Ask HN: What are you working on?

136 points by Anon84 ↗ HN
I know this question has been asked before in one way or another, but in the mean time projects have been started and finished. People have changed jobs or even fields, etc...

What are you working on? And where? Start-up? Academia? Your favorite MegaCorp?

234 comments

[ 2.7 ms ] story [ 251 ms ] thread
I'm trying to develop my own mobile device. Going to be fun. Intel Atom 1.6ghz, Ubuntu-based, hopefully some kind of 5" capitative touch, software written in Qt/C++ (for lack of a better framework/X drawing option). http://avecora.com

Also working with hnuser://jasonlbaptiste on Ramamia (beta), which lets you keep in touch with your family. http://ramamia.com -- as well as status updates for sports games at http://tickrtalk.com (we were launched, but the data source pulled the plug on our API access...)

...and Classleaf, http://classleaf.com - bringing education a bit more into the 21st century by helping teachers create class websites, with homework, test and due dates, events, file attachments, email lists, pages, and more, and class tracking for students. (You can tell I say that pitch too much.) Mostly managing sales staff (they're working on commission, $1000 per sale) to make sales to high schools primarily.

Lastly, working to study/improve SAT/ACT/SATIIs/my abysmal GPA so I can actually get into a decent college come this fall... sigh. Anyway, overview's mostly at http://markbao.com.

I'm really disappointed in how little you're doing with your time ;-)
Lastly, working to study/improve SAT/ACT/SATIIs/my abysmal GPA so I can actually get into a decent college come this fall...

Enjoy having no time for your personal projects and instead having to do mindless "introduction to cutting-and-pasting java 1.1".

Even if his "decent college" has such a class, and he can't pass out of it, I can't imagine that one intro class would take up that much time.
Why do you want to go to college? Speaking as someone who did, I'm not sure you'll learn much that you haven't already.
College social experience, connections, academics, and excuse to be in NYC (that is, if I get into my top choice NYU)
I'm also at NYU. Contact me when you are in NYC.
I believe strongly in higher education (I'm getting my PhD afterall), but sometimes it needs to be said that college is not the only place you can get all these experiences. You will build your own social networks regardless of wherever you go and whatever you do, and your connections from your various businesses are surely just as strong, if not stronger than what you'll find in college. And of course, if you're such a self-starter as you seem to be, you'll take the time to learn the subjects on your own regardless if you go to college or not.

Please realize that there are many different paths through life. We are lucky to be in the top 1% to take advantage of these educational opportunities. But we're also just as lucky to have the freedom to choose how we want to live our life in a way that brings us the most joy.

Speaking as a college dropout, who dropped out due to lack of money and immediately got a programming job, I would like to stress the importance of college. As far as the social experiences go, you will rarely ever be in an environment that is anywhere near the social level of college. As programmers we tend to get jobs where we are locked in on the computer all day off in our own world. I have definitely made social connections from working, but no where near the level that you get with college.

I agree with some of the other posts that you will not likely learn things in college that you don't already know or couldn't easily learn, about _computers_, but my recommendation is to ignore the computer curriculum in college (unless the program is outstanding) and pick a completely unrelated topic that you enjoy. Your college degree will be a piece of paper that no one cares about, so make college about you, not about getting a degree to impress someone else. This also has the advantage of keeping programming fun, a lot of programming classes and homework take topics that are very interesting and make them tedious and unenjoyable.

If I was to go back to college and start over right now, I would double major in physics and linguistics, and minor in greek/roman history and in philosophy. You can still take programming classes, but ignore the earlier classes, figure out the materials, teach yourself anything you need to know, and test out of those classes so you can just take the classes you are interested. There is also a lot of interesting math classes you can take.

(EDIT: typo)

I totally agree, but having a college degree is still a great plan-B, and it gives you a lot more credibility in certain circles which makes dealing with business folk easier in certain contexts.

Can't wait to see the fruits of your labors with Avecora Mark...best of luck!

