Group for discussion of CiteULike. Non-members can start new forum topics. There are specific forums for bug reports and feature requests. http://www.citeulike.org/groupfunc/3124 CiteULike.org en-gb Copyright © 2004-2007 Oversity Limited http://www.citeulike.org/groupforum/1709 <p> really cool...thanks a lot..it will be great if you update your faq with these information. </p> 2009-11-14T18:44:13+00:00 http://www.citeulike.org/groupforum/1709 <blockquote> So, I would like to have a view in which I can choose the papers I want and export it to bibtex format and use that bibtex file for my research paper. </blockquote> <p> The best way to do this is with tags. There are also &quot;private&quot; tags which others cannot see, such as &quot;*my_current_paper&quot; </p> <blockquote> For example, make the papers private or change the priority or create a bibtex file. </blockquote> <p> On the library view you can select multiple papers at once to tag them. You can then export that tag as above. </p> <blockquote> I am not sure how much storage is supported for each user. </blockquote> <p> No limits, but you can add the link to a PDF using [Edit Links]. </p> 2009-11-14T09:52:33+00:00 http://www.citeulike.org/groupforum/1709 <p> First off, citeulike is a very good idea and I believe this is the way papers have to be organized. Thank you very much. I have been using citeulike for about 10 months on and off. I think its due to lack of some features important to me that still keep me from using it fully. I just discovered I can request features. So, here are my requests. </p> <p> 1. The number of papers I read is much more than the number of papers I would cite in my research paper. So, I would like to have a view in which I can choose the papers I want and export it to bibtex format and use that bibtex file for my research paper. </p> <p> 2. It will be nice if we can modify multiple papers properties at a time. Similar to a webmail interface, you can list the papers and allow the user to select the bunch of papers and modify at the same time. For example, make the papers private or change the priority or create a bibtex file. </p> <p> 3. I am not sure how much storage is supported for each user. If there are limits, can we just store the links to pdf from personal website instead of uploading the pdfs. </p> 2009-11-13T22:46:29+00:00 http://www.citeulike.org/groupforum/1706 <p> I tried harder. Slight downside it that it now takes the later date &quot;1 Oct 1992&quot;. In my opinion it's pretty dumb having a publication date as a range, anyway. </p> 2009-11-13T15:57:34+00:00 http://www.citeulike.org/groupforum/1706 <p> Argh. It's because they put the dates in hard-to-parse formats like </p> <pre> Publication Date: 29 Sep-1 Oct 1992 </pre> <p> We try that first (and fail), the resort to the next best &quot;Current Version Published: 2002-08-06&quot; </p> <p> I will try harder. </p> 2009-11-13T15:51:13+00:00 http://www.citeulike.org/groupforum/1706 <p> Thanks, but it still doesn't work for me. Example <a rel="nofollow" href="http://ieeexplore.ieee.org/xpls/abs_all.jsp?isnumber=6179&amp;arnumber=240727&amp;count=85&amp;index=1">http://ieeexplore.ieee.org/xpls/abs_all.jsp?isnumber=6179&amp;arnumber=240727&amp;count=85&amp;index=1</a> </p> <p> Original publication on 1992, citeulike takes 2002. There is no reference to IEEE Xplore what soever (ok, only to IEEE), so I don't understand why the conference is still ICUPC, but the year is 2002. for me it looks like the conference in the year 2002. Same for journals. </p> 2009-11-13T15:38:12+00:00 http://www.citeulike.org/groupforum/1708 <p> just see a lot of babbley text like its not been uploaded in the right format - single refs are fine. I shall email support, many thanks Anna </p> 2009-11-13T15:06:32+00:00 http://www.citeulike.org/groupforum/1704 <p> &gt;&quot;same&quot; article appears more than once within CiteULike... Understood. It's fine then. Thanks! </p> 2009-11-13T14:49:07+00:00 http://www.citeulike.org/groupforum/1708 <p> See anything else? </p> <br/> <p> How big is your Endnote library? If it's massive (&gt; few 1000), try splitting into smaller chunks. </p> <br/> <p> Try just a single record. If THAT doesn't work, please email that to us: support09@citeulike.org </p> 2009-11-13T13:52:09+00:00 http://www.citeulike.org/groupforum/1708 <p> That was my original plan (nice to know I'm on the right track) however, when