Ask HN: What can I build for you today? | Hacker News Reader

Ask HN: What can I build for you today?

35 pointsAlexMuir54 comments

What's your biggest pain point?

I want to build a webapp that people and/or businesses will find useful. I'll start building it right now.

One webapp per comment will make it easy for people to vote.

At the risk of sounding like a jerk, you're asking the wrong people. This is a community of builders, you need to ask the users (or potential users). Go interview small business owners in your area. Call up companies and ask to talk to the office manager. Let them know that you're interested in building something that will solve a problem of theirs so that you can sell it to other small businesses, but that you're not trying to sell them something, you just need their help in identifying painful problems that nobody else is solving.

i partly agree. on the one hand, we're all builders and could take good ideas and run with them.

on the other, we're also thinkers. i have more ideas than i could ever find the time to build, and are usually at least mostly busy with our own best ideas

It's true, we are just a small subset of people who need things built to solve problems. But we also do have problems to solve, which is why companies like GitHub exist.

If building tools to solve software development problems is something you can get excited about, go build those tools. But don't ignore the sea of possibilities outside our world.

We are also a community of fast adopters, and this is one reason we are a good group to build things for.

People here understand what's possible. They are aware when something is sub-optimal. If I go and ask the majority of people on the street they won't even know they have a problem.

Builders are users too.

A way of reading HN that lets me say "That was interesting" and "That is of no interest to me" and then presents, when asked, all the posts from the last 48 hours based on how interesting a Bayesian (or other) filter thinks I'll find them.

I've been intending to write my own, but I've been too busy of late to make any progress.

I'm not convinced - there's already an incredibly powerful filter on HN that finds and reads all the articles and rates them as being of interest to the audience. I suspect that the appeal of something that filtered HN for people with narrower interests (of whom I am one) would be pretty limited. I'm looking for something that is widely useful, rather than a fun coding project (which I'm sure this would be).

You asked for a pain point, I provided one. These days I'm finding most of the "new" page to be uninteresting, and worse, most of the front page to be uninteresting as well. Many, many good, interesting and useful items slip through the net and get insufficient upvotes to hit the front page, and yet on occasion when re-submitted get noticed quickly enough, hit the front page and stay there for hours, sometimes more than a day.

So when you say:

    > there's already an incredibly powerful filter on HN
    > that finds and reads all the articles and rates them
    > as being of interest to the audience.

... I guess I don't know what you mean. And if you just mean the current mechanism of having a wide readership that rates submissions, it's not working for me.

But I would guess that you're unconvinced and won't work on it, but that's your choice, and I fully support it. Someone might build it, it might be me, and I'll be really interested to see how much use it gets.

If any.

Sorry if I came across unappreciative - I do appreciate your idea. I just tend to enjoy seeing items that I wouldn't otherwise consider reading - perhaps this is still a bit of novelty from not having been here very long.

I find that if I'm away for a few days the Best list helps me catch anything good I missed.

No, indeed, and I wouldn't want to read only those items marked as being of interest. I just want to catch those that I should see, but missed because no one upvoted them.

Something that ranked every non-dead article is well outside the scope of my skillset, even just looking at title and source. If you could get that working you'd be onto a winner.

Huh. I've been having the same idea. Except it doesn't have to be limited to HN. I was thinking more along the lines of a standalone system that you can give RSS feeds or whatever (a way to fetch new articles, in any case). You train it like a spam filter, except you mark articles are interesting or uninteresting. It goes out, gets more articles, presents the ones with the best scores, you train it further, etc.

I've been kicking this idea around for a while, but like you, I haven't actually started, mostly because I'm not sure about the best way to do it. (E.g. web app, traditional desktop app, etc. Not to mention the implementation language...)

Check out newsblur.com. The code is on github (all python and js).

That looks almost like what I had in mind... except you have to tell it what you like and don't like about an article (limited to author, title, publisher). What I had in mind was, feed the text of the article to a bayesian filter, and let the system make decisions based on that.

At first sight, telling newsblur that you like an author or publisher, doesn't seem to useful... you might as well subscribe to their feed in a regular RSS reader then. But maybe I'm misunderstanding it.

I thought your idea was great and then I tried out NewsBlur and came to the same conclusion you did. It seems to be filtering based on title, author, publisher, and tag rather than doing any statistical filtering of the text. That's not very useful because I already know who produces great content consistently and I subscribe to their feeds. What I need is a way to filter out the good articles from inconsistent sources.

Wouldn't it be much better to base this off the upvotes of people with similar tastes instead of e.g. the article text? That might get you the best of both worlds (social/Bayesian filtering.)

It's possible that you already meant this, but it wasn't clear to me, at least...

I've thought about that it a lot but also as a way of using text classification to divide up the stream based on context (international, business, hacking, software development, research).