From your page: "Mark Bao (formerly Steven Bao)"

Interesting intro :)

Start-up, while undergrad at Georgia Tech. Fun times :)
(comment deleted)
On my hosting start-up: http://webbynode.com. I am the programmer behind our Manager app. :-)

Our main difference is providing ReadyStacks for easy bootstrap of servers: watch the screencast for more info here - http://webbynode.com/railsvimeo.

Oh, hey, cool. I heard about Webbynode with the video that RailsEnvy did documenting the reasons to use it (which I found on Rubyflow). Can't seem to find it right now, but really liked the features. We're on Linode right now, but looking at our other options...
I'm getting ready to switch from consulting to bootstrapping for a while, so I'm looking for new startup ideas and partners.
Feel free to shoot some ideas by me. I'm looking for ideas/partners too.
hmm me too. you guys wouldn't happen to be in the bay area would you? i've got a couple of ideas that i want to pursue, just looking to build the right team.
I'm in the bay area, and also batting around some ideas. We should chat.
I'm hosting Hackers and Founders Meetup this tuesday night in Mountain View if you guys are interested. www.HackersandFounders.com
Working fulltime for http://transpond.com (web-based widget generator/builder)

In my spare time I'm making tools for web developers at http://www.binaryage.com.

Also look at http://hashpage.com (mashup builder on top bespin, github and google app engine) code: "michael" -> click to generate homepage -> woid -> "edit"

Anyone uses XRefresh, FirePython, FireQuery, FireRainbow, Visor? Gimme feedback! :-)

Starting up my second start-up, in the mobile space. Doing customer discovery right now, so it's a bit early to talk about it, but I have high hopes (and no, it's not making iPhone apps). There are a lot of opportunities in mobile if you don't focus on the app store.

Other than that, doing iPhone and (soon) Android consulting to pay the bills.

Any tips on finding "opportunities in mobile if you don't focus on the app store."?

Also how does one get into iPhone consulting. Do you still get paid if your client's app gets rejected?

Think B2B, for example. The iPhone (and the others - Android, Palm, Blackberry), combined with the app store model provide two things: 1) users effectively get a handheld computer which always has internet access, and 2) developers get much more access to these mobile devices than carriers have traditionally allowed. Pick an industry - what does this allow them to do that they couldn't before? (ideally, something that is difficult to do with a simple mobile version of a website). Chances are they're hearing a lot about this iPhone thing, but don't really know what to do about it - can you help them figure it out?

As for consulting, I started out with a bit of luck, a company I was subletting office space from wanted an app made. Past that, referrals are priceless. And yes, this hasn't come up yet, but they absolutely pay if the app gets rejected - you get paid for the work you do, not for the success (or failure) of the app. Most clients are curious (and fearful) of the approval process, and it's very valuable if you can guide them through the submission process and likely pitfalls - but approval is not ultimately your responsibility.

Sorry for the newb question, but are you saying that ANY app developed for the iPhone (including a custom app for a client) must go through the App Store before being installed on their phones? I'm asking because we have a web-based intranet app that we could create an iPhone-based version of, but it's only useful to the client. So, there's no reason to put in on the App Store. But it sounds like you're saying all apps must go through it. (Edit: I don't own an iPhone so this is new territory for me)
There's a way to do it but you need to be an approved "enterprise" developer. Nobody I know has actually completed the process. This PDF has more info:

http://manuals.info.apple.com/en_US/Enterprise_Deployment_Gu...

Thanks for the link. After reading the deployment notes, I'm kind of shocked that they make it so hard to deploy custom apps that you have no intention of selling via App Store. It sounds like a web-based app tailored to the iPhone is the way to go.
Wow, you can't even load your own apps onto iPhone? Damn, I knew Apple was closed-minded, but this takes the cake. If you spend that much money for that monstrosity, you ought to be able to put what you want onto it, you know, like all other phones i'm aware of. Holy Crap. Just another reason I'll never buy one. Can someone confirm this? I'm having trouble even believing that. If this were true, how do you test it, with only an emulator?
For small deployments, you can use "Ad-hoc" distribution - users have to jump through some minor hoops (give you their device UUID, then drag the app file into their iTunes and sync to the iPhone), but it works. It's mainly aimed at distributing copies for testing or reviewing before an app hits the app store, and I think there's a limit to how many copies you can distribute this way (100 different devices? Don't remember 100%, I'd have to look it up).