I try to import it, it does not appear in my CUL library...any ideas? Thanks Anna </p> 2009-11-13T13:44:51+00:00 http://www.citeulike.org/groupforum/1708 <p> Anna </p> <p> If you export your library from Endnote as either a BibTeX or a RIS file you can import it into citeulike using the 'Import' option on your &quot;MyCiteULike&quot; menu. </p> 2009-11-13T13:08:57+00:00 http://www.citeulike.org/groupforum/1708 <p> I already have an Endnote library - how can I upload it onto CUL? </p> <p> Many thanks Anna </p> 2009-11-13T12:53:19+00:00 http://www.citeulike.org/groupforum/1702 <p> The survey is now complete; we have gathered enough data. Thanks to all the people who participated. Sorry again for the lack of clarity upfront. </p> 2009-11-13T10:44:11+00:00 http://www.citeulike.org/groupforum/1706 <p> I've changed a few things. It's not clear to me that this is always the right thing to do, though. </p> 2009-11-13T09:49:58+00:00 http://www.citeulike.org/groupforum/1675 <p> Hi mmwoodman, </p> <br/> <p> We're still a couple weeks away from having a decent alpha-quality release, but we've got a Google Group to act as an announcement mailing list of sorts: <a rel="nofollow" href="http://groups.google.com/group/paperpile-alpha">http://groups.google.com/group/paperpile-alpha</a> . Sign up there if you want to get updates on the project and try out an early release when it's ready. </p> 2009-11-13T09:49:54+00:00 http://www.citeulike.org/groupforum/1706 <p> Can you post a specific example, please? </p> 2009-11-13T09:30:12+00:00 http://www.citeulike.org/groupforum/1706 <p> When trying to add an article from the IEEE Xplore the wrong publication date is taken. I see there are two date fields in IEEE Xplore format: Publication Date and Current Version Published. The first one corresponds to conference dates or journal issue date, the second to the appearance in IEEE Xplore. So if the paper is published in a journal in 1990, but in IEEE Xplore in 2002, the year 2002 is taken as the publication date in citeulike. This brings only confusion, since all other data (journal number, issues, pages) are correct for the journal. </p> <p> Please, check. </p> 2009-11-13T09:26:23+00:00 http://www.citeulike.org/groupforum/1704 <blockquote> if somehow the id's of some rejections are not being saved </blockquote> <p> They all are (or should be!) </p> 2009-11-12T15:48:56+00:00 http://www.citeulike.org/groupforum/1704 <p> It's going to be difficult to do this. How is a user to be certain whether an unwanted article is repeated? (however convincing their recollection might be). Note that (a) they weren't interested and so hit &quot;reject&quot;, along with perhaps 10-100 other not-very-interesting articles in the same session, and (b) they don't have access to a list of what they've rejected to check. </p> <br/> <p> Why not add a &quot;review all rejected articles&quot; option (after all, you do keep a list, or you couldn't reasonably claim to be excluding them). Then users haunted by recommendation deja-vu could check for ourselves. </p> <br/> <p> Also, such an option would mean that users might recover those rejected recommendations, about which they are now having second thoughts. </p> <br/> <p> Edit: mind you, if somehow the id's of some rejections are not being saved, this wont help. :-/ </p> 2009-11-12T14:51:24+00:00 http://www.citeulike.org/groupforum/1702 <p> What we are planning to do is to replace some of the Google ads with ones supplied by an ad network that is focused on &quot;Science&quot;. The skew in the questions is a function of the proposed network's bias I think, but they know more about how their advertising is sold than us. Computer Science is the leading category in the responses so far, by the way. </p> 2009-11-12T14:22:35+00:00 http://www.citeulike.org/groupforum/1704 <p> It certainly shouldn't be happening and if it is, it's a bug. Next time you see it happen, please post the (CiteULike) URL of the latter (I assume you won't have the URL of the one already rejected). </p> <br/> <p> Occasionally the &quot;same&quot; article appears more than once within CiteULike. Our definition if &quot;same&quot; is if it came from the same URL (or has a similiar unique identifier like DOI, pubmedId, etc.) We don't use, for example, the title or similar metadata. Sometimes an article might get posted from two different sources without us detecting the common