19 May, 11 | by BMJ
My blog about the iPad last month generated some interest, and this update outlines some of the feedback we’ve had to date. The move to make the app free to BMA members has generated some very positive reviews on the iTunes app store. We’d already had some nice comments about the app’s technical functionality, but the pricing issue led to a fair number of 1* ratings. more…
1 Apr, 11 | by BMJ
A chunk of our working week is now devoted to preparing the latest BMJ iPad edition so its appearance in the iTunes app store coincides with the thud of the print issue as it lands on subscribers’ mats. I thought I’d use this blog to explain how our processes have changed to accommodate our new product, and to gauge interest in seeing the BMJ on other e-reader/mobile devices.
But first, some history. The BMJ iPad app launched in early 2011. We are the first general medical journal on Apple’s tablet computer, and our focus was to respond to a wish from some international subscribers to get a sense of what the print issue looks like each week.We’ve been delighted with the feedback to date, particularly with the interest shown by BMA members. When we developed the app we assumed (wrongly, as it turned out), that interest from UK doctors would be limited. After all, most get the BMJ print and online as a BMA membership benefit. So our focus was getting the app live, rather than investigating ways of offering it free to members. The BMA is now offering it for free. You can find out more by reading these FAQs.
I’ve talked a lot about print, but our focus throughout was to make the app genuinely interactive, and it is this which takes up our time each week.
The print issue goes to press on Tuesdays, but many of the pages (research, practice, etc) are ready by Monday. So we meet to discuss what archive links to add to the iPad pages. These are the kind of connections we tend to make:
On Wednesdays, after the print issue has gone to press, we select images for the iPad and ensure they are the right format and size for the device. We then insert all the links we identified on Monday via a content management system, and quality assure the finished product on a test server before getting the issue live by Friday. This usually entails ensuring that links work that images look OK. News, blogs, podcasts, and videos are updated via a live feed, so they bypass the above process, but an end users always sees the latest each time they log in via a web or wifi connection.
The whole process now runs pretty seamlessly, so I imagine we will very soon consider if we want the BMJ on other mobile/e-reader devices. The iPad project means we had to stall our discussions on iPhone apps, or ones for othe mobile platforms. How might we cut the content cake for a phone app? We could, for instance, launch separate ones for news, education, research, and comment. Or we could be more creative and combine content from across BMJ Group products, adopting a similar approach as we have for specialty portals (see my previous blog). Amazon’s Kindle often gets talked about. Perhaps that’s a logical next step for us. It would be great to get your feedback.
David Payne is editor, bmj.com
11 Mar, 11 | by BMJ
The bmj.com topics page is one of the site’s top 20 most visited pages, linking to 290 collections of clinical and non-clinical topics. Each one is powered by tagging tool called Semio, which trawls articles for keywords and then attaches that article to one or more collections. Visitors to the site can sign up for specialty-specific email alerts. So a psychiatrist, for example, can receive email alerts every time an article is posted about mental health.
Some of these clinical collection pages were revamped recently and now pull in not only BMJ articles but also relevant content from BMJ Group’s 39 specialist journals. There are also links to the latest jobs in that specialty (pulled from BMJ Careers), plus BMJ Learning modules and discussions from doc2doc, BMJ Group’s clinical community for doctors worldwide.
We’ve called these new pages “specialty portals,” and launched with oncology, diabetes, and psychiatry. Their design is slightly different from the standard bmj.com topic collection pages, and we were pleased to attract sponsorship for the diabetes portal, which is currently supported by an unrestricted grant from Novonordisk.
As the diabetes portal makes clear, the company has no influence on the editorial content displayed, but you do, by giving us your feedback on the portals we’ve launched so far, via this blog.
Semio is an automated tool and replaced the manual tagging we used to use each time an article was posted on bmj.com. Manual tagging was very time-consuming and quite subjective. We’re continually refining Semio to ensure the content displayed is relevant to diabetologists, psychiatrists, and oncologists (plus anyone with an interest in these three clinical areas). If we restrict Semio too much on the diabetes portal, for example, we risk excluding articles that perhaps mention diabetes alongside other chronic conditions. If we apply very loose tagging rules, we could end up with a list of headlines that fail even to mention diabetes and risk confusing visitors to this page. We plan to introduce some manual “over-rides” in the near future, and to involve practising clinicians as portal
editors. These can advise on which articles to exclude, perhaps write a blog on latest developments in their clinical specialty, and participate in clinical discussions on doc2doc.
The other issue is recency versus relevance. Websites that update regularly tend to attract more visitors. The dilemma this presents is how to ensure a very relevant paper that could be months or even years old still shows on the portal. To address this, we have five manually added “Top 5” articles from across BMJ Group. Again, give us your feedback on this selection. You may think other articles should be in there. Perhaps the diabetes podcast we produced last month should be there? Let us know.
Both the topic collections and specialty portals are bmj.com pages, and they will undergo further changes later this year when the site redesign goes live. We plan to colour code them, and to highlight them more prominently from relevant articles. So a BMJ research paper about cancer, for example, will have a colour coded link to the oncology portal showing at the top, plus any other relevant topic collection links.