You can also physically install the app yourself from your development machine if you get their device (again, there might be a limit on the number of devices, since you have to register each one).

As rdouble mentioned, there's also the enterprise program.

And, of course, there are jailbroken devices with third party app "stores". I don't have much experience with these, but jailbroken iPhones are supposed to make up a significant percentage of users - might not be as useful for business use, though.

Ahh ok, thanks for clearing that up. I remembered after my last post that my professor showed me a free Standford web class on developing for iPhone, and they were giving real, in-person students iPhones to test their apps on. Still, they shouldn't limit this, nobody would overtake the appstore, they just can't stand even the smallest percentage getting around them. Makes me sick. The iPhone could be such a better device, but Apple keeps cutting apps for the use of 'undocumented APIs'. A few apps made the camera do things that Apple never imagined being possible, and they get cut for the above reason. I just wish they would open up more, broaden their horizon a bit. They're such control freaks.
Just switched my personal website from Wordpress over to GitHub pages using Jekyll. So far, so good.
Bringing up a new embedded hardware design (RFID) at work (small and stable private company).
Doing my 9 to 5 at http://www.cukerdesign.com

Nights and weekends are split between developing my side-startup (a service to help makes diabetes management simpler), the occasional WoW session, and spending time with my wife and year-old daughter (not necessarily in that order).

Also planning a natural language datetime parsing service with a friend who was an old boss at another startup.

Recently "finished":

eproject - file/buffer grouping for emacs, http://github.com/jrockway/eproject

cperl-mode for MooseX::Declare - http://github.com/jrockway/cperl-mode/tree/mx-declare

On my agenda:

Emacs/LLVM; compile Emacs to LLVM bitcode, link with extensions written in arbitrary LLVM langauges. Right now, even my simplest attempts immediately segfault, so it is on hold for a while. http://github.com/jrockway/emacs

(Once this works, I want to make ECL run on LLVM. Then we can write Emacs extensions in a real programming language. I hear Python compiles to LLVM now also, so that's another option. Eventually Perl will too.)

HTTP::Engine refactoring; removing unnecessary metaclasses and adding support for requests inside preemptable coroutines: http://github.com/jrockway/http-engine (A threaded web server without the disadvantages of threads.)

Persistent application framework: http://github.com/jrockway/eventful (Generic foundation for web applications, TCP servers, IRC bots, ...)

Persistent command-line application framework: http://github.com/jrockway/app-persistent (Make your slow-starting Perl application start up instantly.)

Path::Class replacement based on Forest::Tree::Pure: http://github.com/jrockway/data-filesystem (I also have a binding to make a Data::Filesystem tree into a FUSE filesystem.)

Once this yak shaving is out of the way, I have a few applications to write:

Filesytem::Kindle - plug in your Kindle and see the books as unencrypted HTML (this one is going to be released anonymously, you didn't hear this from me, I was never here...)

PleasureChicken - email / im / irc message consolidator, indexer, data extractor. The idea is to let the computer read your messages for you, so you don't have to. When you buy something from Amazon, it will stick the UPS delivery date in your calendar, and remind you that you just spent $300 on your Amex. When your boss IMs you at 3am with the word "broken", play a sound. etc., etc. This might be a hosted service one day, as well. http://github.com/jrockway/pleasurechicken

Angerwhale 2 - A weblog for programmers that doesn't suck. http://github.com/jrockway/angerwhale-ng (See also: http://github.com/jrockway/angerwhale for the old version.)

Unfortunately, my time is being taken up by menial $work_tasks, which distracts me enough to not get much done. But that should all be ending soon.

