Hiding the Reply links might make sense if you can't stop yourself from commenting before reading, but hiding the comments themselves doesn't make sense. I find that HN (along with LWN) is one of the few sites with high-value comments, and I often want to read the first few to get context for the article.
I prefer using the comments/discussion themselves as an indicator to whether I'll find the article a worthwhile read in the first place - how often do users actually comment/discuss an article without reading it beforehand? What's the point in that?
Yup. Too many times have I slogged through six pages of text only to visit the comments and find, "This article was complete bullshit, and here's a well-reasoned argument citing excellent sources explaining why you just wasted your time reading it".
I prefer the inverse. Reading the comments from an educated audience first, then diving into the article with a more firm understanding of the landscape.
I like to read the article first, get a feel a of what it's about and then read comments to see further analysis. Reading article first let's me think about it before I get influenced (negative or positive) by reading the comments.
I agree, commenting without reading the article and only reading the comments is a sure-shot way of making a mistake.
To play the devil's advocate, if you read the comments before reading the article it's very easy to fall into the "hivemind" way of thinking without even noticing. If you read the article first you can formulate your own opinion individually and unbiasedly.
I think more often for better, at least for this forum. I read commentary first more probabalistically based on the hype-sensationalism character of the title. I dont subscribe to the hivemind theory, in almost every thread there is well-argued commentary falling on both sides of the issues presented.
I don't want to reward linkbait and hype. A pdf link on the efficient implementation of the radon transform or compiler theory is much more likely to get an unfiltered click from me.
Usually the title of an article is a poor indication of the article's content, structure and tone. By skimming the comments I can usually get some idea of these attributes and then decide if it's an article that interests me.
Also, the HN comments often load much faster than the page itself (full of crazy JavaScript, CSS, ads, useless graphics, etc).
One thing though, if the site's unreachable, often in the comments there's a mirror or a summary. If you can't turn the a into a:visited, you'd never get the link to show up if the site gets overloaded by the HN effect.
I don't have a preference about which one to read first, for ex. with articles like "I stopped using X and why you should stop too", I find more value in reading the comments first, because usually the top comment will be a counter-argument more worthy than the original article.
But with articles that teaches something new to me, for ex. "Scaling X to a million users", I prefer to read the article first.
The people who need it the most won't use it. But you are onto something.
The next step:
Get in touch with Google so that it's included into official Chrome in some way. It doesn't have to be installed by default - just a one time pop-up when a user first visits Hacker News. An offer to install it.
The next step +1:
Encourage people to expand the extension to other sites. Write it in a modular way, write documentation.
24 comments
[ 6.0 ms ] story [ 71.9 ms ] threadI agree, commenting without reading the article and only reading the comments is a sure-shot way of making a mistake.
You're not just getting landscape, you're getting the messy human battlefield of human opinion (sometimes for better, often for worse).
Personally i wish there were a plugin that would prevent voting until you've opened the link and spent at least a minute on page :P
I don't want to reward linkbait and hype. A pdf link on the efficient implementation of the radon transform or compiler theory is much more likely to get an unfiltered click from me.
Also, the HN comments often load much faster than the page itself (full of crazy JavaScript, CSS, ads, useless graphics, etc).
But with articles that teaches something new to me, for ex. "Scaling X to a million users", I prefer to read the article first.
The next step: Get in touch with Google so that it's included into official Chrome in some way. It doesn't have to be installed by default - just a one time pop-up when a user first visits Hacker News. An offer to install it.
The next step +1: Encourage people to expand the extension to other sites. Write it in a modular way, write documentation.