4 ms·

We need to segregate data from presentation. Sorry, but I don’t buy that. The very best presentation is almost always customised to the material it conveys. W

We need to segregate data from presentation. Sorry, but I don’t buy that. The very best presentation is almost always customised to the material it conveys. We try to separate content and presentation as a compromise, partly because it’s too labour-intensive to craft a bespoke presentation for every little job, and partly to separate skill sets because someone can be a great author but suck at design. But it’s still a compromise, and someone who both knows their material and knows design will often be able to create a more effective presentation of that material than they could if they were constrained to some pre-determined, standardised box of presentation tools. By the same principle, it’s not necessarily an advantage to let someone viewing that content customise the design arbitrarily, because it’s almost axiomatic that such a person isn’t yet an expert on the material. They don’t know which ideas are the important ones that the author wants to build on, or which relationships in the data are the most enlightening ones to explore. A well made presentation, created by or in collaboration with a subject matter expert, can direct the viewer through the material in a more considered way. To me, it makes no more sense to use a medium where the viewer has to make presentation choices than it does to send them a few spreadsheets of raw data and let them write their own content or plot their own charts.

hyperion2010 13y ago
Let me rephrase my original statement. We need to segregate data from a single format/layout. I do not think the analogy to spreadsheets applies here. There are (at least) thousands of perspectives on/presentations of scientific data that are completely meaningless. In that case the perspective and presentation matters (though the sharing the original data opens up the opportunity for others to find new perspectives as well, though at the moment they do need to 'plot their own charts') That said, once the author has chosen the perspective that perspective will be conveyed via some medium. At that point the data should be able to stand on its own. I used the word presentation, but 'layout' is probably closer to what I meant. If we take my use of presentation literally then we agree and this is why someone posting slides for a talk without also posting the talk is a terrible idea. With regard to expertise, I would argue that everyone is an expert on what format/layout they find easier to use (sort of like why we let people move their windows around in a window manager instead of forcing them all full screen, re: tablets are major offenders here). I'm not talking about letting people change the words here or put the legend for figure 1 with the data for figure 3.

hyperion2010 13y ago

Let me rephrase my original statement. We need to segregate data from a single format/layout. I do not think the analogy to spreadsheets applies here. There are (at least) thousands of perspectives on/presentations of scientific data that are completely meaningless. In that case the perspective and presentation matters (though the sharing the original data opens up the opportunity for others to find new perspectives as well, though at the moment they do need to 'plot their own charts') That said, once the author has chosen the perspective that perspective will be conveyed via some medium. At that point the data should be able to stand on its own. I used the word presentation, but 'layout' is probably closer to what I meant. If we take my use of presentation literally then we agree and this is why someone posting slides for a talk without also posting the talk is a terrible idea. With regard to expertise, I would argue that everyone is an expert on what format/layout they find easier to use (sort of like why we let people move their windows around in a window manager instead of forcing them all full screen, re: tablets are major offenders here). I'm not talking about letting people change the words here or put the legend for figure 1 with the data for figure 3.