Ask HN: What are you working on?
This question comes from a great thread from a while back that I really enjoyed (http://news.ycombinator.com/item?id=700662), so I thought with the new year here, it'd be a good time to ask again.
I'm still developing readthekanji.com, a site for helping Japanese students learn to figure out how to read kanji. It's been one of the best learning experiences, and I'm loving every minute of it.
So what projects are you currently working on, or planning for the new year? Is it a startup, or research perhaps? And how's it going? Started yesterday: Lark, a toy Arc-like language implemented on top of Python instead of Scheme. (http://www.github.com/swolchok/lark) I've never implemented a Lisp or used it for anything serious, and I've been trying to figure out for myself whether there really is anything to macros or if they're just fancy compiler hacks, so I decided to write Lark as a way of learning what precisely pg is so excited about. It's a direct, lazy translation of ac.scm, except that I've written an evaluator instead of compiling to Python ASTs or bytecode. I'll probably compile to Python ASTs after I have a baseline. I did cheat a bit and steal pieces from PyScheme and Lython as well as someone else's S-expression parser, but it's capable of evaling ((fn (a) a) 1) as well as Arc's "if" currently. I'm not certain whether it's Turing complete right now, but I suspect that it is because I can create and call functions, branch, and bind values to names. (IIRC, just being able to create and call lambdas is enough because you can count, branch, etc. Is that right? I haven't taken PL.) Big problems right now are: 1) Proper lexical scoping. The current model is broken (it might be dynamic scoping), and Python 2.x's closures are broken so I can't just use Python functions to punt the problem to Python. 2) The shortcuts for quote, quasiquote, complement, compose, etc. The S-expression parser I stole doesn't have them. I don't want to use a parser generator because that seems to defeat the point of Lisp's lack of syntax (i.e., being easy to parse). This is a toy that has nothing to do with my research interests (security), so it's not going to be actively maintained or developed. pg will probably make some breaking changes to Arc and kill the project. just being able to create and call lambdas is enough because you can count, branch, etc. Is that right? ♫ All you need is Lambdas ♫ :) what about http://www.picolisp.org ? fuck compilation. fuck lexical scoping. do something for real in lisp, like * entity relationships over a graph db * webframework * flightsimulator Released http://runroot.com/ today. Give it some upvote love at http://news.ycombinator.com/item?id=1032688 (apparently "OMG! AMAZING UTF-8 CHARS!" is more popular than my paltry attempt at making something useful). Love it - A great use of Solaris Containers. Working on the house I just bought with my wife. It's the grandest hack I've ever had to do. From crazy 22ft walls of glass that need replaced, a kitchen from the 50's that needs a remodel to the shoddy wiring throughout the house. There's tons to do, and little time to do it in. This year I start a new position at the company I work for. Hopefully it'll bring more visibility and access to the people who make decisions. There's so much that needs fixing here, and with a little elbow grease I think I can fix it all. Are you the HN member behind http://big-old-house.blogspot.com? I've been enjoying lurking and following that site since it was mentioned here. Can't say that's me, but it's a good idea. I'll have to bug my wife about getting a blog up. Before and after pics would be nice. Perhaps you should make a blog of it. I love these projects. As a hacker, I have high expectations of you! I seem to have invented a new datastructure, so I'm analysing it to check its performance before writing it up and announcing it. I've also invented a variant of an existing data structure that has some nice features. That will accompany it. In the meantime I'm starting a re-write of my alpha 0.1 web service to help connect people with friends and friends of friends. The existing one still exists, and I'm still collecting comments, so if you'd like to know more, drop me an email. And I've got a full time day job, and 8 talks/presentations in the next four weeks. If you don't mind me asking, what do you do once you've invented new data structure? I think I have recently, too, but (besides thoroughly documenting it and posting a library based on it, soon), I'm kind of at a loss. I'm not in academia, so publishing in an academic journal isn't the clear choice. I've been trying to find prior art, but thus far haven't found anything similar. My e-mail address is in my profile. To be blunt, I'm not sure. I've spoken with the head of computing at my nearest university and he seems quite excited by it. We may publish a joint paper - not least it will give him an Erdos Number of 3. Failing publication, I'll just write it up, put it on my web site, and announce it in a few place. What ever happens, I will announce it here for sure. Haha funny I think I've invented a new fundemental data structure as well. Hopefully we haven't all three invented the same one! Mine is going into proprietary software; I don't know if I'll write a paper over it. It doesn't allow you to do anything you couldn't do before, but it takes the best and worst case performance of a well known data structure and reverses the two. So it's useful in places where the other structure would perform poorly. Good luck with yours, both of you. Ok, thanks, and good luck with yours. I hope you share the details of 'what to do after you ...' here. http://skribit.com - helping bloggers cure writer's block and find things to write about.
