SLOW entry of new references

I have just upgraded my RefMan from Version 10 to Version 12.  Some of the new features I like, but entering new references is a royal pain.  RefMan 12 is INCREDIBLY slow to update fields as I type.  I have a six month old desktop running Windows XP with all of the updates installed.  Typing in the title field is easily several words behind my typing, making it painfully slow to type in titles without making mistakes.  Typing in authors is MUCH worse.  In the RefMan10, the author pop-up would update IMMEDIATELY with no lag, so that one could quickly use an author already in the database if needed.  Now it takes FOREVER for the pop-up to update, so much so that I will probably inactivate the pop-up altogether. 

Is there any way to speed up the pop-up so that it is actually usable??

Why would RefMan12 be so much slower with this than RefMan10??

I have over 30000 references in my database (I am a LONGTIME RefMan user!).

Thanks very much for your assistance.

Michael Hahn


This problem began with Version 11 and occurs with very large databases.  Mine was at 32,000 references when Version 11 came out and the slow-down was iimediate.  I eventually contacted Thomson Reuters and they were aware of the problem.  After several  email discussions, their TechRep offered the following soution: “Use a smaller database.”  I was floored at this suggestion.  It is akin to taking your dog to the veterinarian because of excessive drooling, and being told that the solution is to teach your dog to spit.  It is no soultion at all.

After some exploration, I found that TR’s suggestion was the only one that works.  I split my database into a very large one and a very small one and enter new references into the small one only.  And it works.  Of course, I have to remember to set “All Databases” when I search, and any references I want to “Cite-While-You-Write” are in a front window. 

Some time after I figured out the work-around, I heard again from TR’s tech rep, who informed me that they were going to look into the problems that those of us with really large databases were having.  Haven’t heard a word since.  And Version 12 has all the same problems as did Version 11, and some new ones, besides.

On the good side, “Cite-While-You-Write” started to work well with Version 11 and still does well with Version 12.

I started using Reference Manager in 1986.  Now my database is us about 40,000 references (combined), and I have another that is at 15,000 references.  So, what should we mega-database users do?  I think Reference Manager’s days are numbered, and it will soon go the way of ProCite, another TR product that was discuntinued a couple years ago.  If you don’t believe that, look at the number of posts on the Reference Manager forum (ca. 500) versus the other remaining TR product, Endnote (ca. 6000 posts, not to mention the Swedish and Korean Forums for that product).  And consider how few replies are made by the tech reps on this forum. 

And what about this problem for incongruity:  Use Cite-While-You-Write to assure accuracy in your scientific papers, but backup your database and then restore it, and half of the letters with diacritical marks have been magically converted to random symbols, that are then included in the references in your manuscript.

This is an FAQ from the Reference Manager website, under the “Support and Services” tab:

“Install 22: What are the data limitations in the program?

Answer: There is no limit to the number of databases you can create, assuming that you have the disk space to store them. We recommend that you keep all the references you cite in your papers in one main database.

There is no limit to the number of references you can create in a Reference Manager database. It is recommended that the number not exceed 100,000 to maintain efficient performance managing the database.

I can get by with a limit of 100,000 references per database.  When will Reference Manager actually do that?

1 Like


We have a similar problem. We use RefMan12 network and a database containing 13.000 records. When opening the database in read-write share mode the database is very very slow. I have got the advice from technical support, that I should open the database and then sort the column author. It seems stupid, but have changed the very very slow to just slow!!

Our database is very old and has been converted a lot of times - every time a new version has arrived. I suspect, that some of the records are kind of making failures and other not-wanted behaviour. E.g. There is a lot of records without - and these records are very difficult to use in Word and be recognized as PubMed records. Only if you insert the refid in your word document it seems to work ok.

Kind regards


1 Like


I just tried your suggestion of sorting the author column, and it seems to work.  I want to explore this magic a bit further, but it holds promise.

I stand by my prediction of a doomed future for Reference Manager.  The fact that you, a RefMan user, addressed this issue rather than a technical reprepresentative reinforces my viewpoint. 

Why wasn’t this strange little work-around posted somewhere on the Reference Manager site?  I’ve looked through all of the FAQs without finding any mention of the slowing issue, especially vis-a-vis large databases.  It seems a bit rude of them to sit on a solution, even such a quirky one, while I have been struggling with this issue for a couple of years.

Hanne, you wouldn’t happened to have discovered a solution for the problem of losing diacritical marks when backing up a database to transfer it from one computer to another?  That is a huge issue if one uses “Cite-While-You-Write”.

Thank you for your help.


Now here is an interesting follow-up to my original post.  I have noted two things about the slow-down.

1)  I was importing some references on the web this evening, editing the keywords after import (on the imported references tab), and encountering the usual slow-down with the pop-up during entry.  

