Re 2: running code that would take a lot of effort to integrate into an existing system, and no one else is interested in it, least of all the site admins, doesn’t trump anything.
Running new code elsewhere (not on lesswrong.com), and convincing everyone to switch over, would be a sufficient demonstration. I think writing posts to try to convince people or sound them out before implementing such a thing is a good approach.
A failed attempt is more decisive and teaches us more than a series of blog posts.
Mostly, I’m saying “don’t waste time on more posts”. If that means dropping the idea entirely and doing something more useful (perhaps posting on a more interesting topic), ok. If that means creating something and posting about how to use it, ok.
I think condensing the argument for it down from 20 posts to, say, 3 total, would be wise, but eliminating the ‘why should we even think about this’ phase and skipping to the ‘make it’ phase seems too much.
It was half-joking; I don’t actually know how serious you are or how much thought you’ve put into the recommended number of posts. What I meant to imply was that the mention of “20” as a starting point made you pick a higher number as a counter-offer than you would pick if you’d come at the question cleanly.
I meant, he listed 16 articles he wanted to write, and I didn’t remember the exact number but it was around 20, and I thought that was excessive. I figured that 3 would do.
So yeah, I was anchored on what he said he’d do, as a representation of what I was recommending he change from doing. Seems fair.
I think a long post sized explanation might be warranted, and should certainly be allowed, i.e. a comment-sized defense of the idea should not be expected. Even if such a presentation is possible, it would have to assume no inferential distance and so be less effective for at least some readers.
The inferential distance is significant; looking through this thread, the impression I get is that the people who have actually used NNTP in the past do not need to be convinced. Or rather, they need to be convinced only that it is possible, not that it is desirable. Dagon above, for example.
Well, I at least have used NNTP (and also skimmed the RFC as a refresher just now) and still need to be convinced that it’s better than the status quo.
Re 2: running code that would take a lot of effort to integrate into an existing system, and no one else is interested in it, least of all the site admins, doesn’t trump anything.
Running new code elsewhere (not on lesswrong.com), and convincing everyone to switch over, would be a sufficient demonstration. I think writing posts to try to convince people or sound them out before implementing such a thing is a good approach.
A failed attempt is more decisive and teaches us more than a series of blog posts.
Mostly, I’m saying “don’t waste time on more posts”. If that means dropping the idea entirely and doing something more useful (perhaps posting on a more interesting topic), ok. If that means creating something and posting about how to use it, ok.
I think condensing the argument for it down from 20 posts to, say, 3 total, would be wise, but eliminating the ‘why should we even think about this’ phase and skipping to the ‘make it’ phase seems too much.
You’ve been anchored.
Condensing it down to a comment in an existing thread rather than a top-level post would be wise.
What do you mean by ‘anchored’ in this context?
It was half-joking; I don’t actually know how serious you are or how much thought you’ve put into the recommended number of posts. What I meant to imply was that the mention of “20” as a starting point made you pick a higher number as a counter-offer than you would pick if you’d come at the question cleanly.
reference: Wikipedia Anchoring article.
I meant, he listed 16 articles he wanted to write, and I didn’t remember the exact number but it was around 20, and I thought that was excessive. I figured that 3 would do.
So yeah, I was anchored on what he said he’d do, as a representation of what I was recommending he change from doing. Seems fair.
I think a long post sized explanation might be warranted, and should certainly be allowed, i.e. a comment-sized defense of the idea should not be expected. Even if such a presentation is possible, it would have to assume no inferential distance and so be less effective for at least some readers.
The inferential distance is significant; looking through this thread, the impression I get is that the people who have actually used NNTP in the past do not need to be convinced. Or rather, they need to be convinced only that it is possible, not that it is desirable. Dagon above, for example.
I have used NNTP in the past and am not yet convinced.
Well, I at least have used NNTP (and also skimmed the RFC as a refresher just now) and still need to be convinced that it’s better than the status quo.
Fair enough. :-)
ETA: Also, it is relevant that there is an RFC for you to skim, and that it gets read by many, many people not necessarily associated with us.
Decisive, yes. Teaches more, only if anyone is paying attention.
Ok, maybe “teaches us no less” :)