identifier. </p> 2009-11-12T14:14:08+00:00 http://www.citeulike.org/groupforum/1704 <p> I'd like to suggest keeping a list of papers a user rejects from its &quot;My Top Recommendations&quot; list and not adding them again in subsequent compilations of the list. Otherwise, there may be a ever increasing bias, which will make popular papers even more popular, if a user adds a paper to his library to avoid its appearance in &quot;top recommendations&quot; list again and again. In the long run this may significantly reduce the usefulness of the recommendations list, as &quot;top&quot; papers will start creeping into everyone's list, pushing out more interesting &quot;long tail&quot; publications. </p> 2009-11-12T14:00:00+00:00 http://www.citeulike.org/groupforum/1702 <blockquote> The reason we are asking everyone to take part in the survey is that potential advertisers need to know the demographics of the audience </blockquote> <p> Fair enough, but why are the questions so skewed towards the life sciences? Many of my responses were blank, which provides no useful demographic information. Or are your potential advertisers only in the life sciences? </p> 2009-11-12T13:56:03+00:00 http://www.citeulike.org/groupforum/1702 <p> We've changed the wording to make the purpose of the survey clearer. The reason we are asking everyone to take part in the survey is that potential advertisers need to know the demographics of the audience, and it makes a surprisingly significant difference to how much citeulike can charge. </p> <p> Ironically, the words &quot;help support&quot; were chosen as a way to explain why we are doing this. I'm responsible for that wording, so I apologize to anyone who felt mislead. </p> 2009-11-11T13:29:48+00:00 http://www.citeulike.org/groupforum/1701 <p> the gdocs data api can store most any data, but what would be the advantage? a good place to get started might be pulling any RSS feeds into a spreadsheet with importfeed() or importxml() </p> 2009-11-11T10:48:44+00:00 http://www.citeulike.org/groupforum/1702 <p> Agreed. I was looking forward to writing some positive comments about citeulike (basically about the ludicrously quick tech-support) but the marketing aspect only became clear as I was progressing through the survey. </p> 2009-11-11T09:49:47+00:00 http://www.citeulike.org/groupforum/1702 <p> I don't mind ads or surveys and I greatly appreciate the excellent, free service. However, I think it could have been made clearer at the outset that this was a marketing survey and nothing to do with the CiteULike user experience. It felt a little underhand. </p> 2009-11-11T00:07:32+00:00 http://www.citeulike.org/groupforum/1703 <p> I think I may have asked this before - if so, apologies for duplication. </p> <br/> <p> Would it be possible to add a journal to one's watchlist? I can add tags, I can add users (and I've just found I can add authors), but I can't add journals. What do you reckon? </p> <br/> <p> Andrew </p> 2009-11-10T21:52:05+00:00 http://www.citeulike.org/groupforum/1702 <p> we forgive you (and I actually happen to like the huge banner on the left ;) ) </p> 2009-11-10T19:54:11+00:00 http://www.citeulike.org/groupforum/1702 <p> It's very tricky to get the balance right. We don't &quot;like&quot; ads any more than you - a necessary evil. But we felt maybe we could be forgiven for 1 survey after 4 years of free.... </p> 2009-11-10T19:32:39+00:00 http://www.citeulike.org/groupforum/1702 <p> ok, thegoose thanks for the extra edits and explanations - this makes sense, it just was not tremendously clear what you meant by &quot;help&quot; or &quot;support&quot;. </p> 2009-11-10T19:25:15+00:00 http://www.citeulike.org/groupforum/1702 <p> I did the survey too and I had the same feeling of being taken in. If this is supposed to be a serious survey about CiteULike demographics I wonder why there was not a single mention of the social sciences (?!) and only questions about lab equipment and DNA sequencing. For a moment I was tempted to conclude that CiteULike was selling its user base to some third party to conduct market research of no relevance whatsoever to reference management. Please prove me wrong and let us know about the results of the survey ;) </p> 2009-11-10T19:19:29+00:00 http://www.citeulike.org/groupforum/1702 <p> It's mostly about demographics - what percentage of our users are X. If we know this we can focus more effectively. You