2)  The entry of a new reference continues to be slow, as before.

3)  BUT, if one edits an entry on the main database window, editing is MUCH faster.  The pop-ups update quickly and the process is almost as fast as it was under RefMan 10 (though it is still slower).  

SO, RefMan Tech Support, WHY would the editing be so slow in the new entry and imported references windows of RefMan, but much faster on the main database window??

Anyway, I think I will do all of my editing on the main database window from now on.  New entries will just get a rudimentary entry to get a RefID assigned, and I will do most of the entry on the main window afterwords.




About losing diacritical marks.

We have experienced, that references been made by a copy/paste can cause, that invisible codes are saved with the reference. Later, can this cause trouble in Word. The reference looks ok in Reference Manager - but will apear different in Word.

Does that make sense to you and your problem with diacritical marks?

Kindly regards



I think the problem occurs when I save the two files (a data file and an index file) that comprise the RefMan database to a portable device (hard disk or thumb drive) in order to transfer the updated database from, say, my office desktop computer to my laptop computer.  Some letters with diacritical marks seem to be preserved and some are not.  Of course, if you do not catch those before you use Cite-While-You-Write or update a bibliography previously created in a manuscript, then the erroneus letters are included.

I seem to recall a previous disuccion in this forum on this topic–something about the difference between using ASCI and ANSI characters, but I’m not sure I remember that correctly.

I will try to do some experiments and report back on the results.

My current solution to that issue and others with journal formatting of references is to limit my use of the Cite-While-You Write feature to drafts of manuscripts.  That makes including references in a manuscript quick and easy.  But, just prior to submitting a manuscript to a journal, I convert the entire manuscript to original text, and go through each reference and correct any errors or formatting issues. I save a copy of the manuscript prior to converting back to original text, in case I have to add , delete, or reformat references after the original submission.

Speaking of experiments, I have found that sorting on any column, not just the author column, seems to relieve the slowness issue.  Sorting back to the original order does not reverse the solution.  But it is a shame that TR requires its customers to use work-arounds (if they can find them!) instead of correcting the programming.

Thank you for your assitance with this.

We’ve just upgraded a 70thousand citation database to RM12 (against my strong protests, as I had a week of attempting to upgrade 18 months ago, before giving up and going back to v11).

Hoped that v12.02 would have fixed the problems with slow entry/editing. I don’t think that spending 5-10 minutes re-sorting the database each time you open it is going to be very productive - and so far when I have been using the database (on our server) my colleague hasn’t been able to open it at all.

Any suggestions? Have you continued with RM12 or reverted to RM11?

Sarah Cage

Senior Scientific Information Officer

National Poisons Information Service (Birmingham UK)

I got the suggestion of splitting the database 18 months ago - a problem is that you can’t copy synonyms - decided that reverting to RM11 was a better solution. Of course our database is getting towards their upper limit (at least 3 thousand refs added per year, and we’re over 70 thousand). .

We are already doing the majority of editing in separate temporary databases, and only doing keywording in the main database, but we also use the database for bibliographic control of copies etc., so citations do need editing

Havn’t yet tried CWYW, as we’re still trying to sort out input.


I hope you don’t mind me commenting although my query is about slow text file importing rather than manually typing refs in.  I created the following entry on 03-02-2010 08:24 AM, and I have received no response from Technical Support.  Infact none of my entries have ever been responded to by them making this forum virtually useless. 

"Importing text files very slow in Version 12 in comparison to Version 11

I have imported exactly the same text files into network versions of V12 and V11 databases and have found V12 imports text files very slowly.


V12 -  Importing  text files of 1000 refs takes 17 minutes to import

V11 -  Importing exactly the same text files of 1000 refs takes 45 seconds to import"

I am unable to recommend that my team upgrades to V12 until this is fixed. 

Some of our  collaborators are starting to use V12 which is causing problems when using CWYW.  Which leads me to another of my entries on this forum that hasn’t been responded to.

"My problem is that we will be collaborating with V11 and V12 users and I have found that if a Word doc has both V11 and V12 refs from different databases inserted into it the bibliography won’t generate.  Depending which version of CWYW you use (I have both toolbars) the system cannot find the other database.  This happens even if both V11 and V12 databases are open.

This wasn’t a problem when collaborating with V9 and V11 users as the V11 CWYW toolbar was able to interact with the V9 database."

Is it time to start using Endnote?



We have been using Refman 12 network edition for more than 1 year now. We are approx. 20 users and have one database, that we are supposed to share. (15.000 records). We have experienced lots and lots of problems.

Now we give up and are going back to version 11.

I recommend, that you don’t install version 12 until Technical Support at Thomson Reuters solves the major problems in version 12.

I am frustrated, because we have spend a lot of time trying to get version 12 to work - but now we must realize that no work-arounds can make version 12 work.

Kind regards