@wycats (rails core, etc) recently started using us: http://skribit.com/blogs/katz-got-your-tongue and continually working on what pays the bills, http://paulstamatiou.com Making our e-commerce platform http://www.enstore.com available to everyone. With a nice frontend built in Cappuccino/Atlas. Wow looks great! I really need to checkout Cappuccino... You can get a lot of the good looks on any platform by using 280North's open source UI pack, Aristo. http://github.com/280north/aristo Is there a way to use the UI pack and nothing more? Sure you can get the photoshop file here: http://github.com/280north/aristo Thanks, we are aware of that bug, just haven't got around to fixing it. I've been following your company's work on HN and the Appengine group. Just wanted to say, you guys have some very talented designers! Thanks, I'll pass on the love... Gorgeous...those buttons/themes are pretty sweet. Played with the demo a couple weeks ago. Very impressed and looking forward to it. Just curious, is the storefront in Cappuccino, or just the admin section? The storefronts can be anything. Most of them are html templates (we use django) with css and jquery. So for now just the admin is Capp. looks awesome! impressive. Building that Cappuccino/Atlas frontend "Neptune's Pride is a multiplayer game of Strategy, Intrigue and Galactic
Conquest! Neptunes Pride is real-time, but games are played over several weeks. Players log
in at any time of the day to check the progress of their fleets, view the
results of battles and issue new orders. Explore, Expand, Exploit, Exterminate. It's the game you know and love with a twist. A 4x Strategy game with it's
complexity striped away to reveal a sophisticated game strategic command and
diplomacy. How will you conquer the Galaxy?" I really like the artwork on the landing page. Love the style. Lol...is that a turtle with a space helmet? Love it! I'm interested/ been heavily exploring: PCs in the living room (everything from custom linux distro, XBMC Live modified, and windows 7). 140 million HDTVs sold in 2009 alone, hardware is getting cheaper, and content is readily available. Education space. ie- how do we provide easy access to all the materials that exist out there? My hypothesis: everything we could ever want to learn exists already on the net or can be taught to us by a person we can be connected with in seconds. How do we easily organize it? think more of a directory than a search engine. Bringing local businesses into the 21st century. Most don't have a website and still use yellow pages. The existing solutions out there suck and are filled with slime. Email newsletters. Why not create the weblogs inc of newsletters? Look at what thrillist, dailycandy, etc. have done. Create a network of these around a plethora of wide open niches along with building a strong advertising platform for email newsletters (it doesn't exist yet). Human powered purchasing decisions. How do we help people know what to buy with specific criteria that transcends checkboxes and a search engine? Something human powered is the way to go about it. As geeks, I'm sure you're constantly asked- what phone should I get for price x, features y, etc. Purchases such as these are expensive and spending a few bucks more to get a personalized recommendation would be worth it. I'm working on 0xCOFFEE, a compiler for a toy language implemented using Ruby, TreeTop and LLVM. http://github.com/meqif/0xCOFFEE It's quite fun, but I had to fork llvmruby (the ruby bindings for LLVM), since it lacked some things, like allowing access to part of the LLVM API and raising RuntimeErrors instead of segfaulting (especially because of some code mutations that heckle[1] generates). Currently, this is just a little project to keep me busy during the past holidays and the current university exam month, but I hope to create a nice language. [1] http://glu.ttono.us/articles/2006/12/19/tormenting-your-test... You need some sort of related technology called 0xDEADBEEF. If you want a vegetarian port, there's also 0xFA1AFE1. That's a 0xFA1AFE1 lot for 4 bytes. I'm working on a collaboration tool built to help government agencies and aid organizations coordinate during emergencies. I helped start it this summer as part of the Humanitarian FOSS Project (hfoss.org). It was used successfully in beta for Thanksgiving-day feeding coordination and we're getting very close to our first non-beta release. http://collabbit.org
http://github.com/elitheeli/collabbit If you'd like a look around the demo or are interested in contributing leave a comment or send me an email. Is it something similar to Sahana - http://en.wikipedia.org/wiki/Sahana_FOSS_Disaster_Management... ? It's related, yes. Both are classified as "Humanitarian FOSS Projects" and Sahana has been worked on in previous years by people participating in the HFOSS Project summer internship, which is where we started Collabbit. I'm not intimately familiar with Sahana, but from what I understand it's a fairly large system and can do a lot of specific things. Consequently, most of its deployments have been at sites of major disasters or for very large agencies. Collabbit takes a different approach. Our goal is to facilitate communication inside groups and between groups, rather than to provide a means of accomplishing specific tasks. That means that Collabbit is quite simple and broad right now. While our initial use case was volunteer organizations communicating in New York City, we're hoping it can be a valuable tool for smaller groups as well. I'm working on a distributed / fault tolerant network filesystem inspired by plan9's Venti. Several of its internals (code for chunking data with rolling hashes, a purely functional ADT for non-redundant storing of large strings, and a coroutining / non-blocking server framework for Lua, somewhat like Python's Twisted and Ruby's EventMachine) have been broken out into their own libraries. I'm almost done with the first two and am working on the third, but its design is being driven by the filesystem. They'll all be released at once when they're done, no time frame yet though. Probably MIT license. I've also been working an a utility to locate repeated groups of data (particularly geared towards scanning for likely copy-and-pasted code), but that's been put on the back burner due to the above project and lack of free time. I've been putting together a web site, http://www.learngrowdo.com . It's somewhat personal and came out of my struggles taking care of a loved one with a chronic illness. On the site I also plan to sell related software as I develop it. I'm selling my iphone app on it, which was just approved yesterday. "Control Time": itms://itunes.apple.com/us/app/control-time/id348796242?mt=8 http://www.carbonmade.com - The easiest way to display and manage your portfolio online. I'm working on a travel website: "Planning a vacation is hard. You know where you want to go. But what should you see once you're there? Which monuments, parks, cathedrals and museums should you visit when you're there? How do you get around? My website helps you plan your vacation by providing high quality itinerary suggestions, contributed by travelers like you." I hope to do a private release at the end of this month. I assume you have seen http://offbeatguides.com. If not, might want to check it out. What I'm trying to do is slightly different. A guide will tell about everything that there is to do. So does wading through the forums at Trip Advisor. I'm trying to present itineraries that have already selected interesting things to do. Essentially future travelers are going to benefit from past travelers. What's