that my library is a 'hybrid' i.e. a majority of the references in it have been captured by Zotero in the normal way, but there is a significant minority of references have been imported from Biblioscape 6, be a factor in the generation of these errors, and is there anything I can do about it.
Many thanks,
Al
Many thanks,
Al
This discussion has been closed.
Upgrade to 2.0b6.2 and try to sync. If you get an error, tell us what error you see when hovering over the sync icon, and also send in another error report before restarting and post the Report ID here.
As a cross-check, I upgraded to 2.0b6.2 on my laptop, which is greatly more powerful than the desktop machine, and tried to replicate the experiment on that. I did indeed get a sync failure when I tried to sync that, but with a very different set of errors. I have submitted that second error report to you (Report ID: 79071089) to see if there are any indications as to what the problem might be. Both of the computers are running on a network which requires a proxy in order to use it.
Al Gemmell
It's actually entirely possible that you'd get the same error from just trying to load the note in the interface, though possibly not.
If you're having trouble finding the note, you could upload your DB to the DB Repair Tool and e-mail the Upload ID to support@zot...org, and we can run a query for the largest note length. If you do this, you can ignore the download links the tool provides.
We may be able to work around this in the code.
As a check, i will upload the DB to the DB Repair Tool, and email the upload ID to the Support site.
Fatal error: Allowed memory size of 314572800 bytes exhausted (tried to allocate 289267336 bytes) in /usr/share/pear/Archive/Zip.php on line 2190
This surprises me, as according to the computer here, the size of the zotero.sqlite.zip file is only 35.312kb. Is there an alternative way of sending the file to you for inspection?
1) Is there a reason you've created a number of these? As far as I know, there's no way these would appear correctly as pictures within Zotero notes, at least after a reload—though you seem to suggest that they appeared as graphics immediately after pasting? (That would be surprising but not impossible.)
2) What PDF reader did you copy these from? I think there's probably a bug in either it or in Firefox, since data on the clipboard is typed, and Firefox doesn't support RTF data, so either the PDF reader is identifying the data incorrectly or Firefox is grabbing data of the wrong type. The former seems more likely (though I'm not sure how or why a PDF reader would generate RTF data in the first place). At any rate, you shouldn't really be able to paste in raw RTF data. The database itself is 181MB. The notes table itself is about 100MB.
There are a couple of reasons why I tried to copy images - most commonly maps from papers relating to my research interests in glacial history (Ice Ages etc.). The first was an attempt to save space, to keep the size of the database to a minimum, by taking copies of only those elements of some papers which I might want to use for teaching purposes, rather than have the entire paper in the database. The more important reason for 'extracting' diagrams and pasting them into notes was that I could then title the note appropriately from the content of the diagram, and that would allow me to search the database for suitable diagrams if i wanted to put some together for a class. I thought this might be rather faster than searching the database for figure captions contained within complete PDF files.
Could the upgrading of the Zotero database from 1.0.10 format to 2.0b6 format possibly have transformed the captured figures into RTF character sequences?
2) What PDF reader did you copy these from?
I have been using Adobe Reader 9, though I would think some of the diagrams could have been copied from PDF files when I was using Adobe Reader 7 or 8.
The size of the database you recorded (181Mb) as receiving from me is the same as that I have here, so i think the entire database must have got through to you intact, despite the error message I received.
As a further check, I then closed Zotero, shut down Firefox, restarted the browser, reopened Zotero, scrolled to the relevant reference, and the map is still there in the Note!
One thing I did note this time is that all that has been captured by the Copy is the actual artwork. Place names etc. on the original map have not been copied across, though I presume this is to do with the way Adobe Reader functions rather than anything to do with Zotero.
Your comment about not being sure how it might have worked in 1.0.10 as the notes there were plain text makes me realise that I do not recollect going back into any of the diagrams I had saved as notes - I have been presuming that because I could see the image when I pasted it into a note, that was it secure for the future. Today's experiment is the first time I can remember consciously going back to a note later to see whether or not the image was still there.
In your new note, can you click the HTML button and look at the source code? Based on posts in that thread, I'm guessing that you'll find an <img> tag with a file:// link to the image in your temp directory. This behavior is a fluke that should be prevented if possible, since those temp files could be deleted at any time. But it's at least dealing with HTML, and so seemingly not related to the RTF issue.
At this point, I'd recommend going through those large RTF notes (which you can probably find with a search for 'rtf'), copying them into text files in Notepad, and saving them with .rtf extensions. (You may need to adjust the extension outside of Notepad after saving, and possibly adjust your Windows extension display settings as well.) That should give you RTF files that you can open in WordPad or Word, which you can then drag back into Zotero as file attachments. Once you've confirmed that that works, you should delete the RTF notes and empty your Zotero Trash folder. After that, with any luck, syncing will work.
I will now start locating all the large RTF notes and following your suggestions re converting them to file attachments.
Thanks for the assistance with this great product.
An error has occurred. Please restart Firefox. You can report this error by selecting "Report Errors" from the Actions (gear) menu.
This is what I have done. No matter which folder or sublibrary I select, the same message appears in all of them. Restarting Firefox brought back the reference library, but when I tried to empty the trash again, the same thing happened as before. I am not sure what to do now to get rid of the trash.
It's possible that SQLite Manager, being within Firefox, would exhibit the same problem, though, so a different SQLite client might be necessary.
Al
Is it possible that I have got the sync settings wrong (I have no WEBDAV server set up at present).
How can I go about determining which note is the offending one (i.e., the one that's too long)? I have tried to search from the date of my last sync, and I have tried deleting some lengthy notes, but that still hasn't resolved it.
Many thanks!
SELECT title, LENGTH(note) FROM itemNotes ORDER BY LENGTH(note) DESC LIMIT 5;
The top note should be the offending one.
The error you're getting trying to empty the trash is the same one you're getting here, so you'll need to wait for a fix.