Pay-per-answer email service. I want to create email address- smthg like @payperanswer.com

I can give it to recruiters, and all ppl that want something from me. Senders need to credit say $1 to sent me something to that inbox and once i reply to that email money go to charity of my choose (and 50% go to payperanswer.com or something).

I wanted to build something like that myself, but dont have time atm.

Extra bonus - convince Steve Jobs to use, price it like $1000 to Apache SW Foundation, and publish all emails publically - what would be HUGE!!! :))))

I would be interesting in being a part of this project! Contact me if you want to take this from inception to conception!!! Email is on my profile.

sorry i didnt get it.. why do you need my contact? feel free to implement this thing, i don't hold any patents for this AMAZING idea :)

I rely on your sense of justice and just added red Ferrary to my amazon wishlist

I need a book listing service similar to http://www.7bks.com, but which shows ratings (by me) and category. The list should be embeddable in my website. (eg: http://sivers.org/book )

(The service could then extract the user ratings and make a top books list for each category.)

Sounds like the creator of 7books has a couple of features to add.

Hey, I'm the creator of 7books :)

Embeddable lists are definitely on my roadmap. Drop me an email (tom@) and let me know exactly what you want and I'll see if I can build it for you!

It seems we need a better way of tracking what existing webapps are available.

Great idea. Sort of like an Appstore for webapps (without payments). http://www.appappeal.com/ was the first thing I came across - it looks awful. Anyone used it?

agreed. i kind of started working on a solution to this at some point in time, but i got discouraged. might try and finish it up this weekend.

For anyone making this, let me search not only for "image editor" or "spreadsheet", but also for web apps "like [desktop app] Photoshop" or "like Excel", etc.

A sexy dashboard that grabs data from Google Analytics API and displays it all - suitible to have a monitor up on the wall showing various stats from 1-10 websites, highly customisable, very beautiful.

Heard of http://geckoboard.com/ ?

I now have, and will check it out on Monday - thanks!

edit: don't suppose you have an invite code?

Hi guys, I'm the founder of Geckoboard. Thanks for your interest (and thanks for the mention lachyg).

We currently have a a backlog of > 1000 invites. It's still early days and the product is a little raw but if you're interested in checking it out mail me (address in my profile). If you include your HN handle I'll get your invite out straight away.

heard of? Yes. On it? No. Got an Invite?

There was a similar YC startup. Does anyone remember what it is called?

Leftronic - looks great but I haven't been able to check it out in person.

it looks beautiful, wish I could try it out.

There seems to be some slight differences between leftronic and geckoboard though. Geckoboard collects data on your behalf, not sure if leftronic does it

Write a small PHP plugin that will allow our users invite their friends. This plug-in can be embedded as an iframe, and then it will ask the user to select his email service. Options would be gmail, yahoo and any other email services that provide an API. On clicking next, he will authenticate, and then his contacts will be displayed, with none selected. He can select as many as he wants. Pressing next once more will allow him customize the invite email, and a final 'invite' press will invite all his friends.

I believe most of us here could use this.

Uh oh, make sure you provide a way to export your contact data too- you don't want to make google mad!

You can use OpenInviter: http://openinviter.com/

I think you have to signup to download it, but it's free and you get the source.

where are you going with this invite idea? I'm working on a phone based invitation product at http://televite.totalengage.com I'm just curious what your plans are :)

I would like a shared to-do list that is suitable for touchscreen devices.

I like this idea.

Building it - http://gettasklist.com

Cool!

The webapp that shows exactly what you're looking for in this "Ask HN" post.

I've created Wappr.com. Too much hassle to moderate/filter (100% neglected right now). I hope you have a much better solution.

Please, please make an application that will make it easy to track supporting and opposing views on a topic. The idea being to replace discussions (mailing list, real life meetings) that get lost in word games and politics. The application would bring transparency to a decision and allow any people not acquainted with the discussion to get up to speed quickly. An evidence-based transparent decision process I guess.

http://www.procon.org/ does this, but for more popular topics.

A plugin/gem for rails apps that allows you to build a feedburner-like feed subscriber tracking system so you don't have to go through FB all the time. If all it does is accurately report the number of subscribers, I'll be using it on a number of sites. If it does more, all the better.

Feedburner in a gem, basically.

A tool similar to Delicious that aggregates content that you've Liked on Facebook (within the wall) and marked as a Favorite on Twitter. There's a few of us that are using those two methods to mark content to come back and read later.

Not exactly, but http://historio.us/ is similar.

Sounds like Instapaper

A sync tool between Mac Outlook 2011 and Google Calendar and Contacts

A web app manager. The browser window is far from optimal for using multiple web apps at once.

Upvote a post on HN if I click on the post in the New section.

Reply on news.ycombinator.com