often important is: How much can I fit in a day? How do I get around? What was my personal experience, and how can other travelers benefit from it? In a near future, you'll be able to search on itinerary based on your travel profile (age group, kids/no kids, fast, lazy etc.) I searched for Bangkok Thailand and found the content returned very familiar - it'd come solely from http://wikipedia.org/ and http://wikitravel.org/. The FAQ confirms that's what the site does. I won't try it again. I was thinking about this earlier today actually.. I know you posted this the other day so I may email it to you. Before I go on a trip, I'll plan out where I need to go and usually copy the addresses into notes on my iphone or into google todos. When it's time to go there, I'll copy the address into google maps and go from there. Perhaps I'm just doing it the hard way, but something that streamlined the process would be great. I was thinking something like, Google Calendar API + Google Maps + Todo list + iPhone App/Web App. Brick and Mortar startup, with online enhancements: Currently in the process of purchasing an art gallery space in Denver, which we will renovate, and hopefully open in the fall of 2010. (Would be sooner, but we are having a baby in the spring.) I intend to have an exhibition every 3 months that will seek to blend technology with a physical installation of some kind. Details TBD, based on online collaboration --
The web site will spend the 2 months between exhibitions working towards collaborating on the next exhibit. Each collaborator will hopefully be able to visit in person, to create a true blending of online and physical life. Where/when/how will you announce? I'm not in the art world, so I wouldn't naturally see the announcement, but I would like to visit. There is an organization in Denver that markets the Art District. I'll likely join them, and therefore be included in their marketing. Otherwise, the location I am working on gets a lot of traffic, and the sign in front will be the major announcement. Our web site will also have an announcement -- it is currently just my wife's blog, but is connected to her etsy shop, so we do have a small following. If there is interest, I can post on HN as well. I wasn't sure if the community here would have a strong interest in my project or not... Finally, simply word of mouth - as my day job will pay the mortgage on the property, I don't actually need to profit, so I can just work with my friends on the first few shows, and get the world out gradually. In terms of timing, I am hoping for August. Once we actually move in (hopefully on the 31st), we need to physically prepare the location (needs new floors, doors, drywall, lights, things like that). So there is quite a bit of work to do yet. My startup, CloudFab, is doing well. I'll have more to say about that in a few weeks. In my spare time, Hackety Hack is coming along, got a release out for Christmas, hoping for 1.0 early next month. Then I have one more small project that's still secret. http://www.80legs.com - web-scale web-crawling for everyone. Launched in September and growing revenue now. We've solved a lot of big data store issues in our back-end. Challenges have shifted from technical to business :) Extractiv - web-listening and content-extraction that combines semantic analysis and web-scale reach for a complete picture of what information is on the web and what the web is discussing. The core technology is working, currently turning it into a real product. While in principle I think the approach is great, I have recently learned that some of your "volunteers" are in fact people infected with spyware. Can't find the corresponding article atm, though :-( Yeah I met these guys when they presented at Rice a while back. I can tell you they aren't doing anything with Spyware. It was actually really frustrating because they had to keep answering this question over and over. They're not infected with spyware. Some of the computing power comes from the users of freeware applications, who are asked _during install_ if they want to enable the grid system that powers us. The article I read was about one of these freeware applications (something starting with "G", and a green Logo or mascot I think, I can not remember unfortunately), which installs the spyware if the users does "Express Install". Buried in the TOS is "we may use your computing power". It was by a third company, which in turn is supposed to be used by 80legs. Perhaps 80legs is not even aware of the shady practices of that company. You're talking about Digsby. If you managed to follow the story to its conclusion, you'll see that they dramatically changed the install process to follow Plura's (our sister company) affiliate agreement. I suppose I had read an older article. Thanks for the clarification! Have you thought about the economics of having a program where people earn money from allowing use of there processing power vs paying for it regularly/ having your own processing power. Even if it was barely worth it for the user after factoring in electricity costs the novelty factor might attract people. We have thought about it.. there is a downloadable app an individual can use, but it's not too likely they'll earn any significant money. Some people leave their computer on all the time anyway, so they would probably be interested. Working on http://herefilefile.com, an iPhone app that lets you access all of your computer's files from anywhere. Currently doing a UI refresh, planning a big marketing push including ads, contacting interested journalists, designing a web site / blog, planning the support workflow, and other fun stuff. Things are going really well for us so far. Won a nice iPhone app competition, getting some pretty good buzz building, and looking to launch to a decent crowd of interested parties! I'm creating an online power monitoring system so you can see how your home or office uses power. A web platform is great for this because we can centralise the processing / graph generation and analysis and then put simple, cheap devices into people's buildings to do the actual measurement. See: http://www.gridspy.co.nz/ We offer live data (updated every second while you watch) and multiple channels, all at a price that is cheap for a solution like ours. http://www.lendfriend.net - friends and family lending site built in ASP.NET MVC & jquery. I wasn't the first one to come up with this, but I got the idea after I funded my first startup on credit cards, had some friends with money, but didn't want to complicate things with a DIY loan. We're launched, but we have a lot of work ahead of us. If anyone has any ideas/suggestions/criticisms please let me know. Developing a generic price arbitrage engine making it easier for sellers to track price movements of their products vis-a-vis their competitors. The first release will target sellers selling on Amazon. I should have a private beta by the end of the month. Just wanted to add , there are 2 features that i am most excited about (at least for the first release): a) Speed
The web front-end and the back-end processing engine have been written keeping in mind processing time, everything is optimised to get data quickly in front of the user. b) Scripted Rule-based pricing
A user can script the pricing engine by specifying rules written in a custom-DSL. My first attempt at doing a proper DSL. I'm writing an iPhone musical instrument with an emphasis on just intonation (as opposed to equal temperament). I wrote a just intonation library in Java, if you're interested: Just sold Bac'n ( http://blog.bacn.com/2010/01/bacn-acquired-by-baconfreak-com... ). Focusing 100% on Urban Airship ( http://urbanairship.com/ ) now, which is fantastic fun. I'd love to read that book once it's published. Fixing poverty through water and sanitation:
