3 ms·
Yes, I've been following with interest the evolution of the treatment of items that some might regard as duplicates. I agree that there are items that deserve
by ColinWright 11y ago
Yes, I've been following with interest the evolution of the treatment of items that some might regard as duplicates. I agree that there are items that deserve a second chance, because the HN firehose pretty much guarantees that good items (by some definition) will be missed, and sink without trace. Leniency in the definition of "duplicate" is one way to go, as is the encouragement of reposts of items you think are good for HN and deserve a second chance.
But having multiple copies of the same story - HN defined duplicate or not - means that there is less space for other new stories. I realise I'm not saying anything you don't already know, and I'm not offering a different strategy, I'm just saying that I see you're doing things (whether I agree with them is irrelevant) and so my previous strategy of cross-linking duplicates (by my definition) has been discontinued.
You'll notice, by the way, that those specific three were submitted within minutes of each other. It all adds up to my conclusion that, in effect, I've stopped bothering. I recognise that you are working on this, and I appreciate your efforts.
- nkurz 11y agoYou've obviously thought about this lots. Is there a better way that you envision? My instinct is that it's best solved with more explicit curation of submissions, but I can't yet picture a full working system.
- ColinWright 11y agoI think the primary unit should be "the story". I think that submissions of the same story should come under an umbrella meta-submission. This will allow different points of view, without splitting the discussion. New submissions of the same story can be identified as "identical" or "related", and either dropped or merged. All this can be curated by people who simply happen to notice the duplication, and most users need not do anything. The addition of a "meta-submission" might be too much of a step for the software, though. With some thought it should be possible to integrate the idea into the existing structure, but I know nothing about the internals, so I can't say for sure.
- dang 11y agoThat's actually pretty close to something we've been working on. I don't know when it will be ready though. Catching not only identical stories but also related ones is important, because often the same story appears in a dozen or even several dozen guises.