4 ms·
Am I the only one who assumes the NY Times considered the futility of trying to stop every single 2000-bit hack and is rather trying to change the way intellige
by klenwell 16y ago
Am I the only one who assumes the NY Times considered the futility of trying to stop every single 2000-bit hack and is rather trying to change the way intelligent consumers of news and culture with disposable income approach the Times, and perhaps quality content generally, online.
I think they've done a fair job of trying to handhold users through the transition. I expect the success or failure of the paywall not to depend on the fact of the paywall itself (obviously), but on how easy and frictionless they'll make it to pay. If they can make it graceful enough, I will happily pay for content I know there are loopholes for accessing. If it's a clusterfuck, as with other DRM schemes, I'll either seek out the leaks or give up.
I'm hoping they've adopted something like this approach, a slightly more rigorous version of the NPR pledge model, and are relying on the goodwill and intelligence of their readership and not its stupidity.
- ultrasaurus 16y agoI assumed they want to see how people will react (they are in essence A/B testing Canada vs the USA). And that they're oping that people will end up using the service more or less as intended. It's an interesting experiment, I read more than 20 articles from NYTimes.com a month, but I don't need more than a few. If the blogosphere starts pointing to the BBC instead, I'll end up reading that. I should probably mention that it's my code (if we can call it that) we're talking about. The traffic is killing my blog now, but I a few hours ago made it out of curiosity if I could hack that (it was 20 minutes all in).