http://akvo.org/ Which is also becoming an open source platform for development aid. This is great. I remember having to endure a few days without running water in my apartment thanks to burst pipes and wondering how I would cope - then I realised how lucky I was to be having it return eventually. I've had a few conversations with Mexican colleagues who told me how it is common to find people getting sick in some poor coastal towns because they don't know how to build latrines properly. It's such a big issue. I hope you find every success with this. Building support for matrix datatypes and operations (arithmetic, eigenvalues, and factorization) in PostgreSQL. I think I will use the Gnu Scientific Library, but I am still looking around at options with looser licenses. Try LAPACK. Incredibly fast, and a minimal license. You'll have to deal with FORTRAN, though. I don't want to write my own binary dataformats, and I don't want to deal with FORTRAN. Eigen doesn't help with the first problem, but it does with the second - written in C++ with some nice expression template magic for efficiency. Comparable performance to LAPACK and LGPL licensed. http://tweeteorites.com/ - builds timelines of what your friends are favoriting on Twitter http://amid.st/ - social placemarking app A book called "Reinventing Discovery", about how collective intelligence is transforming science. The manuscript is due to the publisher (Princeton University Press) in a couple of months. Compiling our JVM language (Gosu) down to bytecode. It's slow going; the bytecode side isn't hard, but the language has been evolving for 7 years or so and has all sorts of obscure edge cases, as well as some looseness in the semantics that has to be tightened up in bytecode, so getting every little detail right so all the existing code still runs 100% correctly is difficult. We're still hoping for an initial open source release some time between probably June and September. I'm working on a dissertation linking the international balance of power to the characteristics of civil wars during three periods: pre-Cold War, Cold War, and post-Cold War. Working on an update to http://stormpulse.com that will give us city-level maps and general/severe weather forecasts. I'm starting to think my project seems mundane compared to some on this thread, but I'm bootstrapping a cycling reference site at http://brightspoke.com/. Our mission is to put more people on bikes by creating informed consumers. We're not quite ready for a "Rate my Statup" post, but you can be sure we'll have one. In fact, I've just been working on a milestone plan for 2010 and that's one of the milestones. http://frogmetrics.com/ - YC summer 2008. In my spare time, researching how to catalyze mass behavior change to improve society. How does one catalyze mass behavior change to improve society? That's the question I'm trying to answer. I see the answer as having two parts: 1. More rigorously define what "improve" means. Beliefs about what an ideal world would be varies drastically between people and cultures. To avoid the "colonize the savages / white man's burden" problem, we need a framework for discussing and evaluating goodness in a more rigorous way, especially one that is viewer independent, cultureless, and timeless. In the 14th century, people lit dogs and cats on fire for amusement. This sort of thing would be greeted with horror by many people today. I wonder what we're doing now that people from the future will view with similar disgust. 2. Identify the behaviors / beliefs / etc that prevent such a world from existing now. It seems likely to me that this will involve formulating a scientifically falsifiable model for how people adopt and change beliefs and behaviors. It's not clear how much precision will be attainable for a model this. It could be that we are only ever able to understand very generally which stimuli or environmental conditions will give a higher probability of shifting individuals towards certain behaviors and traits (eg, empowered analytical altruism). We can then identify the levers that cause this shift to occur, eg altering what children are taught at school. In a nutshell: 1. [if possible] More rigorously define what good world(s) would look like. 2. [if possible] Figure out a model for belief and behavior selection that can be applied to causing / raising people to come to adopt behaviors and beliefs that bring about this better world. Last month my mom emailed me asking if I knew of any new/good places she could take her friend to eat in Philadelphia. I went on Yelp and found a few suggestions for her. Had my mom known about Yelp, she could have done that herself. I went over the next day and showed her Yelp and now she uses it every time she wants to go out to eat. I'm tired of my mom treating technology like it's a chore so I'm working on something that will introduce her to new tech that can really make a difference in her every day life. Like most of you, I'm immersed in the latest and greatest technology every day but she's too intimidated and too busy to discover it herself. I find that once I get her past those two bumps in the road, she actually enjoys the benefits of whatever it is I introduced her to. Right now I'm compiling a list of awesome sites, services, gadgets, etc. that I think my mom would actually like and benefit from... if she only knew they existed. Then I'll send her a weekly email with a 2-5 minute video showing basic use of a new piece of technology and explaining how it can benefit her. My gift to my mom. I'm working on Flixa.tv, a digital distribution platform for independent filmmakers as well as other indie video content. There's a bit more information and a long interrogation of the idea here: http://news.ycombinator.com/item?id=1028673 Still looking for a co-founder, so contact me at chris.chen@flixa.tv if you're interested! Sheet music reader for the iPhone: http://www.wonderwarp.com/opus I have been working on a health\food tracking site www.eat.ly as well as one other unlaunched project. A simple blog project that has been going well is www.multiplayergames.com which is in need of some buddypress\wordpressMU and design help, but since it continues to perform well I am hesitant to change anything. Outside of those a handful of other smaller projects:) How well? I can't stop trying to dev my own PHP framework. I must say that this one is the best by far:
http://backend-php.net I love the community of Hacker News, and I'd like to create something similar but for a South African community:
http://zacoders.net I just finished editing documentation of this outdoor projection / interaction project we did in new zealand: I'm working on an open source eye tracker: and a c++ toolkit for creative coding: that looks like so much fun. nice work! Trying to find ways to attract more US-based mappers to OpenStreetMap. First stop: support my buddy Lars' map rendering to get this beautiful map http://toposm.com/ma/ to cover the entire United States. Anyone out there have some spare CPU cycles and/or memory to donate for the next couple weeks? I am a co-founder of http://ridewithgps.com and we have been intending on integrating OSM support into our route planner, as well as figure out a way to funnel mapped routes back through to you all. We were thinking a cool way to vette cycle routes would be a heat map of sorts. We have some more business related priorities (payed accounts and the like), but this is up there; I'd love to see those map tiles cover the entire US. Wow, that's gorgeous. Something I'd been thinking of working on was a tool to produce really beautiful printed maps from the OSM data. The existing OSM renderers are screen-oriented and fairly utilitarian, and their output isn't nearly as pretty (or as high-resolution) as a good paper map. But the data is all there; it's just a matter of rendering it better. This is definitely in the direction that I was thinking of. Although the holidays clobbered my productivity on it, I've been working on a management app for WoW raiding guilds: http://srsguild.com/ But I did start hacking on a fully email based todo/reminder system over the break that I hope to have functional in another week or so. I am writing a maze generator in Haskell. I'm working (slowly) on an interpreter for a minimal, non-strict, pure untyped lambda calculus. Bonus features include syntactic whitespace, some basic optimizations, simple tracing/debugging features, and support for "compiling" multiple source files into a single program. No, it's not supposed to be useful. Very cool - I've used Tripadvisor for international restaurant recommendations plenty of times. Happy to hear it! Any feedback on the recent redesign is greatly appreciated. I'm doing website user-session/click flow analysis, developing a graph/network visualization library in processing suitable for these purposes. Crunch data in Hadoop, output to interactive graph visualization web app. One step, 'just works.' Hoping to open source this work later this year. I work at Google running open source compliance, code release, outreach and on public sector engineering management. Management, meaning, I don't code much anymore, although I added a totally trivial constant (1729, Hardy Ramanujan's number) to the calculator recently :-) Working on an general game playing artificial intelligence. Nothing grand, I just have a few ideas I'd like to test in real life. It can theoretically play any game, but it's interesting to see if it's actually feasible, since it's turning out to be quite a resource hog. Can you tell me any more about this? I guess my explanation sounded more awesome than it actually is - general game playing has been around for a while now (see for example games.stanford.edu and the wikipedia article on general game playing), mine's a stab at it with some homegrown techniques, largely out of personal interest. As I said, I have no idea if my techniques are feasible or actually work yet, but if they do (and I have the time), the general game playing competition at the 2010 AAAI conference isn't out of the question. Is that happening this year? I bet it involves A*. I recently finished a decentralized content API for my own personal website and for a future community that is planned to be an article publishing portal. It is only a private console with an API, so there isn't much to look at. I tied up a number of loose ends with my own personal website, integrated the API from my decentralized content web application and my google books library. My current project is to finish reading The Little Schemer and doing the exercises alongside it. I've been learning a lot and thoroughly enjoying it. My next programming project will probably be to move my decentralized content API from PHP+MySQL to an Erlang backed key-value store and a Scheme powered content API; I may still use PHP for the web interface. http://www.optask.com - outsourcing (research and admin assistance) at the drop of an email. We're working to open up the world of task-based work, just helped some customers with file conversion and compiling a spreadsheet of apartments on the Upper East Side. Just linked up with our first socially-responsible BPO. Really sweet that we can create fair-paying digital work for folks around the world. Currently also interested in live video streaming (how will this space be impacted by the iSlate? Is there room to innovate beyond Justin?), mobile surveys, eBook readers (Kindle's way too expensive. I want a cheaper, more open product). Will be @ CES, let's link up if you're around. I'm building an open source rails framework to create sub-communities on facebook that are focused on a particular interest, but also providing other views into the app, such as a standard web interface, phone interface, and iphone/android native apps. Facebook connect provides you with some interesting options for taking elements out of facebook and using them in other places, and really allows you to build one application and treat facebook as a view format rather than a completely separate entity. I'm also building in some cool functionality to bring widgets into the app, where you can build out your pages based on what widgets you are interested in, similar to widgets in wordpress admin or igoogle. Working on web based speech recognition. Have a demo at http://www.speechle.com but its buggy. I learned a lot making it but it gets frustrating at times. Decoding is done using Sphinx. Flash's "allow" button is hidden for me (FF3.5, OS X): http://abstractnonsense.com/Screen_shot_2010-01-05_at_1.11.1... Thank you. I was not aware that could happen. Can you please try again? No problem - it works now. I'm working on some mathematical models of the flirting signaling process and some other social phenomena. I'm hoping to get some interesting / nontrivial existence and/or impossibility results with cool qualitative interpretations out of it I am working on making 500 of these: http://www.jgc.org/blog/2008/03/building-temperature-probe-f... for schools in Uruguay. I'm really surprised that the mic input is DC coupled. That's _very_ odd. But then I've not seen the schematics (anyone got a link?), so maybe it is. I'd certainly consider a low power rail-to-rail op-amp and dump the 9V batteries. You can possibly get rid of the pair of 9V batteries by using a floating ground that you make with two resistors. But then what would he use for the voltage source, + and - 4.5V? It appears he wants the + and - 9V available for the amplifier. But my EET is extremely rusty. The 5 V from the USB. Leave the opamp with it's ground connected to the joint between the two resistors, choose the bias so it drives the output from the -2.5 V the opamp sees (which really is the 0 from the computer) to the +2.5V (which is the +5 of the computer). I'm only giving the circuit a cursory look, but I thought he wanted a 9V reference voltage added to the sensor voltage going into R7. But I guess a lower reference voltage may be adequate too. The midpoint voltage of the voltage divider used for the reference is 0.45V, well within the -2.5 to +2.5 swing. The reason why I think it matters is that batteries are a nuisance, they run empty and will cause a problem with long running experiments, so it's worth the extra time to engineer them out of the circuit. From what I can see they're only used to power the op-amp. The 'floating ground' trick is s.o.p. when designing op-amp circuitry that needs to be fed from a single supply. I said reference voltage, but I meant the voltage for the summing amp would need to be 9V if that's the step up range he's going for. I totally agree about eliminating the batteries if possible; I just wasn't sure how to get adequate voltage available to the amp without the 9V batteries. But again, it's been many years since I've even looked at a schematic. The floating ground idea looks extremely cool though. Reading the post again he might be able to completely get rid of the whole circuit by placing two Ge diodes in series with the 0 terminal of the LM35, that will raise the 