I'm sure plenty of you work on projects with no motivation for financial returns, but suppose you could quit your job right now, and work on any project full time. What would it be? Why?
I'd go after fixing the software deployment and packaging issue. Actually getting software on systems is still a PITA.
There's a huge gap between dev and ops here - I'm thinking that building tools that allows a developer to make an ops ready deployable package as easy as pushing to source control would be the ideal. Then tools, that work against source control and can generate packages off any revision/flavor would be great for both continuous testing and eventual distribution.
We've been stuck in the "download the tarball and compile it" mindset for far too long.
>"Suppose you could quit your job right now, and work on any project full time."
Well, I just did the former though I can't really justify the latter. :) However, I mostly quit because my work was again taking over my personal life, so the intention now is the find something a bit more humdrum than yet another poorly managed startup so I have the energy at the end of the day to work on the side project.
Similar to Ivan, games. I want to make a Dwarf Fortress-like game for iPad, with a top-down 2.5D perspective, sprite graphics in a cel-shaded style, and culled down to the core elements I find to be most entertaining (with influences from other games I enjoy). It's a game I want to play, but I'm probably the only person on the planet willing to make it. However I'd also like to raise Dwarf Fortresses's visibility through my work, and get it some additional exposure to ensure DF development can continue. I've also a few concepts for iPhone games that scratch some other itches of mine, but the iPad game I at least have written a basic rendering and pathfinding engine already.
However I need to line up other work for the time being (should only take a few weeks). I'm a perfectionist that prefers the "when it's done" release cycle.
I'd like to create a sprite-based 2D rpg, also heavily influenced by Dwarf Fortress, where terrain, cities, forests, rivers, people, monsters, quests, dialogue, groups, and stories are all generated procedurally. Push a button and an interesting game comes out. That'd be cool.
Well, Dwarf Fortress's scope is... impossible to match. I struggle to think of any game with even similar complexity. Making a Dwarf Fortress clone would take many years, however culling down the feature set (and the depth of each features) to the actual "fun bits" at least makes it possible. So it's not even cutting the breadth of features, but also their depth. No need for complex climate or economy simulations, etc. No need for insane descriptiveness of objects (save for artifacts). It's a lot of complexity that reduces its appeal to a lot of folks (though caters very well to a specific targeted audience). Dwarf Fortress with a touch interface and cute graphics won't magically broaden it's appeal -- the core game needs to be streamlined and no longer be daunting to newcomers.
For example, no massive world at worldgen with full history and legends. Most people don't want to deal with complexity, they just want an embark site to start in and get to work on. It takes a lot of time to create those things that the majority of folks would never notice or appreciate.
There's a lot of satisfaction with gathering/growing food, building basic shops and homes, defending against threats, creating a trade industry, and a bit of dungeoneering. However one of Dwarf Fortress's strength is it presents a lot of competing interests -- where you need to weigh building defenses with industry, etc. Plus the occasional "oh shit" moment where the game attempts to stomp on your sand castle.
From reading about DF, and especially the NYT profile of DF and the creators, I get the strong impression that DF is very much being done the hard way, even ignoring the C++ aspects, and one could cut down the complexity tremendously with good abstractions and things like DSLs.
(Think the Viewpoints OS strategy - with good enough abstractions and DSLs, you can do an OS or languages in very small sourcebases; I'm still in awe of their TCP/IP strategy - parsing the RFC illustrations! http://www.moserware.com/2008/04/towards-moores-law-software... )
"What would it be? "
- Open source Federated P2P socnet.
"Why?"
- Software can be as closed as it wants 'cause we can always drop it for another option anytime we want since the data is still under our control, but imprisoning our digital lives into these walled gardens put up by "for-profit, maximize-shareholder-value-at-all-costs, turn-user-into-product, shove-ads-into-the-user's-stream-and-if-possible-download-a-few-into-their-throats-as-well" companies like Facebook, Twitter and Google, is a recipe for disaster as our data, our friends, our relationships, our entire digital lives, go out of our control in realtime. The situation is really, really bad and to make things worse, the people currently working on this problem are so geeky, so out of touch with the common man's mentality that their technically smart solutions will never ever go mainstream. I'm very passionate about improving the situation and I desire to make social, at minimum, a federated system like email so migration can get a little bit easier when better alternatives pop up down the line or if the user feels the service provider is starting to get a tad too "evil" for his taste.