will have noticed that we have started showing ads more prominently. This is an unfortunate fact of life as we need to get income (for this free service) from somewhere. Advertisers pay more - this is the &quot;help CiteULike&quot; bit - when they know the target audience. And the ads may annoy you slightly less if they're somehow relevant. </p> <p> Thanks for doing the survey, BTW. </p> 2009-11-10T18:50:49+00:00 http://www.citeulike.org/groupforum/1702 <p> I have done the survey. It's all about whether you have purchasing power in your lab and how much you have got. Can't figure out how any of that information would help the guys at CiteULike. </p> 2009-11-10T18:18:47+00:00 http://www.citeulike.org/groupforum/1701 <p> There's always _edit_ to hide your shame :-) </p> 2009-11-10T17:13:57+00:00 http://www.citeulike.org/groupforum/1701 <p> oops apologies, it was a shameless copy and paste from the mendeley blog where I posted the same question. Let me try again: </p> <p> Has anyone thought of writing a plugin to integrate CiteULike with Google Docs via their data API? :) </p> 2009-11-10T17:12:44+00:00 http://www.citeulike.org/groupforum/1701 <p> Ask Mendeley? :-) </p> 2009-11-10T15:45:59+00:00 http://www.citeulike.org/groupforum/1701 <p> Has anyone thought of writing a plugin to integrate Mendeley with Google Docs via their data API? Check out this project: <a rel="nofollow" href="http://code.google.com/p/gdbe/">http://code.google.com/p/gdbe/</a> </p> 2009-11-10T15:44:27+00:00 http://www.citeulike.org/groupforum/1700 <p> On each article page, there is a link back to the original under: &quot;View FullText article&quot;. </p> 2009-11-10T11:21:42+00:00 http://www.citeulike.org/groupforum/1675 <p> I spent all day yesterday uploading ~100 PDFs the manual way: </p> <p> 1. open PDF 2. Search Pubmed with title 3. Get pubmed id 4. paste into CUL 5. upload PDF </p> <p> So I would be happy to help automate this process, or test something (I use Linux). Can you give some more details? </p> 2009-11-10T10:36:42+00:00 http://www.citeulike.org/groupforum/1700 <p> If you have access to the article's PDF, you can upload it and read that. We don't have access for the majority of sites (they are subscription based) </p> 2009-11-10T08:20:09+00:00 http://www.citeulike.org/groupforum/1700 <p> A list of articles is useless to me? I thought I would be able to store them here and be able to read them from here? If this just serves as a list - it's not very useful. </p> 2009-11-10T04:41:19+00:00 http://www.citeulike.org/groupforum/1675 <p> Interesting discussion here. This really gets at the core problem of connecting old folders of PDFs to nicely structured metadata. I'm convinced that the way to go is through DOIs and title/author extraction, which *can* work reasonably well if you are smart about it. </p> <br/> <p> We've got something in the pipeline that should do pretty well in this arena (open-source Mendeley-Desktop substitute). Gonna be some time before it's done, though! </p> <br/> <p> Contact me directly if you run Linux and might be interested in alpha testing... </p> <br/> <p> greg </p> 2009-11-09T15:05:31+00:00 http://www.citeulike.org/groupforum/1695 <p> Well, now it appears <i>underneath</i> the CiteULike logo, which is equally weird but also not right. </p> <p> I feel for you in browser compatibility hell... </p> 2009-11-06T13:21:33+00:00 http://www.citeulike.org/groupforum/1695 <p> I've tweaked some stuff but I haven't got Camino to test. </p> 2009-11-06T11:57:18+00:00 http://www.citeulike.org/groupforum/1695 <p> Why does the google add sit on <i>top</i> of the citeulike logo? That can't be right. (Camino on OS X) </p> 2009-11-06T02:01:57+00:00 http://www.citeulike.org/groupforum/1697 <p> I've added a plugin for Nature Precedings so you can post directly now. </p> 2009-11-05T16:20:29+00:00 http://www.citeulike.org/groupforum/1697 <p> There is a DOI hidden in the headers: 10.1038/npre.2009.3496.1 </p> 2009-11-05T15:46:59+00:00 http://www.citeulike.org/groupforum/1697 <p> It's an hdl, not a doi ... if you read closely you see </p> <p> hdl:10101/npre.2009.3496.1 </p> <p> and NOT: </p> <p> doi: 10.blahblahblah </p> <p> I think doi's are a subset of hdls, at least it seems so if you read </p> <p> <a rel="nofollow" href="http://www.handle.net/factsheet.html">http://www.handle.net/factsheet.html</a> </p> <br/> <br/> <p> EDIT: although the hdl/url given on the