A free ear training and music theory software http://www.trainear.com/ that few people use because no one cares about ear training and music theory
I do. I'll give it a try.
I've used it, it rules. jimmyr.com is pretty cool too, used to be my homepage.
what beat it
news.ycombinator.com, hah. that was most of what i read anyways so i just switched over one day
My wife is a beginning fiddler and currently uses www.learntohear.com. [She cares a great deal about ear training and music theory.] She likes how your site lets you choose songs that correspond to specific intervals.
thanks for checking it out. if you guys have any questions feel free to contact me. the songs are the main benefit. I couldn't master the higher intervals until i associated songs
This is completely awesome and something I've been looking for, for a few years. Thanks!
Looks like you were wrong, people do seem to care. There must be plenty of band kids out there that would find this useful.

Popular by Nada Surf is down, have you heard of grooveshark.com? http://listen.grooveshark.com/#/song/Popular/21866669

Thanks for the updated link. The site gets about 200 new visits and 200 returning visits a day. While this is good it's a low turnout compared to how much I've linked it everywhere possible.

On the other hand my retarded anime episode crawler at crawlanime.com gets several thousands hits for a half ass'ed effort with curl. my roleplay site at Eliteskills.com/rp is crawling with people making god damn twilight roleplays at a depressing rate. I'm rewarded more often for making stupid shit than for making well designed sophisticated apps for improving actual skills.

That's where my frustrated statement comes from.

Ah, twilight could get anyone down.

Well, your intro video is very good, a good script. And, you jargon disclaimer is hilarious. Most good intros need a good disclaimer to remind new users that the ideas behind the jargon are simple and the real goal. The ideas are the forest, the jargon are the trees.

Your site is very, very visually busy. Can you reduce the clutter? This includes the app itself. Maybe you could widen the content area, encroaching upon the whitespace.
The ear trainer is very cool.

I think it might benefit from a simpler interface, and/or making it more obvious where I should look first. I wasn't paying much attention when I first checked it out and I thought I was supposed to hit the piano keys to answer the questions. Then I read that I could hit Replay, but I didn't see the Replay button, so I just hit Play. This changed the interval, which I thought was the intended behavior and didn't like that. Then later I saw there's a Repeat button, and then I saw the song answer choices.

If I were you, I'd do something like:

-Make the Song Answer Choice buttons bigger and/or brighter.

-Bring the Play and Repeat buttons closer to the answer choices and the piano.

-Change the instruction to click "Replay" to be consistent with the word "Repeat" (or change "Repeat" to "Replay").

-Hide the statistics, songs and other buttons from the main interface; instead, have them accessible with Show/Hide buttons or tabs.

I know there's a tutorial video, but I think you could still alter the interface to make it easier for people who don't want to watch videos (for example: me. I'm on a relatively slow connection right now and YouTube videos take ages to buffer.).

You might also like to grab random people, put them in front of it, and watch what they do, see what confuses them, etc.

I think it could be really successful! Maybe post it to Hacker News separately as a "review my app" kind of post.

Thanks for the thorough feedback and correction for repeat. I do plan to make a simplified version eventually but for now I'm still adding features and trying to document the most recent ones.

For now there's video tutorials there that guide through the basics of using the ear trainer and further instructions on the help page. There's a few things that are counter intuitive but are far more efficient that way. For example when you click the wrong answer the correct associated song starts playing. Clicking star wars and having it play here comes the bride might seem confusing but it's necessary for improving the song associating with the correct interval. Also automatically playing a new interval when you're correct is confusing at first but the more clicks saved the more efficient training can be.

I can't guarantee anything if they don't watch the tutorial or read the help page for now. Making it beginner foolproof would be great but not if older users have to make 5 - 10 clicks when starting up to get it back to the most efficient settings. I tried a version with elements hidden but in the end it's better just make another version and call the current one the 'advanced version'. The song buttons are packed because they're dynamic in length. The songs can be named anything in the song editor. Also if you select all chords or scales in the first dropdown it already overflows the expected area. There can be a lot of answer choices.

Thanks again for the detailed comment, It'll try and reciprocate my time if I notice somewhere I can or if you tell me something you want feedback on.

Formerly: 2 years at Publictivity as founder/CEO

Now: Ramamia.com co-founder with Mark Bao. Part time side project, but may make it full time. Also, CTO of MIT Enterprise Forum in florida (nonprofit).

more: jasonlbaptiste.com

i should be working on my startup (on profile page) but i am recovering from catching mono and have been sick for several weeks now.
I quit my job a month ago for my second crack at bootstrapping a business: http://www.shirtstastegood.com

I'm in nyc and I would love to hear any feedback! Thanks.