I specifically see "games" a lot from developers and I always wonder if they have all 5 or 6 talents necessary to pull it off.
I think most developers are interested in the engine development (a known quantity) but when it comes to the art assets (3D models, animations, textures, level design) that seems like a giant unknown to me.
I would enjoy working on a rendering engine, but if I had to model characters and animate them, I think I'd release the game 6 weeks before my 97th birthday.
If you have talent in all the required spots to do this, I'd agree with the other poster... so 2 or 3 proof of concepts in the app store to hone your skills.
1. Game 1, simple terrain traversal game. Get used to open-world rendering.
2. Game 2, character-focused game with stats. Get used to modeling and animation.
3. <some more stuff>
4. Game 4, Skyrim for iOS
Between coding OpenGL and jumping into web dev, I spent 2 years professionally working as a 3D artist. As a kid I worked a lot on video and audio editing, and as a front-end developer I've mastered Photoshop. I have the talent but need the money safety net before leaving everything behind and pursuing a "crazy" idea like this.
Keep in mind I don't know your situation or what idea for a game you have in your head, but are you coming at this from too much of a black-or-white angle?
No you can't create Skyrim iOS edition in your free time, but you could create something much much sillier and simpler right?
Get the ball rolling, the creative juices flowing and a few apps in the respective app stores before quitting and going full-force?
I realize there isn't anything prophetical here; if you are an all-or-nothing type, then that will just be a much harder decision to make and I am hoping you get a chance to make it at some point.
I’d hire a team of 8–10 sharp developers and write a better Photoshop. There’s quite a bit of low hanging fruit in terms of (a) improved image processing that can be done on modern computers that wasn’t possible when Photoshop’s core features were designed 20 years ago, and (b) much more sophisticated, flexible, and intuitive user interfaces. The needs of those photographers who deeply care about bending images to their will have been sadly neglected by the whole industry for 15 years, with most of the recent improvements (e.g. in Lightroom, Aperture) coming on the workflow side, for photographers who need to worry about organizing tens of thousands of photos rather than getting any particular one just right.
"There’s quite a bit of low hanging fruit in terms of (a) improved image processing that can be done on modern computers that wasn’t possible when Photoshop’s core features were designed 20 years ago,"
That's an interesting area of thought. What things today are the way they are, merely because things were the way they were when the canonical instance was produced?
> What things today are the way they are, merely because things were the way they were when the canonical instance was produced?
Almost everything we interact with is completely path-dependent: our mathematical notation and the organization of our mathematical abstractions, our weights and measures, our language, our cultural institutions, &c.
Or when you get down to it, everything about our bodies and ecosystems.
However, at some point conditions may emerge that allow you to leave the trodden path, which I think is where you're looking now. These are not our dad's computers.
Funny enough, I've had a similar idea in my notes.txt for a few months.
"microscholarships, crowdsourced, donorschoose for people"
I suppose students would have a page about themselves and donors would pick students to support based on their circumstances, talents, grades, and personality.
So really it woud be more like Kiva; loaning money to people based on charity. The problem with that is what makes Kiva so rewarding is the quick turnaround times, where you loan someone something and you get part of it back only a few months later to help other people with. That wouldn't work with student loans.
Except that it wouldn't be a loan, rather it'd be a scholarship or fund. The student should not be expected to pay it back, so I guess it would be someone like 'ChipIn' in that regards. If the site were built similarly to KickStarter, where the student posted a video and could post their grades, they could more easily solicit donations from friends and family. The organization could send the money to the university as a grant rather than to the student directly to lower the chance of the system being used for fraud.
Suspension is properly handled by using the right tires. Anything else adds significantly to weight and cost, and detracts from durability and pedaling efficiency. Also, the existing infrastructure for parts and manufacture has to be considered. I agree that it's a great goal, but those machines have been evolving for a long time.
