Open jeremysimmons opened 9 years ago
It's just the md5 kind if md5... Which library / implementation? I'm not sure. But the miracalesalad link you sent works the same as ours for this example: http://inlightapp.com/webArticle/article/53ca276ee4b0356b396e72ed which miracalesalad calculates as: ddd354c67d0b69c3c3ad6175c53f74bf2 And our link: https://delicious.com/link/dd354c67d0b69c3c3ad6175c53f74bf2
However it's not always that simple. If there are anchors (#placeholder) we strip them all off, so we don't end up storing unlimited variations of the same page forever. That does break some single page apps, so we're rethinking our strategy. There might be some other cases than anchors that we strip off as well.
Is that explanation good enough?
Hi Dave - Thanks for the quick reply. Your explanation doesn't help me calculate how the hash is calculated by the api. If there are special rules for which part of the initial url to use, I would like to know those.
Neither my hash or the one I got from the delicioius api work for the link format you sent: https://delicious.com/link/:md5 hash for the url http://www.codemag.com/article/0611041.
Most of the links for https://delicious.com/link/:md5 just have a message Delicious is processing this link, please try again later.
That is correct. I don't think it's the md5 implementation, more likely some business logic on our side. I'm not sure what that could entail, but I suspect we have matched a URL to another version of the page. I'll try to check it out.
Hi Dave - Any news from the inside on how the URL hash is being calculated? I'd very much like to be able to compute the hash locally. The api /v1/posts/get?hashes={md5} is another reason. Thank you
What is the MD5 implementation you're using to hash the href field from a post? I cannot understand why my hashes are not matching your data.
I have cross-checked my computed value with a 3rd party which matches my computed hash. http://www.miraclesalad.com/webtools/md5.php