Firefox busy after each field edit
Hi,
Using beta 2.0b5 w/ FF 3.0.11 on Windows XP Pro (SP3), I am finding my system to be unacceptably slow. When I finish editing a field in Zotero, such as an author's name, the system slows down. According to Process Explorer, whenever this happens FF is taking up about half the CPU cycles. (I believe the system limits individual applications to 50% of CPU cycles.)
Is this possibly because Zotero is trying to sync the library after every field is edited? If so, is there a way to tell Zotero to suspend syncing? Since I was adding reference information for several articles in a special issue of a journal, it would probably be best to suspend syncing until I've finished all of the articles rather than sync after finishing with each individual reference.
Using beta 2.0b5 w/ FF 3.0.11 on Windows XP Pro (SP3), I am finding my system to be unacceptably slow. When I finish editing a field in Zotero, such as an author's name, the system slows down. According to Process Explorer, whenever this happens FF is taking up about half the CPU cycles. (I believe the system limits individual applications to 50% of CPU cycles.)
Is this possibly because Zotero is trying to sync the library after every field is edited? If so, is there a way to tell Zotero to suspend syncing? Since I was adding reference information for several articles in a special issue of a journal, it would probably be best to suspend syncing until I've finished all of the articles rather than sync after finishing with each individual reference.
Syncing happens 15 seconds after the last edit, though, so that may not be the problem you're experiencing. How many items do you have in your library? (Select All, look in right-hand pane.)
I have just switched over my EndNote library to Zotero and have the same experience (Likewise using Zotero 2.0b5, Firefox 3.0.11 and XP ). I started testing Zotero on a small library ~300 entries, which worked really well. Since importing the rest of my library (~3000+ entries) Zotero has slowed down incredibly every time I select a Tag or edit a field. It seems however to be slow only when I have the Tag selector open. Perhaps this might give a hint as to what could be slowing things down. By the way I have the Sync feature disabled. Removing the pdf text indexing as well caused no change.
Thanks again for a fantastic tool
I tried disabling, but it's hard to tell if this makes much difference. Both w/ and w/o auto-sync take a while but not as long a previously. I'm not modifying records exactly like the ones that gave me problems, but then again, the Internet may be less congested, etc.
I'm definitely having failing syncs. I posted something about this earlier today.
I have 2826 entries in my library.
A few things happen if I close the tag selector. (1) I still failing syncs. The exclamation point has become an almost permanent fixture. (2) I tried reopening the tags shortly after importing a new reference and got the message, "An error has occurred. Please restart Firefox. ..." (3) Response time is still slow, but not as slow as when I reported the problem. I tried editing new references with and without the tag selector open. Although both are unacceptably slow but not as slow as when I reported the problem, I cannot discern one as being faster than the other. BTW, how would one isolate Zotero from other things Firefox or the OS might be doing in the background in order to determine what's causing the slow performance?
zotero(5)(+0000000): Binding parameter 1 of type int: 1808192
zotero(3)(+0000001): Refreshing item boxzotero(5)(+0000914): SELECT creatorTypeID AS id, creatorType AS name FROM itemTypeCreatorTypes NATURAL JOIN creatorTypes WHERE itemTypeID=? ORDER BY primaryField=1 DESC, name
zotero(5)(+0000000): Binding parameter 1 of type int: 2
zotero(3)(+0000183): Switching lastCreatorFieldMode to 0
zotero(3)(+0000095): Switching lastCreatorFieldMode to 0
zotero(5)(+0000000): SELECT A.itemID, value AS title FROM itemAttachments A NATURAL JOIN items I LEFT JOIN itemData ID ON (fieldID=110 AND A.itemID=ID.itemID) LEFT JOIN itemDataValues IDV ON (ID.valueID=IDV.valueID) WHERE sourceItemID=? AND A.itemID NOT IN (SELECT itemID FROM deletedItems)
zotero(5)(+0000001): Binding parameter 1 of type int: 1806876
zotero(5)(+0000000): SELECT N.itemID, title FROM itemNotes N NATURAL JOIN items WHERE sourceItemID=? AND N.itemID NOT IN (SELECT itemID FROM deletedItems)
zotero(5)(+0000000): Binding parameter 1 of type int: 1806876
zotero(3)(+0000000): Refreshing item box
If the quick search found a lot of things, this could appear to crash Firefox (unless you've got the terminal open). I can just clear the quick search before editing the item - but is there any way to not do what Zotero is currently doing?
Firefox 3.5 latest dev build of Zotero.
Regards,
Jon.
Autocomplete suggestions are nice, but only if they come up without any delay. As soon as there is a delay for more than a few hundred milliseconds, they become as annoying as speed bumps.
Similarly, the Quick search lag is terrible with larger libraries - I never use it as 'Quick search' anymore, instead prepending " to get into Advanced mode. Wish the 'search if I press Enter'-mode were an option.
(I always have the tag selector closed and auto-sync off.)
This is with auto-sync off. I haven't tried this on editing Place or Publisher.
Running Zotero 2.0b7.6 in Firefox 3.5.7 on Win 7.
1GB RAM, Intel Pentium T2080 @ 1.73 GHz
The Debug ID is D251919676.
My problem ID 1289202139
As for the corruption of names, it sounds like an encoding issue. Can you provide a URL of a page that consistently causes this issue?
Re configuration (poirm} - Office for MAC had an update a few weeks but don't think that this where the problem lies for me. I have not had to update Zotero plugins recently that I recall.
Re Embedded RDF (ajlyon). The sites from which I obtain bibliographic references are almost exclusively EBSCO and ABI PROQUEST accessed through my University VPN. I cannot capture my references through the brown icon top right of the browser URL bar and Dan has looked into this in the past. It is something to do with how my University structures access. I normally get known translator issue when I try. So I tag target references, push to the "folder" provided by EBSCO or PROQUEST, go to folder view, select the references and export to Bibliographic Reference tool. I have always had to to do it this way with only intermittent times when it works through the browser. The issue of capture as web sites however is a recent phenomenon - say 6 months. And it is not all that frequent - and I catch it through serendipity only so I might still be catching legacy issues. I will observe each download in future to see if it happens and from where.
The corruption of names - I caught that only once but I have not gone through my 1000 references one by one. I don't know from where I obtained that reference so I will be more observant in future as I obtain references and if I spot it again in the near future, I will post to this thread the details.
I think that the corruption of names is somewhat more likely with the export - import workflow that you're using, but we do have ways to deal with it. Again, we'll need an example of such an entry.
Hopefully, we can pin these down soon and fix whatever is behind them.
My operating system is a Windows XP Pro, and I'm using the latest versions of Zotero FF and Standalone.
But you may want to delete the apostrophes from the entries in Zotero and retype them manually. They're probably encoded in some weird way that Word, with its sometimes flaky Unicode support, is mangling.
Poirmw: Naive as I am, I think this is a Word for WIndows plugin thing. I use Word for Mac and have not yet had that issue. My corrupted file entry was in Zotero itself and seems to be focused on three names one article, all the names using accented vowels. I often get funny symbols in my free form notes I attach or if I cut and paste into the abstract box. Here is an example from a note I associated with an article
Propositions numbered P1. 1.CRM i
That O with a hat and the C with the cedilla (sp?) happens a lot but I ignore them when reviewing old notes as it does not obscure what I have written. So I have never chased this. But these things occur in Zotero fields and not in word.
Today everytime I try to link a pdf file to a record in Zotero it hangs firefox and I need to force quit and reopoen; very frustrating. I can never seem to get an ID code it is either greyed or when the browser is hung I cannot get to it.