Tim Berners-Lee sitting at a computer desk, typing on his machine. Around him, 10,000 camera flashes flare, fireworks fire reds and blues into the sky, and the shades of a thousand costumes, filling his eyes. Then, on cue, he sees 50 metre high letters, brillant orange, spelling out “This is for everyone” in homage to his creation, to the Web. He smiles, feeling overwhelmed by the brilliance, and the sheer scale of the event, and the strangeness that his abstract creation has bought him here, to the start of the Olympics, a celebration of muscle and bone.

But, somewhere, admist the multiple hues of London 2012, a doubt gnaws at him. He knows that there is a flaw, a problem, something wrong with his creation. Surely, he realises for the first time, it would have been better, more dynamic, more capable of capturing all of human emotion, if it could do colour?

Of course, it did not happen like this. The web has been able to do colour images, inlaid with the text for a long time; I remember watching Mosaic download 10Mb full-colour image of a biscuit from China, line at a time in the early ’90s. Sadly, it would appear, news of this has not yet reached Oxford. My PhD student, Michael Bell has recently had his paper accepted at ECCB’12. The paper is looking at annotations in Uniprot, and particular the usage of words especially over time. As you might guess, it has got quite a few statistics in it, with lots of graphs. Publication in Bioinformatics is part of the deal from going to the conference.

Imagine my surprise, to get back the page proofs with this question attached.

Please advise whether the figures should be retained in colour. Please note, colour figures in print will be charged £350/$600 per figure. We offer colour figures online only free of charge, but this will be uploaded as supplementary data only.

— OUP

Now, the paper in this case is, to my knowledge, online only, so at least we are not going to get asked for £350 per figure (that would be around £3000 for the whole paper, depending on how “figure” is defined). But, it would appear, too much to expect the colour figures to be actually embedded in the PDF or, perhaps, even the web version of our paper.

All of this seems pretty pathetic really. I thought that I had plumbed the depths when PLoS One told me that they could not put ontology terms in my paper in a fixed width font, as Courier was not possible on the web; thankfully, we did manage this in the end. But the inability to do colour, embedded, for free in an online paper seems to hover dangerously close to incompetance.

There is only one solution to this. I publically state now, that in future the canonical version of my papers should be considered to be those produced directly by the authors — either myself, or one of my lab — without further intervention. These versions will be either published here, on my blog, or on arXiv. Versions published elsewhere will be solely for the purpose of REF and/or other forms of career advancement.