4 ms·

Ask HN: Idea for paid q&a for programming problems

1 points by palidanx 13y ago

What do you think about an idea where anyone can submit short programming questions for a $1, and then get it answered in 'real time' from the community. I often have a problem with some tech things throughout the day. For example I might be working on a regex driving me insane, only to solve the issue maybe an hour later. This would be different than forums and stackoverflow where the goal would be to get the tech question as quickly as possible for those who are programming. It would be kind of similar to fiverr.com, but more focused on the programming community. Ideas/feedback?

hardwaresofton 13y ago
I've had a really similar idea (different domain, but very VERY similar) -- and this is one of the problems I ran into while implementing it: How do you pick who gets to answer? Is it only a speed thing (which might be bad because then you might get some bad answers? but it's hard for the user to know that an answer is wrong?) That sounds pretty viable, a nice way to make a quick buck (or maybe even offer some charity contribution) for knowledgeable people. Though, you should probably make it clear and above board that the service caters to people who are NOT looking to google for hours. I think making it clear that the service is offering an alternative to that is awesome. Like I said, I had a really similar idea that I have actually started to flesh out, in a different domain so I think this is a good idea/something worth pursuing.
- palidanx 13y ago
  I was thinking about that, and here is my idea of the workflow. When someone posts a question, multiple people can answer, but the question asker only sees the first answer received. He can either choose to accept or reject that answer. If he rejects that answer, he can move on to the second answer and then accept/reject accordingly. Because if it is like a forum where you see the entire thread, you might run into a problem where you choose between the answers.
  - hardwaresofton 13y ago
    So yeah, here was my problem with that: How does the person know what's a good answer and a bad one? If they see the answer, then people figure that out and only go for speed, and possibly produce bad answers, which COULD sound like good answers to the person who's asking the question.
    - palidanx 13y ago
      So here is like a hypothetical example: + Let's say I have a problem with a regex which just isn't working 1) I post the question to help solve and fix this regex 2) I receive answer 1. I test it, and it doesn't work, so I reject it 3) I receive answer 2. I test it, and it does work. I accept the answer as correct, and they get paid accordingly. My view is when the question is asked, nobody can see the answers except for the asker until the entire thread is marked complete. People can choose to answer a question based on reputation of the question asker (how many they reject), and they can review their question history to see if the person is doing the right thing.
      - hardwaresofton 13y ago
        Ah, that actually makes a bit of sense -- but what about quality of solution? like if they see a second answer and if they decide the first answer was better? This model actually seems pretty viable, my idea was actually related to real-time translation apps, so it's a little harder (for the user) to establish which is better or not, but if it's a program, then they can definitely check if it works (at least) right away (and probably with some good google searches or even an explanation, find out what the person did that was different/special/good)

1 points by palidanx 13y ago

hardwaresofton 13y ago

I've had a really similar idea (different domain, but very VERY similar) -- and this is one of the problems I ran into while implementing it: How do you pick who gets to answer? Is it only a speed thing (which might be bad because then you might get some bad answers? but it's hard for the user to know that an answer is wrong?) That sounds pretty viable, a nice way to make a quick buck (or maybe even offer some charity contribution) for knowledgeable people. Though, you should probably make it clear and above board that the service caters to people who are NOT looking to google for hours. I think making it clear that the service is offering an alternative to that is awesome. Like I said, I had a really similar idea that I have actually started to flesh out, in a different domain so I think this is a good idea/something worth pursuing.

palidanx 13y ago
I was thinking about that, and here is my idea of the workflow. When someone posts a question, multiple people can answer, but the question asker only sees the first answer received. He can either choose to accept or reject that answer. If he rejects that answer, he can move on to the second answer and then accept/reject accordingly. Because if it is like a forum where you see the entire thread, you might run into a problem where you choose between the answers.
- hardwaresofton 13y ago
  So yeah, here was my problem with that: How does the person know what's a good answer and a bad one? If they see the answer, then people figure that out and only go for speed, and possibly produce bad answers, which COULD sound like good answers to the person who's asking the question.
  - palidanx 13y ago
    So here is like a hypothetical example: + Let's say I have a problem with a regex which just isn't working 1) I post the question to help solve and fix this regex 2) I receive answer 1. I test it, and it doesn't work, so I reject it 3) I receive answer 2. I test it, and it does work. I accept the answer as correct, and they get paid accordingly. My view is when the question is asked, nobody can see the answers except for the asker until the entire thread is marked complete. People can choose to answer a question based on reputation of the question asker (how many they reject), and they can review their question history to see if the person is doing the right thing.
    - hardwaresofton 13y ago
      Ah, that actually makes a bit of sense -- but what about quality of solution? like if they see a second answer and if they decide the first answer was better? This model actually seems pretty viable, my idea was actually related to real-time translation apps, so it's a little harder (for the user) to establish which is better or not, but if it's a program, then they can definitely check if it works (at least) right away (and probably with some good google searches or even an explanation, find out what the person did that was different/special/good)