There is search. As I say in my other, longer reply, there's SearchYC and there's the link at the bottom of every page.
But people won't search.
No one will bother. People will just submit the link and if it's a duplicate, who cares? It's too much work to see if anyone has submitted the same thing in the past few minutes, let alone the last few hours, or days, or weeks.
No, but if something interesting just pops up you simple don't have the time to search but the first few recent new-pages before someone else will post.
EDIT: Well yes, people don't search, it should be done automatically ;]
> ... if something interesting just pops up you simple
> don't have the time to search but the first few recent
> new-pages before someone else will post.
Why does that matter? If someone else posts it, good for them.
And it only takes about 30 seconds to run a SearchYC query, and about 30 seconds to scan the last two or three "newest" pages. One minute isn't long to wait.
And what if every time you submitted the system went and tried to deduce what to search for, presented you with its results, and made you check them before submitting. How would that feel.
> Maybe that would be an improvement. If it worked.
Exactly, that's a very good idea, and easily implemented.
Concerning search, which should be integrated nevertheless, I'm not used to do things again and again if I can automate it, it's just not my thing and if forced I'd rather not post again than doing it (which would maybe not be all that bad ;).
My point is that it's easy to implement for YC, simple filters and keyword search, c'mon, if I can do it, they can do it, I'll definitely create a prototype..
I use http://searchyc.com. It used to be the case that if it less than an hour old it was on "newest", and if it was more than an hour old it was on SearchYC.
Recently, however, it would appear that SearchYC has been lagging, and I wonder if it has been blocked by PG/YC. PG has said elsewhere that he's developing his own search engine for HN:
Perhaps he's deliberately blocking it to make his search seem better. Seems unlikely. Perhaps they scraped too aggressively and triggered a block. Seems more likely.
But the subject comes up a lot, and many people are dissatisfied. Lots of people therefore scrape, and that puts extra load on the server. You're proposing to do the same. It would be better if a distributed scraping system were devised, and that had a proper API. Alternatively, if PG provided a proper API, but why should he?
Let PG focus on getting good discussions, and someone else to provide a better filter or tagging system. More scraping, more load, but if someone does something "properly" then maybe the others can finally stop.
Interesting. Coincidentally I just wondered about how to not get in trouble with HN for providing a custom live feed (I decided to make my own rather than relying on unknown services or learning how to use searchyc).
My solution to that problem will be easy - I will hack Hacker News, sort of, by injecting client script which is fully capable of everything, and would not be considered a foreign service.
16 comments
[ 3.3 ms ] story [ 44.4 ms ] threadEDIT:
"An Open Letter to the Apple News Community on Steve Jobs’ Health, Professional Boundaries and Not Being Douchebags"
-- http://forkbombr.net/apple-news-open-letter/
EDIT: And this is a duplicate (http://news.ycombinator.com/item?id=2234776)
Any chance of getting a search function into HN?
I'd vote for an integrated search rather than relying on external services with unknown cache intervals.
I always wanted a better feed view of HN. I guess grouped by automatically derived tags and instant live-update could improve on that matter.
But people won't search.
No one will bother. People will just submit the link and if it's a duplicate, who cares? It's too much work to see if anyone has submitted the same thing in the past few minutes, let alone the last few hours, or days, or weeks.
EDIT: Well yes, people don't search, it should be done automatically ;]
And it only takes about 30 seconds to run a SearchYC query, and about 30 seconds to scan the last two or three "newest" pages. One minute isn't long to wait.
And what if every time you submitted the system went and tried to deduce what to search for, presented you with its results, and made you check them before submitting. How would that feel.
Maybe that would be an improvement. If it worked.
Concerning search, which should be integrated nevertheless, I'm not used to do things again and again if I can automate it, it's just not my thing and if forced I'd rather not post again than doing it (which would maybe not be all that bad ;).
My point is that it's easy to implement for YC, simple filters and keyword search, c'mon, if I can do it, they can do it, I'll definitely create a prototype..
http://news.ycombinator.com/item?id=2228842
http://news.ycombinator.com/item?id=2228854
http://news.ycombinator.com/item?id=2230042
http://news.ycombinator.com/item?id=2230361
http://news.ycombinator.com/item?id=2230497
http://news.ycombinator.com/item?id=2230702
http://news.ycombinator.com/item?id=2234776
Here's a lengthy discussion of the question:
http://news.ycombinator.com/item?id=1329334
In that PG says it's hard, but SearchYC seems to have got something that's good enough. Ironically enough, I found that using it:
http://searchyc.com/submissions/PG+search?sort=by_date
Recently, however, it would appear that SearchYC has been lagging, and I wonder if it has been blocked by PG/YC. PG has said elsewhere that he's developing his own search engine for HN:
http://news.ycombinator.com/item?id=1955703
Perhaps he's deliberately blocking it to make his search seem better. Seems unlikely. Perhaps they scraped too aggressively and triggered a block. Seems more likely.
But the subject comes up a lot, and many people are dissatisfied. Lots of people therefore scrape, and that puts extra load on the server. You're proposing to do the same. It would be better if a distributed scraping system were devised, and that had a proper API. Alternatively, if PG provided a proper API, but why should he?
Let PG focus on getting good discussions, and someone else to provide a better filter or tagging system. More scraping, more load, but if someone does something "properly" then maybe the others can finally stop.
My solution to that problem will be easy - I will hack Hacker News, sort of, by injecting client script which is fully capable of everything, and would not be considered a foreign service.
I might however delete it, if that's of any help.