NNTP worked really nicely for information distribution until two people called Canter and Siegel shit all over it. I've thought for some time that the primary virtues of NNTP were that it was text-based and client-centric.
You don't need a whole lot of stuff in a news article. Text, basic formatting, images, references, maybe a table or chart. Just because you can make elaborate interactive graphics doesn't mean that you should, most of the time. HTML is way overpowered for news and (more to the point) it makes it too easy for publishers to distribute crap to downstream users like Betty.
This brings me to the other side of the equation; the client. I do not just want an embedded browser. I've grown to hate the way that publishers, and their minions, the designers (^_^) have slapped CSS all over everything in order to brand every aspect of the web experience. I actually have preferences of my own about fonts, layout and so on, and I really miss being able to configure client software to display things the way I liked to look at them rather than the way some Jolt-chugging framework addict with a me-too manager wants to show it to me.
There are huge untapped opportunities in news delivery, of the sort I'd be willing to pay for. Why is there no way, for example, to pull up a globe that shows me where the hottest news stories are right now, or to use the same mapping layer for all geographic data contained in news stories, or build and share correlates, or set rules for source integrity to filter the obvious content farms that keep cropping up in my Google news feed, and so on?
News is ever-changing, but it's also consistent enough in lots of ways that it should super-enhanced by machine learning by now: the action and outcomes of individual sporting fixtures is highly unpredictable, but the context and seasonal progress of sporting fixtures is nearly invariant. Likewise, I don't know which country in the world will supply next week's most consequential events, but I do know that the list of countries and their associated geographic features vary very little over time. All stories involving weather take place in the context of seasonal variations over annual and decadal timeframes; by extension so do a vast number of economic news stories that involve commodities, and monthly and annual cycles are inherent to all sorts of economic and commercial activity. For a good many topics, there are only a few interesting facts worth parsing out of a 500-word story; much of the content is just contextual boilerplate. And yet when I go to Google Finance all I get is this mess of news headlines, uncorrelated data series, and market weathervanes. There is a great set of tools in Google Trendsbut there doesn't seem to be any systematic undertaking to integrate them with the feed of news.
My best guess is that most people are simply not interested enough in news to see a viable subscription-based model, or even an ad-based one based on carefully managed ads (like Google search results rather than the hideous eye candy that pollutes most news websites). I'll shut up now rather than waste an hour and thousands of words sketching specifications for how digital news ought to operate.
You don't need a whole lot of stuff in a news article. Text, basic formatting, images, references, maybe a table or chart. Just because you can make elaborate interactive graphics doesn't mean that you should, most of the time. HTML is way overpowered for news and (more to the point) it makes it too easy for publishers to distribute crap to downstream users like Betty.
This brings me to the other side of the equation; the client. I do not just want an embedded browser. I've grown to hate the way that publishers, and their minions, the designers (^_^) have slapped CSS all over everything in order to brand every aspect of the web experience. I actually have preferences of my own about fonts, layout and so on, and I really miss being able to configure client software to display things the way I liked to look at them rather than the way some Jolt-chugging framework addict with a me-too manager wants to show it to me.
There are huge untapped opportunities in news delivery, of the sort I'd be willing to pay for. Why is there no way, for example, to pull up a globe that shows me where the hottest news stories are right now, or to use the same mapping layer for all geographic data contained in news stories, or build and share correlates, or set rules for source integrity to filter the obvious content farms that keep cropping up in my Google news feed, and so on?
News is ever-changing, but it's also consistent enough in lots of ways that it should super-enhanced by machine learning by now: the action and outcomes of individual sporting fixtures is highly unpredictable, but the context and seasonal progress of sporting fixtures is nearly invariant. Likewise, I don't know which country in the world will supply next week's most consequential events, but I do know that the list of countries and their associated geographic features vary very little over time. All stories involving weather take place in the context of seasonal variations over annual and decadal timeframes; by extension so do a vast number of economic news stories that involve commodities, and monthly and annual cycles are inherent to all sorts of economic and commercial activity. For a good many topics, there are only a few interesting facts worth parsing out of a 500-word story; much of the content is just contextual boilerplate. And yet when I go to Google Finance all I get is this mess of news headlines, uncorrelated data series, and market weathervanes. There is a great set of tools in Google Trendsbut there doesn't seem to be any systematic undertaking to integrate them with the feed of news.
My best guess is that most people are simply not interested enough in news to see a viable subscription-based model, or even an ad-based one based on carefully managed ads (like Google search results rather than the hideous eye candy that pollutes most news websites). I'll shut up now rather than waste an hour and thousands of words sketching specifications for how digital news ought to operate.