If you like any of the shirts, I made a coupon for HN readers: http://www.shirtstastegood.com/coupons/lXgcYWo4XEQ9fVAq
Love the shirts. Wish you the best of luck.

How about taking it one step further, and giving viral video content creators an affiliate cut in the shirt profits. With Store front widgets, that could be placed anywhere. Transitioning into a platform.

We do sketch comedy, and I could see us using a service like that.

We are working on a viral video incentive program right now!! The store front widget is a great idea, we did not think of that. Do you mean place the video embed and accompanying t-shirt into a widget to be included on blogs etc? That way the content creator has an easy way to spread the t-shirts as well because they get a cut of each sale! Thank you so much for the input!! I love it!
exactly, an embed code that could be placed on blogs, or other sites. It seems like a great angle for everyone.
Those are pretty good! I can definitely see the appeal (which probably means I spend way too much time online).

How tough is the online t-shirt market? It seems like every time something goes even remotely viral, everyone and their dog tries to sell t-shirts. Do you even see random websites and opportunists as real competition (do they actually sell much?), or does your competition consist mainly of the other large t-shirt sites?

Thanks! There are a ton of t-shirts out there made from viral videos, but they are really scattered throughout the net. The biggest provider would be zazzle, but the designs are generally not created by graphic designers. We want to become the definitive place to look for high quality youtube inspired t-shirts.
Great idea, good artwork!

Minor amateur opinion: the artwork seems too large on the shirts / too overstated.

Thanks! We appreciate every bit of constructive criticism. We did make the artwork for a few shirts deliberately big. For example, we wanted We'll Do it Live to really be in your face just like good old Bill. Or do you think we are too bold with all of the shirts? Thanks again for the input Adam!
transitioning away from my legal work back to startup work, learning rails while working on 2 early ideas and 1 underway.

the 2 ideas i'm at will to share:

1 - .org - wikipedia-style database of case-law commonly used by pro-bono lawyers -- create something that would assist legal clinics manage their information locally but also something that stores that data and shares it with other clinics doing similar work. caselaw is supposed to be free and without copyright but somehow its come to be locked up by the lexis/westlaw duopoly. I understand they add value, but we believe there is room for a simple database of opinions & briefs, even w/o the shepherdizing.

2 - .com - brainstorming site for early stage startup development, using principles from Covey & Napoleon Hill. i.e. weekly meetings with small groups of founders discussing multiple ideas, voting immediately upon disagreements, etc.

I'd love to discuss the last idea as its the one I'm actually working on coding right now but the folks I'm working with want to remain hush-hush for now.

I'd be fascinated to hear folks' ideas for startups with a broad focus on the role technology can play in community development, government & regulation & even stuff like troll regulation / comment scoring. If anyone wants to IM or chat, hit me up at hitesh@gmail.com.

Your .org sounds like a possible .com
I can see that. We thought we'd keep it wiki-style at the beginning, possibly setting it up as a foundation. we think its more likely that legal clinics & others are more likely to trust it & use it that way. We would start as narrow as possible, possibly focusing on mental health working with the clinic I used to work with and then expanding outwards.
I started to write a comment on how/why this could easily be a business. I have recently been involved in this area.

But then I remembered some interview with the founder of betterworldbooks.com: "This idea would be best embodied in a company." In this case, maybe it's not. Lowering the barrier for a non specialist lawyer to work on the kind of cases that attract pro-bono attention is good work.

I think there's something to be said about the indication users of the site will get from knowing that profit-maximizing will never be a goal of the software. It may pay some salaries (as wikipedia does) but it will never sell out its purpose first and foremost as a collaborative tool.
htsh - would like to get in touch. give me a ping - me at euwyn.com
excellent. will email after my breakfast & coffee this morning. -hitesh
Forgot, something I sometimes do on the side is work on my fake Sean Hannity fan-site @ http://www.fannity.com. This is a intentionally ignorant joke that attempts to fool people, both liberals and conservatives, with sometimes hilarious results. A story of mine was posted on blackplanet.com and it led to some hilarious comments left by a few thousand visitors from that site over a weekend.

