4 ms·
So what's wrong with 1975 programming? (2008)
- pvg 10y agoAn HN evergreen - https://hn.algolia.com/?query=1975%20programming&sort=byPopularity&prefix&page=0&dateRange=all&type=story https://hn.algolia.com/?query=1975%20programming&sort=byPopu...
- bootload 10y agoHN does need a gold or classic posts tab so we automatically note this reposting. Appearing again and again hints to the fact the article is ^more interesting^ than usual.
- pvg 10y agoYes and there's a probably a whole other interesting post to be written about the critiques that have come up over the years, including some that claim the original is no longer representative of how varnish works today.
- Mathnerd314 10y agoStackoverflow has "community wiki" questions, maybe HN could have community wiki comments? OTOH my limited experience suggests that the majority of front-page articles will get reposted several times. Or someone could publish a "Best of HN" collection, then change all reposts to link to that.
- bootload 10y ago"Best of HN" collection" that @Mathnerd is a good idea. It would not be hard to do this remotely via a script.
- Mathnerd314 10y agoYeah. It's a little tricky though, because the top posts (https://hn.algolia.com/?query=&sort=byPopularity&prefix&page=0&dateRange=all&type=story https://hn.algolia.com/?query=&sort=byPopularity&prefix&page...) are all one-time events, and the # of people on HN has been increasing so upvotes aren't comparable across time. The criteria should be something like "reached the front page multiple times over several years".
- bootload 10y agowhat is the hn.angolia search term for the "past" of a story? i.e.: for "So what's wrong with 1975 programming?" it is https://hn.algolia.com/?query=So%20what's%20wrong%20with%201975%20programming%3F&sort=byDate&dateRange=all&type=story&storyText=false&prefix&page=0pe=story&storyText=false&prefix&page=0 https://hn.algolia.com/?query=So%20what's%20wrong%20with%201... Having this means you can measure how popular they are.
- rakoo 10y agoMost importantly, posts with that label should have links to all previous discussions
- bbcbasic 10y agoWas it all fixed in '76?
- nimchimpsky 10y agoyes.
- kelnos 10y agoSo, I get the frustration expressed here, but PH-K just comes off sounding like an arrogant prick by the end of this. I know I'm supposed to attack the argument and not the author, but it's very hard to read a piece like this when the author makes you feel like he's talking down to you the entire time. I assume this was posted because of the presence of antirez's 4-years-later rebuttal (well, for some workloads) that's also on the front page today (https://news.ycombinator.com/item?id=13226341 https://news.ycombinator.com/item?id=13226341). It's interesting to see the difference in writing style: antirez seems to just aim to educate, and reasonably points out how redis's needs and memory access patterns don't allow for good performance if it were to just rely on the OS's VM system. The article about Varnish, however, comes off as someone telling us how stupid we are for not knowing how hardware actually works.