At the risk of sounding cheesy, I would work on something that helps soothe or at least mitigate the destructive effects humanity has on the planet, ecosystems etc. This could be anything from renewable energy, self-sustaining houses, solar powered cars, arcologies, biodegradable or fully recyclable materials and so on. It's a far cry from the web stuff I am working on now. I would definitely need to acquire some new skills first (which would not be an issue if I don't need to worry about money).
Also, education. Things like Khan academy are amazing, but not yet accessible to everyone in the world. Thinking about that, working on internet access for all seems like a pretty noble cause too.
This idea is sort of wild, but auto-generated scripted media. Auto-machinima? I think in the perhaps distant future, even if your favorite tv show is no longer being produced, you will be able to input a script and some source material and have the result be a decently rendered creation that looks and feels like the TV show you're used to.
Imagine downloading a script with very detailed stage and setting direction. A piece of software would exist that would take as input the script and as much source material as you have (seasons of a tv show or a movie) and create the setting, model the characters, synthesize the voices and output the finished product.
I think a lot of the tech to get started with this exists in the fields of game development, video editing, machinima, voice synthesis, etc. If money were no issue, I would work on putting it together. It would probably just be a toy for years before it could generate something watchable, but would be lots of fun.
I kind of hate this type question. I feel like I have been working on my top priority "full time" (every minute of every day) for over a decade and cannot for the life of me figure out a) how to monetize it when money is a huge personal obstacle and b) how to effectively spread the word.
I and my oldest son have a dread disease and are basically well at this point when that is supposed to be impossible. Is the public curious? No. It is incredulous -- as in "you are a teller of tall tales".
Having said that, I did recently quit my job and moved a thousand miles to live on the beach and support myself doing freelance work. When the income is a little less dicey I may try to finally pursue my dream of creating a web comic or similar entertainment. That might actually pay. I very much need the money and I am extremely burned out on getting kicked in the teeth for trying to help people.
I'd learn AI, machine learning, natural language processing, compilers, interpreters, drawing, playing guitar/violin/piano, snowboarding, martial arts, socializing better and effortless, programming/natural languages and whatever else I can come up with.
I'd love to come up with answers to questions others have not yet answered and discover questions no-one else has discovered, yet.
I'd also like to travel the world and get to know as much people as possible.
However I fear that even with unlimited amounts of money the real problem I will face is time which is why I hate procrastinating and even more the fact that I can't seem to stop it.
Inexpensive robotic aerial platforms for film-making and the attendant software that controls the platform as well as the various aspects of post-production, such as match-moving, involving footage shot from same.
I'm currently leading a small project where something like that would be extremely handy.
Copenhagen Suborbitals (copenhagensuborbitals.com) is doing five testlaunches of rockets this summer from the baltic sea. My job is to get live video from the launch to the Internet. There are several interesting problems in this, one is getting good footage of the launch and preparations.
The rockets will launch from a specially built self propelled vessel that sails from Copenhagen to The launch site aacompanied by a navy vessel that will act as support and crew ship. On launch the navy vessel will be around one mile from the launch platform, so getting good footage of the launch and preparations will be difficult.
I've thought that aerial remote robotics with cameras would be a good way of getting some really great shots. It's not easy though; you're operating over water, with one mile to the support vessel, and you'll need to steer the remotes via the Internet since there won't be available space for an operator on the navy ship (there will be a reasonable good internet connection available from the support ship)so you need remotes that can be operated over the Internet and send footage back.
If this sounds interesting, or if you have any ideas let me know, either here or by email. My mail is in my profile.
I just tried to find a link but have failed but, a few months ago, there was a video showing a camera-equipped quadcopter that autonomously tracked a water-skier. Not only was the flight path dictated by the course of the water-skier but the camera oriented continuously to frame the water-skier using a radio-beacon on the water-skier. The guy who built the rig had a tutorial site showing how he built the beacon and receiver.
Beyond that, I haven't much insight into how such a platform would be built but, figuring that out is what the no-worries-about-money time would be spent.