0 by 0.4 V, just enough to get over the threshold. Wow, now I know where to look the next time I have a hardware project. Nice. Agreed. I am redesigning it for the production version. I'd like to pull everything from the USB if possible. Have you tried simply raising the ground of the LM 35 by placing a diode in series with the 0 terminal? That way you might just add a single 10ct component and get rid of the problem completely. Just enough to raise the ground of the LM35 above the input sensitivity threshold. One Si diode would probably work, two Ge in series would be slightly better (because you get more range). So that would be: What happens to the voltage across the diode when the temperature changes? Looks to me like you've just replaced a calibrated temp sensor with two sensors in series! The temperature of the diode will change right along with the temperature of all the other hardware, it's not as though you'll go out of your way to get the diode heated up. The way it's drawn is not the way it is built, the diode would normally sit near the computer end of things, not right on top of the sensor. The thermal drift of the other circuit is at least as large as that of a diode at room temperature, probably a whole lot larger. http://www.invoicemore.com - online billing and invoicing system (bootstrapped startup) At $15/month for unlimited invoicing, it's currently one of the cheapest on the market. I'm making a phone party line similar to Omegle/ChatRoulette using Twilio. Working on an ultra-light mashup to help track books that I've read. It's my first Rails project, hosted on heroku, mashes up Amazon+Facebook Connect. I'm working on a website with my Mom! - http://www.wordsonthefly.com The aim of the site is to provide tools and tips for communication and writing. She's a veteran in marketing and publishing and I've been in the software field for a couple of years now so we decided to collaborate. The site's pretty basic at the moment, a blog and a tool for templating short pieces of writing, but I am looking forward to upgrading to a VPS and possibly producing some more interactive features. Some great projects in here, I'd love to see a thread like this more often. I'm working on a new eCommerce shopping cart - IONCart. Check everything out at GitHub - http://github.com/leftnode/ION-Cart or on my blog - http://leftnode.com/category/ioncart/ I want to release the minimum viable product as soon as possible, which will be open source, or you can pay for it and get all future commercial releases for free. I'm working as an undergrad research assistant for a distributed systems ploject within UW's CSE department. Been working on open sourcing a PHP framework I created a while back. Uses MVC, but the primary goals are to be lean and schemaless. It pulls columns from the database and assigns them to variables dynamically so you don't have to update the schema and then worry about the database (most frameworks require you to flush and rebuild). More to come in the next week. It can be found here if you're interested: I'm working on my cricket blog Against the Spin (http://againstthespin.com), trying to apply sabermetric-style analyses to cricket. Does anyone know whether this has been done with hockey? http://www.learnivore.com - (programming screencasts aggregator) is my main side project currently. I'm learning a lot (audience building, iphone web app programming) in the process of running it. This year I plan to bootstrap 2 or 3 similarly-sized sites with my wife (including one brick-and-mortar business with a site most likely), and work on a largish project for the pharmaceutical industry in association with one of my customers. Same answer with the old thread. I'm working on a Python IDE for Windows. Recently released an alpha indeed. http://pfaide.com I'm working on http://wtales.com/. Think collaboratively authored choose your own adventure books. So far wtales is just a side project and is just getting off the ground. To kick things off, I am running a kick start program where people can earn amazon gift cards for posting stories (more info on that at http://wtales.com/kick-start-landing/). Hey Sean, I've been musing about automatic story generation / dynamic choose-your-own-adventure games recently. I'd love to chat with you about this stuff if you're interested. You can contact me from my webpage from my profile if you'd like to chat. Groupie An app that allows you to create join and manage social groups on the iPhone. Every group has a message board, live chat room, gps-enabled map and member directory. Thus far we have over 70 groups and 300 users. Many thanks to members of HN that helped make it possible. You know who you are. Url: www.groupie.mobi
Video: www.groupie.mobi/whatisgroupie
App Download: http://www.groupie.mobi/images/Apple_BTN.png I'm working on my dream requirements management tool, it's basically just a big graph visualisation tool http://blissapp.wordpress.com/ Back to the data warehousing day job tomorrow though, so lets hope that the development of Bliss doesn't slide another year... Also, trying to get my head around the ycombinator concept, wondering if the recursive magic that i've seen might be useful for fast graph traversals. Just got official go-ahead from client today to produce a MIS & reporting extranet app written in Perl/Catalyst/DBIx::Class with jQuery and Flash charting. Woo-hoo. Go perl! http://www.UniversityTutor.com Online tutor directory http://BuyersVote.com StackOverflow for product reviews http://FeedmailPro.com Email newsletters for your blog And my personal blog, http://www.StartBreakingFree.com Yes, I have too many projects :) Working on a simple, private, short-term mobile group communication tool called Fast Society. It automatically connects friends who are going to a concert, on vacation, or just out for a night of drinking over SMS and has a ton of cool functions to communicate during and share content after the event. Launching soon at http://www.fastsociety.com/ (still waiting on carrier approval) I just launched a mobile group communication tool recently. A fun yet grueling process. I am eager to see yours in action. Will it be available on the iPhone? feel free to reach out to me to chat. lefstathiou@gmail.com my app is called groupie. www.groupie.mobi Have you considered ad-hoc networking (e.g. iPhone short-range bluetooth)? Besides a small SaaS (that I use internally, but making public to gauge interest), I've been having fun with namespacing Mootools (http://github.com/ericclemmons/mootools-namespace) and using Rhino to generate dependency maps for simple concatenation (which works on most frameworks' dependency scripts so far). mobile coupon site for local merchants. It will be live in two weeks on http://dealbk.com I considered building a similar app - good luck! I'm working on a realtime strategy game in the browser using "full-duplex" AJAX. I started working on it about a year ago but I had to take a considerable timeout after burning out at my former boring job. It's slowly getting somewhere, though. So why in the browser? Because that makes it instantly hackable! No more bad pathfinding or shitty interfaces to put up with! Have a demo? Unfortunately, no. I've been getting the plumbing together; a layer that abstracts the ajax and session stuff so that all I'm left with on the server side is a file descriptor (one per session) to read from and write to. I have some JavaScript code that allows you to select "units" (I made some ugly debugging sprites in Gimp) and move around with them, and scroll around and stuff like that. But I have yet to bridge it with the mostly