3 ms·

Note: The dupe filter doesn't catch a missing www. https://news.ycombinator.com/item?id=14571426 https://news.ycombinator.com/item?id=14571426

by danjoc 9y ago

sillysaurus3 9y ago
As it shouldn't. It's a different URL. It's pretty annoying when dupe filters "catch" submissions with ?repost=1 appended to them, for example. It was either Reddit or HN that did that at one point. Sometimes there are legitimate reasons to resubmit. It's also a form of letting the community express themselves -- if a thread is repeatedly flagkilled, having some mechanism to resubmit it is a way of resisting the urge to cry censorship. It often gets a thread to the front page (albeit with cement boots). That's a bit tangential, but I'm just providing a few counterpoints to resist the urge to tighten the dupe filter. The manual method works pretty well.
- danjoc 9y ago
  >It's a different URL. It seems ripe for abuse. #walawalabingbang is also able to pass. https://news.ycombinator.com/item?id=14573297 https://news.ycombinator.com/item?id=14573297 That opens the door to unlimited resubmissions.
  - sillysaurus3 9y ago
    That's a rather inefficient way to get your account banned. ;) Jokes aside, I've been consistently impressed with how well manual methods have worked for HN. It was very unexpected that you can simply throw human resources at problems of scale and do so well. There is lots of smart software powering those humans, but the manual approach avoids false positives. Stripping a hashtag has also caused problems, by the way. I remember one submission in particular where the OP complained that the hashtag linked to a different entry in a blog, which was set up to make that the only way to link to it.
    - Houshalter 9y ago
      Deepmind's blog does that which is the most annoying thing. E.g.: https://deepmind.com/blog/#decoupled-neural-interfaces-using-synthetic-gradients https://deepmind.com/blog/#decoupled-neural-interfaces-using... It screws up url searches when I want to find if the link was posted before on HN or reddit, for instance.
- PhasmaFelis 9y ago
  > It was either Reddit or HN that did that at one point. HN, then, because Reddit loves reposts.

Note: The dupe filter doesn't catch a missing www. https://news.ycombinator.com/item?id=14571426 https://news.ycombinator.com/item?id=14571426

by danjoc 9y ago

Note: The dupe filter doesn't catch a missing www. https://news.ycombinator.com/item?id=14571426 https://news.ycombinator.com/item?id=14571426

sillysaurus3 9y ago

As it shouldn't. It's a different URL. It's pretty annoying when dupe filters "catch" submissions with ?repost=1 appended to them, for example. It was either Reddit or HN that did that at one point. Sometimes there are legitimate reasons to resubmit. It's also a form of letting the community express themselves -- if a thread is repeatedly flagkilled, having some mechanism to resubmit it is a way of resisting the urge to cry censorship. It often gets a thread to the front page (albeit with cement boots). That's a bit tangential, but I'm just providing a few counterpoints to resist the urge to tighten the dupe filter. The manual method works pretty well.

danjoc 9y ago
>It's a different URL. It seems ripe for abuse. #walawalabingbang is also able to pass. https://news.ycombinator.com/item?id=14573297 https://news.ycombinator.com/item?id=14573297 That opens the door to unlimited resubmissions.
- sillysaurus3 9y ago
  That's a rather inefficient way to get your account banned. ;) Jokes aside, I've been consistently impressed with how well manual methods have worked for HN. It was very unexpected that you can simply throw human resources at problems of scale and do so well. There is lots of smart software powering those humans, but the manual approach avoids false positives. Stripping a hashtag has also caused problems, by the way. I remember one submission in particular where the OP complained that the hashtag linked to a different entry in a blog, which was set up to make that the only way to link to it.
  - Houshalter 9y ago
    Deepmind's blog does that which is the most annoying thing. E.g.: https://deepmind.com/blog/#decoupled-neural-interfaces-using-synthetic-gradients https://deepmind.com/blog/#decoupled-neural-interfaces-using... It screws up url searches when I want to find if the link was posted before on HN or reddit, for instance.
PhasmaFelis 9y ago
> It was either Reddit or HN that did that at one point. HN, then, because Reddit loves reposts.