This would take more money than just quit-my-job money, but it I had the money, I would ask around for suggestions (here on HN and in other places) on how to build a new crowdsourced general encyclopedia that would provide a competitive nudge to Wikipedia. (I'd probably try to hire away some of the existing Wikipedia team, depending on what kind of noncompete agreements they are subject to, and would look to bring in a new kind of staff for overall direction of the project.) I was an editor in a few earlier workplaces, so I'm interested in the challenge of how to manage a mostly-volunteer, world-facing general encyclopedia project. Wikipedia sets an amazingly high standard for a project that started out so haphazardly, but here on HN and elsewhere I've seen criticisms of Wikipedia that get me thinking about how to do a similar project better. Of course, different readers define "better" in different ways, and the first challenge would be figuring out what intentional differences from Wikipedia would help a competing project be successful over the long term. The reason I would do this is to enjoy the result of thousands of new, well written articles on a variety of subjects, a resource my children and grandchildren and people all over the world could use for decades to come. The reason I would ask other people for advice on the project is that it would be interesting to hear how to improve upon something that is already free.
Speaking as someone who worked for the Wikimedia Foundation: noncompete agreements? That just doesn't exist in our culture, as far as I know. But your idea seems to overestimate the staff's importance when it comes to writing the encyclopedia.
The community does all the content and most of the policy decisions, by themselves. The staff is there for things like ops, fundraising, legal, bugfixes, research, PR, and projects that require a longer-term vision.
Ian Baker aka Raindrift recently did a cool project to expose how things really work on Wikipedia. People usually underestimate the process.
There is a way to kill wikipedia and it is simple in essence, just build a beautiful encyclopedia by field. Start with a world atlas, with stunning images, then continue with animals, a book on birds, plants, books on flowers. Even greek mythology, history, religion, art.
Something to rival Facebook that puts more control into the hands of users. The "wordpress"/blogging phenomenon came shortly after low-cost hosting. Now VPSs are where web hosts were 10 years ago.
Seems like most people probably wouldn't need more than something like a TonidoPlug to serve content to people on their friend list and, complementing that, to pull content from their friends' TonidoPlugs to make a nice looking news feed. A dirt simple to use, one time $100 box that puts you on the new social network the users control would be nice.
Nice question, some answer are pretty sad though. Particularly the one of binarymax ("PhD in Computer Science"). Money shouldn't be an issue to learn.
If I have a lot of time (not only money). I would certainly work in Artificial Intelligence or Artificial Life. I believe this is where I would be the most efficient in discovering something useful for the rest of the Humanity.
Sadly, even with a PhD in Machine Learning, I lose most of time resolving trivial issues or making trivial softwares.
An ambitious machine learning project that I was not sure would work out in the end. I think with a lot of machine learning projects it's the journey of discovering trends in data which is the most interesting. However, it's only after you start the project and learn what the actual challenges are that you know if the solution is even possible, much less accurate.
I would continue to do what I already do on the side: teaching my girlfriend how to do qa/support/product/design/engineering while working on our side project (dynamic documentation). She gets to learn some marketable skills and understand what my job is like, and I get to spend more time with her.
Personally, I would work on creating a school/open location that trains students to make fun things with code... games, simulations, or anything else that draws people in. I want it to be a place that is interesting so it may spark that fire that leads them to a life of programming.
I would build my own computer. I'm reading 'The Elements of Computing Systems" and I started the exercices : it's really rewarding. Then I'll build a REAL computer with a VHDL programmable chip. Then I'll design my own VM, my own compiler..
Help the greatest possible number of other people do that exact same thing.
(It's very tempting to reply with more multiple answers, in the form of a goal-subgoal tree. So for instance nodes in that tree might be "Help people think more effectively about their life situations", which in turn is connected to "Help people acquire better critical thinking skills", which, given my life history, has the greatest leverage when connected with the node "Help people become smarter about software development". But this kind of tree is dynamic; as you acquire new skills in the pursuit of intermediate nodes you often reprioritize nodes of the same level.)
Probably most interesting Ask HN discussion I have ever seen.
I would work on new algorithms and application of high-performance computing to genetic sequencing, protein folding and, in general, modeling biological processes.
Strong AI has always fascinated me so I'd probably do some research in the field (would likely take a few years of catching up before I can significantly contribute to the field though).
