Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: An idea that you could not implement?

news.ycombinator.com

Post reply on HN
Print view View original

51–60 of 125 posts

Re: Ask HN: An idea that you could not implement?

#51

Post by Animats »

bootload wrote: post #33

good question, going through these ideas I keep in mind: "The winds of change originate in the unconscious minds of domain experts. If you're sufficiently expert in a field, any weird idea or apparently irrelevant question that occurs to you is ipso facto worth exploring." [0] 1/ So the idea I haven't done is build a hacker search engine. I'm not sure what this would look like, but I'd bet being able to search throug…

being able to search through source code at high/low levels might be interesting.

It was called "Google Code Search". 2005-2013. [1]

[1] https://en.wikipedia.org/wiki/Google_Code_Search

Re: Ask HN: An idea that you could not implement?

#53

Post by developer2 »

TAForObvReasons wrote: post #12

The recent election should have made it clear to everyone that many communities, centered around industries like manufacturing or fossil fuel energy, have been decimated in the last few decades or so. These are obviously large trends, but in our technological and economic progress we should figure out how to make sure these communities are not left behind. I would like to see a larger effort for the SV community to e…

a) Doesn't actually answer the OP's question whatsoever. b) Drones on uselessly and in a negative tone about the election; get over it.

Re: Ask HN: An idea that you could not implement?

#54

Post by sqs »

bootload wrote: post #47
findjashua wrote: post #41

Earlier quoted context omitted.

something like sourcegraph?

"sourcegraph" dunno, haven't looked at this, will do. Having access to any public source code would be interesting to find ^canonical^ examples. I know where this does't happen, cough stack overflow cough

Sourcegraph CEO here. Sourcegraph shows local and global usage examples for code. It's useful for understanding unfamiliar code or writing better, idiomatic code more quickly. https://sourcegraph.com

To see how we built it and future plans, check out https://sourcegraph.com/plan.

Re: Ask HN: An idea that you could not implement?

#55

Post by ivansavz »

judahmeek wrote: post #21

I've been dreaming of creating a platform that used crowd sourcing to map concept & procedural masteries to higher level masteries and to occupations. Basically, a giant graph database mapping concepts, procedures, systems (versioned), and occupations. I think such a database could be used to improve learning management systems for school faculty and autodidacts alike. It could also serve as the basis for a performan…

That's very cool. I've been thinking about "detailed" prerequisite structures (e.g. [1]) for a long time, but I never thought about connecting with occupations/jobs/competencies. In addition to replacing resumes, it could also replace exams (e.g. mastery-based grades).

[1] https://minireference.com/static/tutorials/conceptmap.pdf

Re: Ask HN: An idea that you could not implement?

#56

Post by bootload »

Animats wrote: post #51
bootload wrote: post #33

good question, going through these ideas I keep in mind: "The winds of change originate in the unconscious minds of domain experts. If you're sufficiently expert in a field, any weird idea or apparently irrelevant question that occurs to you is ipso facto worth exploring." [0] 1/ So the idea I haven't done is build a hacker search engine. I'm not sure what this would look like, but I'd bet being able to search throug…

being able to search through source code at high/low levels might be interesting. It was called "Google Code Search". 2005-2013. [1] [1] https://en.wikipedia.org/wiki/Google_Code_Search

yeah I saw that. Was/is it any good? Is that all hackers search for? I'll take a look at this. Might be some gaps I might find interesting.

Re: Ask HN: An idea that you could not implement?

#57

Post by bootload »

sqs wrote: post #54
bootload wrote: post #47

Earlier quoted context omitted.

"sourcegraph" dunno, haven't looked at this, will do. Having access to any public source code would be interesting to find ^canonical^ examples. I know where this does't happen, cough stack overflow cough

Sourcegraph CEO here. Sourcegraph shows local and global usage examples for code. It's useful for understanding unfamiliar code or writing better, idiomatic code more quickly. https://sourcegraph.com To see how we built it and future plans, check out https://sourcegraph.com/plan .

"Sourcegraph show local and global usage examples for code. It's useful for understanding unfamiliar code or writing better, idiomatic code more quickly." ~ https://sourcegraph.com/plan

Great, will do.

Re: Ask HN: An idea that you could not implement?

#59

Post by _nalply »

Natural language understanding. This will open Pandora's can, sort of, because really intelligent surveillance and ad targeting becomes possible. Combine a neural network and a lexicon and map an uttering to a graph with probabilities where the graph nodes are entries in the lexicon.

Note that this is partly independent from speech recognition. Perhaps speech could be translated to a string of word alternatives with probabilities, and language understanding gives a hint which alternatives make more sense.

In other words, language understanding could help speech recognition.

Re: Ask HN: An idea that you could not implement?

#60

Post by egfx »

slinger wrote: post #5

I don't think these ideas would take over communities neither the world, but they are awesome: 1) Some kind of crawler that goes through YouTube videos indexing them and its subtitles and then when a user searches a word or idiom the web app would show some part of a video that speaks exactly that search term. I know there are other services that does that, but I want something more natural since these services use s…

Both ideas are personally music to my ears. I also happened to work on some of the titles you mention and a huge player of UO. What a game! On that thread, an arcade arena style PvP, where play is progressive and evolving. Basically LORD (legend of the red dragon) with today's VR tech. But Riot will undoubtedly see it and steal it though.
Post reply on HN