to-be-implemented server-side logic. The client-side part is just for "dead reckoning" and generally making things flow between game state updates from the server. So both sides will share some logic, and I'm still thinking about how to centralize this. It might take me as much as half a year to get unit movement working across the network; once that is in place, the rest should be relatively easy. Luckily it's all hobby. ;) A lightweight Python implementation of MapReduce. Are you planning on making this open-source or writing about this anywhere? Would especially love to follow your progress if you have a blog. It will be open source (MIT license). I'll definitely post to HN when I make the initial release (should be within the next month). http://www.vocabdojo.com : A website used to study vocabulary for the SATs, GREs, etc. I made a Django app to do this after I discovered that making 900+ flashcards is a huge pain in the neck, and then decided to turn it into something that everyone can use (after switching to Tornado, of course :D) I have the following domains: http://blackhatsystems.com, http://bluehatlabs.com, and http://policyworkbooks.com. Does anyone have ideas for them or want to collaborate? Drop me an email. I've got joytouse.com and notlit.com (I agree, the first one seems like it would sell devices for hedonic pleasures). I'd like to contact you, but your email address isn't public. It needs to go in the 'about' section of your profile, not just 'email'. Scratch that, I got it from WHOIS. A wave killer using node.js, xmpp and websockets How can you kill Wave when nobody knows what it's supposed to be used for? http://www.snapproofing.com
Proof reading service for college students. Also working on my BreezyFAQ app for plugging in searchable Frequently Asked Questions to your site. Site is not live but it works fairly well at http://www.snapproofing.com/faq Hah! I was just saying the other day that there needed to be a Mechanical Turk like site for Proofreading! Working through 'The elements of computing systems'. Just finished testing the cpu and moving on to writing an assembler. Still tinkering with http://www.startupwiki.co.uk And on a server admin tool http://hg.errant.me.uk/eventscripts-xa Cant find another "big" project to engage me :( Perhaps gonna have a go at reinvigorating my blogging platform project. A stateless election web service for more complicated aggregate methods (like Schulze STV): http://vote.cognitivesandbox.com A simple illustration of a possible use for the service: http://www.modernballots.com A novel about what it means to be a novel. My first Mac app: http://www.windowflow.com - it gives you keyboard shortcuts for moving, resizing and tiling app windows. I'm getting ready for a big new release today so if you try and it and it's not quite right for you, give it another go later. It is a place for me to play with python :). Working on my startup SocialBlaze - it helps companies do social media marketing / brand monitoring, http://www.socialblazeapp.com In private beta now; looking to launch a public beta in the next few weeks. Some good feedback from beta testers so I'm excited :) http://www.deptofnumbers.com/nyc/home-sales/new-york-city/ - An analysis of New York's property sales records by borough and neighborhood. I'm trying to build a openly defined price index for residential sales across the city. check out the publicly available figures from radar logic, they aggregate residential real estate transactions in the boroughs and NYC and publish a daily price per square foot. an interesting technology they havent figured out how to capitalize on but relevant to you nonetheless. I'm currently working on a build system. I'm effectively scratching an itch. It's my first serious project, where I actually plan to use it on a regular basis. You can find the repo at http://github.com/sahchandler/buildit :) Putting up boxes in the local area with free stuff to give away in it. This is one of the projects of the SocialBar in Hamburg, Germany. To separate this from my worklife no hacking is involved. http://adzerk.com - first beta customer starting this month. I'm working on my browser-based game (kind of like Kingdom of Loathing, but with cavemen) at http://www.shinyrockhunter.com and a more business-focused product that needs a bit more work before it's a minimally-viable product. I'm working on a graphical model MCMC sampler for doing Bayesian statistics in Haskell. I've been working on a codepad.org clone for .NET languages: http://dotnetpad.net It's been a ton of fun. I do CMS development in my day job and it's nice to build a site that actually does something instead of corporate brochure-ware :) Expanding my startup, http://theplanis.com , out to another hundred users - although I'm facing a difficult decision between that and one which already has a well-paying customer and angel funding on the table. Off-Road Velociraptor Safari HD, an up-scaled version of our most popular game from Blurst.com. Im thinking of building a few simple games using clojure. Im also building some simple tools to help me learn calculus and linear algebra. Man, university sucks, not enough time to hack, and they make you use C++ :D Still better than high school though. http://www.snailpad.com - snail mail for small business and freelancers. I'm hoping to find a way to meet someone from MailChimp, Aweber or other places to see if they'd be interested in integrating. Online payroll system for South Africa. It's my startup. Just launched a while ago. How's it going? Got a few trail users. Busy climbing the search engine rankings. I'm working on an e-learning platform for low-income students in the United States. For fun? --> http://www.wotsummary.com a full summary of the Wheel of Time Series. (I'm working on a couple of design updates (and the summary for TGS) this week) small web app we built over the last few days to generate nice invoices and track if they are overdue or paid. (not finished yet, but open signup and core functionality working) Workflo - A microblogging platform for your company. A "ask hn: rate my app" will be put up soon http://www.buenacarta.com - Yelp for South America, in English and Spanish (ruby on rails, alpha quality, and running on a slow slicehost instance) We're building real-time data services. If you've got interesting data, we want to talk to you ASAP. andrew at timetric dot com. We're based in Clerkenwell, London. I'm a front end guy who's turning to physical products. Building the world's coolest kitchen outfitter... Sign up here if you like: http://gastronautics.com Products like this? Saw this the other day and I had a Pavlovian response. Check out: http://www.sousvidesupreme.com/ Still working on http://www.MyBankTracker.com/ and loving it. Released our iPhone app a couple weeks ago, which runs on MongoDB. turned http://thatpoll.com into a twitter only site to create/answer polls. about to add a feature where if enough people respond to a poll with an answer that isn't included in the original answer set, it automatically gets added. the funnest part is the deployment method with git and capistrano. so fun that now I have to create some deployment method for my real job. winscp'ing php and swf's onto a production server doesn't feel like the best way to go about things. Still working on http://www.RateMyStudentRental.com, trying to improve student rental housing one school at a time. http://www.tellmycell.com - A simple, affordable, do it your self SMS marketing platform targeted at small businesses Assembing the parts for an open source 3D printer, which were 3D printed on an open source 3D printer.