That's what I did after a nice stint at Microsoft, and it's amazing --- working on compilers is even _more_ fun than you'd think!
But, there's always the gloom & doom at the end. Post-doc positions are easy to find, but there are very, very few academic research+teaching positions and competition is quite intense. Still, it's a great ride so far. And the worst-case "build compilers for a big bank" scenario isn't so terrible.
I feel that there are certain aspects of cs that I really want to focus on for my own personal benefit. Sure I could study on my own (and do), but being in a program of pure laser focus for 3 to 5 years would be deeply rewarding for me.
I would most likely study some fusion of cs and art. These days I don't feel like I have enough time to move fast enough. I am trying to get into gpgpu (via CUDA at the moment) and I need to get into hardware. Having the full time to do this, without needing to worry about financials with my current lifestyle, is my dream.
I did a PhD (not in CS) because I wanted to understand a set of scientific problems in depth, do the research, understand setting up experiments, monitoring them, etc. There are other ways, but it takes 2-3 years of semi-structured work to wrap your head around some of these aspects. At the same time, it helps to understand what you like/don't like and try out new methodologies and areas of research.
Not the only way to get there, but a very fun one.
First, my idealistic and self-indulgent project, and second, the project that would assure me I was actually providing value to the world.
1. My self-indulgent dream: an IDE built around text editor integration. Think Eclipse, but constructed as an exoskeleton that Emacs and Vim could slip into and become real IDEs.
Of course my priorities would come to the fore. The core would be written in a concise but statically typed language, plugins could be written in the same language or in a clean, concise scripting language. It would be trivial to run code from within the editor. And it would be nice if one day it generalized into an application framework like Eclipse.
2. My actually socially productive work: interactive educational software. I would construct interactive software specialized for very small units of learning. An app to teach French verb conjugations, an app to teach the concept of electron valence, an app students could use to interact with a writing coach, an app to teach basic programming, and so on. I would try to implement what had been learned from studies about learning, and it would be awesome if I could provide a research platform for education research.
In the end, I hope that the software work on teaching individual subjects would generalize to a set of templates for interactive learning that would let people construct reasonably effective interactive learning tools for many different subjects, with a reasonable investment of effort. Just as a software engineer would look at a problem and construct a solution out of well-known systems, an educator could look at the material and skills they want to teach in a class and construct an interactive course from different parts: spaced repetition to teach certain information, an interactive coaching module and a drill module for medium-complexity skills involving several steps, a submit/teacher review/revision module for writing tasks, and a portfolio module for the semester project. Those are the kinds of general modules that come to mind now, but the aim of constructing completely ad hoc software for a variety of subjects would be to discover novel design patterns for interactive learning.
Custom modules created by teachers, researchers, and hackers could be trivially published, for free or for sale, and incorporated into course-specific modules or explored by self-guiding students. Utopia! (Whew.)
P.S. If I had to show ads to cover support and operations costs, I would only show them to adults, and I would let people pay to turn them off. (Nobody should have to feel guilty for running AdBlock, so I think it's important to let AdBlock users pay extra to turn off the ads they would see.)
"1. My self-indulgent dream: an IDE built around text editor integration. Think Eclipse, but constructed as an exoskeleton that Emacs and Vim could slip into and become real IDEs."
For the idea of Vim within an exoskeleton, you might be interested in these ideas:
I've been working on a variant of the latter idea on and off for about 4 years now. We got some funding (through an NSF grant) but I couldn't maintain full-time development on it. I'm presently working on a project that sprang up after I identified problems in the original design.
I'd love to hear more about your ideas; ping me @gmail if you want to chat.
Ah, but this is if money doesn't matter :-) My job is very challenging, and after eight hours at work, I can't sling code anymore, much less tackle an entirely different set of technologies than I use at work. Also, I don't think I could do justice to the idea without being face-to-face with teachers and students, which would be a full time commitment. I can't put myself in a position where I'll end up disappointing either my employer or a classroom full of kids.
My current job pays well, though, and I plan on retiring at least fifteen years before I'm too old to program, so the "money doesn't matter anymore" scenario is a realistic one. I have a feeling that whenever I retire, educational software will still feel like the right thing to work on.