Also, at one point the fake news site 23/6 thought my fake news site was real (and actually advocated that Sikhs put flag pins on turbans). This is perhaps the most hilarious thing that happened on the site: http://fannity.com/?p=553&cpage=1 . The users "backdoorman" & "gacracker" were in on the joke and pretending to be Hannity fans.

to clarify #2 - the question is whether businesses can be crafted somewhat similar to how open source software is crafted, which is democratically. Somehow, entrepreneurs could get together and butt heads and form small idea-groups that go off and talk on their own, with features like the software figuring out how much each person's contribution is worth, etc.

With free software, say someone has a good feature they want implemented, they have a means to affect that change in theory in a way a user can not do with closed source software. If they can't win the project over, they can fork the idea and start their own variation. the theory is that these kinds of mini-democracies lead to the most efficient software, could this work for businesses as well?

Besides YC, I've been working on Arc. (No writing lately. I can't seem to focus on more than 2 things at once, so it's always a choice of Arc xor essays.)

Specifically, I've been trying to do things to Arc that will make News shorter. I'm running out of room, though: News is 1886 LOC, and it's rare now when I can find something that will cut as many as 5. So I'm going to try writing some other types of applications to make short.

1886 is short. Anyone else surprised?
Read the source code for arc.arc, srv.arc and news.arc, and you'll probably learn something new about programming. I did, at any rate. Apparently PG isn't just some investor who can't hack... :^)
Care to give some examples of things you learned?
speaking for myself, the idea that everything has an id from the same range and that the context figures out how to use that id was a neat thing. I'd have split it up into several structures simply because that seems to make the most sense but this works well and in fact makes the code very compact.

Articles, comments it's all the same.

I'm playing with keyword extraction algorithms in arc. Yesterday I discovered it was super easy to pull the classic porter stemmer C library (http://tartarus.org/~martin/PorterStemmer/c_thread_safe.txt) through mzscheme's foreign-function interface and into arc. That's not really about the language, but it suggests arc made a really good platform choice. Just by making mzscheme salient it's responsible for getting me hacking more on lisp.
Any thoughts yet on what other types of applications you'll be working on?
> Besides YC, I've been working on Arc. (No writing lately. I can't seem to focus on more than 2 things at once, so it's always a choice of Arc xor essays.)

Fatherhood, Arc and YC are three things, though:-) Don't know if it's your intention to share anything about the first one, but I'd be curious if you did, and very understanding if you didn't.

I'm creating an online spiritual-learning (Christian) site for kids: http://kidbuilder.net.
you're in the wrong website. try: love the almighty dollar, with all your heart, and your mind, and join a startup with your fellow, as you promote yourself.
I'm just working my day job over at eirestudio.net - design/dev work. I have a few side projects going which are mostly for fun and learning more JavaScript.
I'm working on a solution to unify all the major SCMs (ClearCase, Git, Perforce, Subversion, etc.) and abstract the data it in such a way that I can provide information that can be digested by all levels of an organization. I'm calling this my full spectrum tool as it's intended to be as useful for developers as it is for executives and everybody in between. Basically I'm trying to change how we communicate and access information when it comes to software development.

I've been working on this fulltime for over a year now and it seems like I can't make a dent in it. Don't get me wrong, I have a very solid foundation so far that includes an indexer and a very flexible server/client framework that should work for any organization. However, even with this I still have a boat load of work to do.

There are definitely days where I think I've gotten way over my head, but I figure I've invested too much to walk away now.

Have you ever thought of open-sourcing it (or even parts of it)?
It's still too early for me to think about this, but I'm sure some aspects of it will be open sourced.
Working my 9 to 5 (which is more like an (8 to 8), and trying to learn Django well enough to start an open source product I think Django makes sense for.

Django is fun, and surprisingly productive even while learning, though with little free time it's going slower than I'd like.

Making a new spreadsheet. Trying to make it fast enough. Learning about spatial data structures.