See RepRap.org for general info and www.erikdebruijn.nl for my blog. Twisted Life, a video game written in Flex that uses cellular automata: Remindum (iPhone App, http://7mills.net/rem) -- lets you to create reminders in your Google Calendar fast and easy. I'm working on http://greaterdebater.com a social news and forum website for link-sharing and debate Developing an online platform for independent/unsigned musicians. Functioning as a radio/library, store and ticket application. Looks like some great projects on this page. Right now I'm migrating a Django-based newspaper website to CouchDB. Much nicer for storing articles. And I'm always working on my secret toy programming language. I moved back home and I am working 60 hours a week of manual, back breaking labor :( Saving money to move to the SF Bay Area this spring :) I'm working on my (as of now) experimental debate site: http://argumentclinic.net (see the about page) I'm trying to get together all the right resources so I can start selling my notebook scribbles as full-fledged games one day soon. I'm working on Sup, the email client for nerds. http://sup.rubyforge.org. http://www.headup.com - connecting users with great content as they browse. The Online Slang Dictionary - http://onlineslangdictionary.com A Java library for Tumblr API access with a focus on making it usable for development of a Tumblr Android application. A PHP library built from the ground up for PHP 5.3: "Crowdsourcing" for grassroots activism: http://dreamact.info/ http://www.caterpi.com -- Crowdsourcing freelance translation I'm working on http://www.peekmaps.com/ Lots of fun! A quick toy: http://projectsuggestions.com/ I've created a relational database that can horizontally scale and now I'm putting it to use. Working on Wikispaces. Thinking about how to make it even better for teachers and students. I was recently shown this by a colleague. Will definitely be trying it out with my students soon. http://done.io/ and a new thing called Forrst. Writing a genetic programming system in OCaml to evolve minimal boolean logic circuits. Just wondering, do you ever get users/companies concerned about the security/safety of their documents? quick project/task tracking app. something easy & simple so we can replace the current system we're using (sticky notes and excel ftl...). hoping to open source it. a novel poker-related web app. i say "novel" because i don't know if it will end up being popular, awesome and profitable, or crash and burn, but i think its neat. currently working on SWIX
http://www.swixhq.com We are a social media metrics company and we just entered public beta. Programming a library for interfacing GWT and MongoDB multitasking: customer support, new customer development, writing copy Launching a digital marketing startup. I've had the opportunity to fully immerse myself in the CodeIgniter framework and I love it. I have used it to build a custom CRM and also an internal tool that will allow us to launch new micro sites in 60 seconds, including buying the domain name, parking them on the server, and spinning the content and publishing everything. Very fun. Have learned a lot about linkbuilding and SEO as well. It's interesting to play with different linkbuilding techniques and watch the results happen in just a few days because of how low the competition is in our niches. Am planning on developing a fully automated linkbuilding tool that will require me to wear my gray hat while I work ;-) How does the CRM compare with sugar? I've been looking to replace that for a while. Well I had actually played with using Sugar initially, and while Sugar is nice, it was a little too robust in some areas and lacking in others for our needs. We have a certain sales model that is basically "hit and run" (We move quickly from client to client with short sales loops instead of detailed and long sales processes, just the nature of the product) so we needed something that matched our model and was finely tuned for how we work. I personally feel like CRMs are just like CMS systems for developers. Yes, there are good solutions out there for your initial build, but eventually, it's better to build your own to get exactly what you want. You shouldn't compromise anything for your business in my opinion. I'm making a Home Kitchen Inventory Manager complete with recipe suggestions. still in dev, but just for you: http://fridgereport.com (its hosted on a server in my closet... sorry if it is slow) It is made with ASP.NET MVC Framework and C# I am working on HockeyBias It is a site that covers hockey news using a simple layout a la the drudgereport and protoblogger. It is a startup I unveiled in late November and it is attracting more visitors almost every day!
The diode is 'up side down', you simply use the voltage drop to raise the 0 of the lm35 (like a zener diode), it doesn't know any better and will add the output voltage to the ~0.6 V the diode provides. +
|
|
+---+--+
| lm35 +---- out
+---+--+
|
---
\ /
v
---
|
|
---+-- gnd