Portals, of course, are just one way of clustering content online, and Semio is just one tool out there. Some sites suit the “handpicked” approach better. The portal for junior doctors is a good example of this. Each week my colleague Helen Macdonald (a GP trainee and BMJ assistant editor) looks at what’s been published across the Group and suggests some links for our junior doctor portal (although there is some automation also). A similar approach is used for the site we set up to highlight the NHS reforms in England, and the one we developed three years ago to mark the 60th
anniversary of the NHS.
Again, we’d be interested in your feedback. Should our portals include feeds to content from other sites, for instance? And although 290 collection pages sound like a lot, are we missing any?
David Payne, editor of bmj.com
12 Nov, 10 | by BMJ
Some 15 years ago, the BMJ was the first general medical journal to launch its own website. At the time, this was a paradigm shift, as Tony Delamothe wrote at the time:
“Users will now be able to scan the list of contents, read the editor’s choice, pick out from the page headed “This week in BMJ” what interests them most, and download the full structured abstracts of these papers. Some articles will be available in full text (for example, in this week’s journal all the articles on electronic developments). As more of the “hard” copy of the journal is produced electronically we hope to increase this proportion. Past issues of the journal will be available from an electronic archive, as will our advice to authors on the conduct and presentation of research. Details of other journals and books published by the BMJ Publishing Group (with ordering facilities) will also be provided.”
The next milestone in 1998 arrived in the shape of full-text articles online. As Richard Smith and Tony Delamothe wrote:
“We are restricting most of the BMJ’s homepage to text because images, sound, and video take a long time to download with currently available software. But there are tantalising possibilities for the future—for example, our recent fortnightly reviews on Parkinson’s disease could have included video clips of parkinsonian gait and tremor. The linkage functions of the web mean that access to a paper could provide simultaneous access to any correspondence and editorial comment it provoked. Linkages could provide the full text of references to papers, and the full text of references to references (and so on), so that readers need no longer take authors’ interpretations of earlier work on trust. Authors’ original data could also be made available through an electronic archive linked to the paper.”
However daring this seemed, both editors were still quietly cautious in their predictions for the future: “Predicting the future of electronic publishing is as risky as any other branch of futurology … future electronic developments will be closely informed by what our readers tell us they want and need.”
Much has happened to bmj.com in the intervening years. We introduced publishing ahead of print, continuous publication, and processing our articles via our webhost HighWire’s content processing platform, HighWire Express. Many features and technical innovations have been added to the site—such as folders, web extras, blogs, podcasts, videos, to name a few—all of which reflected new technical developments and the desire to provide users of bmj.com with the varied outputs they were expecting, and all of which contributed to making bmj.com the huge and varied beast it is today (For more information on the history of bmj.com, follow this link: http://resources.bmj.com/bmj/about-bmj/copy_of_history-of-bmj.com). But, to misquote a popular film release, technology never sleeps.
On 25 August 2010, between 5.30 pm and 6.30 pm GMT, a long anticipated moment had come: after months of preparation, bmj.com in its entirety had been migrated to a brand new web platform, affectionately known as H20. In the words of our webhosts, Stanford University’s HighWire Press, H20 is “an infrastructure that is flexible, permeable, and standards compliant throughout. Its architecture integrates with Web 2.0 and other emerging technologies and services and offers optimal speed and 24x7x365 systems monitoring and response … It can easily be built upon using Web 2.0 applications, feeds, widgets, and web services.”
H20 enables customisation of design and add-ons depending on publishers’ individual requirements, and in our up and coming redesign we shall explore those capabilities fully. In the meantime, bmj.com already looks cleaner and more contemporary after the migration and offers some standard features that come with the platform and keep users in context at all times, including, but not limited to easier ways to enlarge figures and thumbnails, expandable/collapsible author affiliations, related links, and other article enhancements, and the ability to view multimedia or video within the article.
For the BMJ editors and the H20 team, the August launch had been preceded by many months of comparing articles, with the new format appearing on a beta site. Many log sheets were completed through to iron out mistakes and glitches, and this work continued at full speed for several weeks after migration, to ensure that what users found was no different to what they were expecting. And although the migration process itself did not seem to take long, that evening the H20 team worked until midnight to ensure that everything was as it should be, logging problems as they presented themselves. Whether this was the way graphics were displayed or font sizes rendered, or whether it was missing titles with rapid responses—everything required meticulous checking and rectification. One of the problems with our articles was that, owing to the site’s evolution over the years (see above), they were all in different formats of HTML and XML, so a lot of conversion work by external suppliers was necessary, which, in turn, required thorough checking and probably threw up things that had happily lain dormant for many years. The journal archive, which dates back to 1840, exists as as pdfs and did not need to be converted. However, the quality checking continues—in-house and with the help of our users and authors,
The impact of the new web platform and its potential will become obvious over the coming years and the new features offered will become even more relevant as we are preparing for a redesign on the site. All in all, it was amazing to see how something so technically complex went through with comparatively minor mistakes.
Birte Twisselmann, web editor, bmj.com
Keep abreast of the technological developments being implemented on the BMJ journal websites.