That is similar to my more radical answer, moon-bots. A small army of tele-operated digging and construction oriented robots sent on a cheap trajectory to the moon. We have the technology to start construction on a colony. At least, I think we do.
Event registration, chip timing software, live/historical race results. The whole package. Right now the pieces exist but they're usually separate pieces and they look and function like 90's software.
I'd try and bring modern look, feel, and technology to these areas.
Strava is a great example of bringing modern web to the cycling/running world if you ask me.
Build a better "Distributed Proofreader" system for Gutenberg like crowd-sourced book digitization projects which can use real-time feedback to reduce the human effort required for the task. I hope that this will make it easier to build PG like repositories for more countries and for more languages. Even those where good OCRs are not available.
The problem of applying machine learning techniques to search problems (I mean "search" in the traditional AI sense that includes theorem proving and planning, not in the Google sense). To me this is the most exciting area of AI research, and it's still very much wide open.
I would keep doing what I already do. I care deeply for the organization I work for, and I have many friends there. But I would engage only in what seems fun and meaningful, and avoid all things that I don't care about. Hey, I seems so fun I would do it for free :)
I'm probably cheating myself here, because what I want to do should be possible right now. If 'money was no concern' I'd finally lose all arguments against
.. learning system level C (embedded kernel/driver stuff)
.. commit full-time on open-source projects that I love
What I have found helps me get over these "I know what I SHOULD do..." humps? If I pick up a book or find a good source of information on the subject and start reading it, I find that it lowers the barrier to entry for me just enough and starts filling my head with enough good ideas that I have to move forward.
You might pick up a good Kernel beginners book (do you know C? If not, start there.) and just start reading it... thumb through it... start seeing some of the API calls and comitting them to memory so the next time you see them you think "Oh yea, I remember that from Chapter 1..."
Don't worry about coding, which IDE to use, which build system to use, installing Ubuntu on your laptop or moving in with Linus... just get that book and start reading before you go to bed at night.
It'll either break down the barrier to entry for you, or make you realize it isn't what you thought and you can focus on something else.
Absolute worse case scenario? You learn a few cool tricks and put the book on your shelf to collect dust. No biggie.
I spent a considerable portion of my teenage years wasting away in front of books like Windows Internals in the hopes that I would be slightly better at writing rootkits and exploits than the next guy. It's a pretty fulfilling skill and meeting someone with real (I stress real) deep kernel knowledge basically never happens, which can give you a reputation for solving the unsolvable, which is both positive and negative. As you may have surmised however, it's effectively commercially useless to have reverse engineering, assembly or kernel mode driver development skills -- That considered, I'd highly recommend you give it a shot, 'Understanding the Linux Kernel' and 'Windows Internals' are some of my favorite technical books. If you have only a rudimentary understanding of memory paging and assembly, I'd recommend 'Reversing: The Art of Reverse Engineering' and 'The art of exploitation' both published by No Starch (Which also publishes two highly recommended, but certainly not my favorite - "Designing BSD Rootkits" and "Art of Assembly").
Open source projects, stuff that's fun and creative. I had a great time with Hecl ( http://www.hecl.org ) for instance, but don't have a lot of time for it these days.
(Another "PhD" response begged the question of "to what end" - in my case, artifact preservation and restoration museum work would be the ultimate dream job. It's totally impractical, though.)
I'd start a foundation that funds research on medical 'open secrets'--treatments that show great potential but can't be patented and hence do not get funding consistent with that potential.
The advancement of artificial intelligence, not sure where to start, but I guess that's the point of this. If I had no concerns about money I would devote most of my free time to it.
Without thinking twice, I would commit myself to working on projects that tackle the main sociopolitical issues that humanity faces today and will face in the long-term future.
Contribute to open source projects. I think they're making a huge difference in the world and deserve more than just a few dollars from my account each month.
There's a huge gap between dev and ops here - I'm thinking that building tools that allows a developer to make an ops ready deployable package as easy as pushing to source control would be the ideal. Then tools, that work against source control and can generate packages off any revision/flavor would be great for both continuous testing and eventual distribution.
We've been stuck in the "download the tarball and compile it" mindset for far too long.