page, i.e. </p> <p> <a rel="nofollow" href="http://hdl.handle.net/10101/npre.2009.3496.1">http://hdl.handle.net/10101/npre.2009.3496.1</a> </p> <p> ... does not seem to work </p> 2009-11-05T14:29:41+00:00 http://www.citeulike.org/groupforum/1698 <p> Works for me. Are you sure you are using the &quot;advanced&quot; bookmarklet and not the classic one? The &quot;select&quot; feature only works with the former. </p> <br/> <p> If you are - tell me which browser you're using and I'll look into it. </p> <br/> <p> In any case, you can copy that URL into the form on MyCiteULike -&gt; Post URL </p> 2009-11-05T14:18:44+00:00 http://www.citeulike.org/groupforum/1698 <p> On &quot;http://www.ncbi.nlm.nih.gov/pmc/articles/mid/NIHMS19407/&quot; I select &quot;doi: 10.1021/pr050034b&quot; and use the bookmarklet, resulting in: I couldn't post the page you were looking at to the site because: </p> <p> * It looks like you're posting a page from PubMed, but it's not an article. Are you posting the results of a search, or maybe some other page, by mistake? Use PubMed to choose the article you're interested in, and then try posting that. </p> <p> It's like it doesn't recognize that I selected the doi. </p> 2009-11-05T14:10:27+00:00 http://www.citeulike.org/groupforum/1647 <p> The &quot;quality&quot; of the recommendations is driven by what other people have in their libraries. It just so happens that the disciplines most strongly represented are the in Life Sciences (biology, medicine, etc) and there's naturally a strong interest in web 2.0 stuff. </p> <br/> <p> Imagine a system based on real-life bookcases. Assume that most people have a dictionary on their shelf, irrespective of there other interests. I you don't have one, its seems reasonable that a recommendation system would suggest a dictionary to you based on the fact that almost everyone else has one. </p> <br/> <p> The good news is that we're moving forward with more advanced algorithms (see the Blog) and hopefully the next iteration will turn up things of more relevance. </p> 2009-11-05T14:02:17+00:00 http://www.citeulike.org/groupforum/1697 <p> That's not a DOI - DOIs are always 10.nnnn/xxxxxxxxxxxx </p> <br/> <p> I couldn't find a DOI for it - I'm not surprised because it's a preprint or similar. </p> <br/> <p> There is some quite good metadata in the HTML headers, though. I'll see what I can do. </p> 2009-11-05T13:44:22+00:00 http://www.citeulike.org/groupforum/1647 <p> The recommendation system is a great idea, but looking at mine I wonder why other CiteULike users in my research field have so many more relevant articles in their recommendations list than I do. Although my library contains mostly articles in geoscience, geophysics, or seismology, the majority of the papers in my recommendations list (http://www.citeulike.org/profile/nilsma/recommendations) is not relevant to these topics (10-15 relevant articles). Instead it contains a lot of articles in the fields of biology/medicine or e.g. on &quot;Web 2.0&quot; applications, and deleting all irrelevant articles does not help to improve my recommendations list. The recommendations list for my top neighbour (http://www.citeulike.org/profile/stevenatnorsardotno/recommendations) appears to be much more relevant and contains articles that should probably also appear in my list. </p> <p> Could the poor performance of the recommendation system on my library be related to my &quot;general&quot; tags such as &quot;review&quot; or &quot;book&quot;? Or are there other factors than article metadata influencing the recommendations? </p> 2009-11-05T11:02:16+00:00 http://www.citeulike.org/groupforum/1697 <p> I can't seem to post to citeulike from nature precedings. </p> <p> example, <a rel="nofollow" href="http://precedings.nature.com/documents/3496/version/1">http://precedings.nature.com/documents/3496/version/1</a> </p> <p> When posted by its DOI:10101/npre.2009.3496.1 to <a rel="nofollow" href="http://www.citeulike.org/post_url.adp">http://www.citeulike.org/post_url.adp</a> </p> <p> gives the citeulike message: &quot;Sorry. Can't do it. I couldn't post the page you were looking at to the site because: That DOI does not appear to be a known type.&quot; </p> 2009-11-05T10:13:23+00:00 http://www.citeulike.org/groupforum/1692 <blockquote> But I tried this just two weeks ago, resulting in article 5982750. I'm not sure how you obtained 6047498; I can't see the bibtex you imported, </blockquote> <p> I just cut and pasted the bibtex exactly as you supplied into the &quot;Import&quot; form. All bibtex imports get a unique article_id. </p> <blockquote> It appears, however, that you changed Davis's name from &quot;John S. Davis, II&quot; to &quot;John S. Davis&quot; </blockquote> <p> We don't support any suffixes (Jr, Sr, II, etc) :-( </p> <blockquote> I'm not sure what you mean by add hyphens; do you mean you hand-patched the result in the database? </blockquote> <p> I just clicked on [Edit] and added a few hyphens here and there. If you [Copy] my article into your lib, then [Edit] that, you'll see the raw data. </p> <blockquote> rewrote many of the names in last,first notation. </blockquote> <p> No, just the one mentioned above. The BibTeX syntax as in &quot;Khaled {El Emam}&quot; doesn't actually imply (as I understand it) that last name = &quot;{El Emam}&quot;. &quot;{}&quot; are just groupings, cf. &quot;{David, II}&quot; </p> <br/> <p> If there WERE a clearly defined AND standard then we might be able to implement it. Your example above seems to have a consistent grammar (Last name = last word or {...} etc) but I could show you lots of counter-examples. We import metadata from many journals and those &quot;are all over the place&quot;. In fact, in my experience, most tend to use &quot;lastnames, firstnames&quot; as that's mostly unambiguous. (And then some mess this up with extra &quot;PhD&quot;, &quot;MA&quot; etc. in a way that doesn't even parse!) </p> 2009-11-04T13:56:29+00:00 http://www.citeulike.org/groupforum/1692 <p> Yes, CiteULike has gotten better at parsing names since I tried it six months ago. Back then, the above test actually caused CUL to spit out syntactically incorrect bibtex. </p> <p> But I tried this just two weeks ago, resulting in article 5982750. I'm not sure how you obtained 6047498; I can't see the bibtex you imported, so I'm not sure what changes you made to the source bibtex before import. It appears, however, that you changed Davis's name from &quot;John S. Davis, II&quot; to &quot;John S. Davis&quot; (his father) and rewrote many of the names in last,first notation. I'm not sure what you mean by add hyphens; do you mean you hand-patched the result in the database? </p> <p> I understand the challenges of name parsing, but CiteULike should correctly parse correctly-coded bibtex. The code for bibtex is open-source so one could study that code to understand how bibtex handles its syntax. </p> <p> I have a bibtex bibliography of 700 references I'd like to import and really don't want to spend hours verifying the import (and later, export) of every reference. Believe me, I've already spent hours cleaning and verifying the bibtex (and its output when render by bibtex and latex) and don't want to fix bugs inserted by CUL. </p> <p> So, I'm appealing for a little help. </p> 2009-11-04T11:55:10+00:00 http://www.citeulike.org/groupforum/962 <p> OK, it was a little bit of a surprise for me too, that it worked without tweaking. </p> <p> Anyway, I've now updated the script so that it removes the ads, as per <a rel="nofollow" href="http://www.citeulike.org/groupforum/1695">this thread</a> </p> 2009-11-04T07:59:51+00:00 http://www.citeulike.org/groupforum/1695 <p> Heh, <a rel="nofollow" href="http://userscripts.org/scripts/show/56486">done</a> . <img src="http://s3.amazonaws.com/uso_ss/3805/large.gif?1257410453"/> </p> <p> But I guess those ads are helping to pay the staff's wages. It is an ethical dilemma. </p> <br/> <p> Still, it looks like my greasemonkey script is only used by a couple of dozen users or so, so the revenue impact should be very small. (hah, not yet 8am, no coffee yet, but my powers of post-hoc rationalisation are as devious as ever) </p> <br/> <p> So it ain't my fault if it all goes titsupnogood </p> 2009-11-04T07:47:09+00:00 http://www.citeulike.org/groupforum/1695 <p> Another GreaseMonkey session coming on? </p> 2009-11-03T17:44:59+00:00 http://www.citeulike.org/groupforum/1695 <p> The ads on the left bar used to be 120 pixels wide, now they are 160. If you use firefox, you can use adblockplus to get rid of them. It can't change the layout, however. </p> 2009-11-03T17:27:52+00:00 http://www.citeulike.org/groupforum/1695 <p> I understand why you had to do it. But it's foul. It really messes the ergonomics, and the ads take a huge chunk of the desktop. </p> 2009-11-03T16:57:39+00:00