I finally wrote some of my ideas down for a coursera.org course: Startup Engineering. This is the result:[quote]
1.) Project Management Social Network:
There are many networks and websites with project management like Kickstarter, Sourceforge and Google Project Hosting.
But one platform for them all would probably be a blockbuster startup.
The facebook of project management where you can
* browse categories
* recruit people
* manage your projects (scrum, todo lists, UML models, documents/spreadsheets etc)
* expose it to the world
* add crowdfunding possibilities
I am surprised that this has not been done properly yet.
Another thing that I am into, rapid skill acquisition, could be a part of that network, too. Think gamedev.net with all the articles.
2.) a) Meta Data Services:
The IT landscape will have to become more abstract and automated in the very near future (exponential thinking).
Everything is connected and the way development works currently won't work in the semi near future.
One thing that bothers me (has been bothering me for years as a Java Programmer) is how applications can't work together.
There are interesting things like PMML and other standards by the W3C but they don't take off.
There should be a pretty low level layer over the internet LAMP/WAMP setups that offer meta data services.
At some point computers will want to understand the world (Siri could be smarter, no?). We should create a model of the whole world that way.
Software should be written so that it can work with evolving models which it can retrieve directly from a UMLMS (UML management system) with UMLQL (UML query language).
b) Abstract Programming:
CASE (computer aided software engineering) software/services could use the Meta Data Services which that layer provides to generate code, build it and possibly deploy it.
At first people would design the software and how the builders turn the abstract software into compilable/deployable code.
At some point that can probably be automated that people won't have to program software anymore.
3.) Shared Intellectual Property World:
There is a lot of money to be made in the entertainment industry. Big franchises have a huge appeal because of that.
Many people want to jump on board with commercial products but they can't legally do that.
I would love to see a social network around an open effort that creates a world made up of shared intellectual property rights.
4.) Entertainment Studio:
See 2.) a) and take a look at procedural content generation. Virtual reality will play a really big role in the near future.
People want to enter epic worlds and the company that offers procedural content generation services will make a lot of money.
Those services can then be used to make movies, games, images and things like that.
With those services there could be a software that helps filmmakers with rapid prototyping for their movies.
As soon as there are databases with textures, animations, models and sounds
(hopefully captured by electronic pets and drones)
the quality will be good enough to create top notch movies in very little time.
I think it should already be possible to take little steps in that direction. Think MakeHuman and a better bridge to Blender.
That would be such a cool playground. Especially once cool things like multibody dynamics are added ...
5.) Fermi Scenario Social Network:
People don't play thoughts nearly often enough. Heard of TVP (The Venus Project)? They complain that they'd need data to take the next steps and start planning.
But according to Enrico Fermi we can estimate pretty much anything.
A social network that collects some data, lets users posts Fermi problems and lets others try to solve them might be fun.
6.) Sound From Scene Graphs
Getting the sounds for animated movies is pretty weird. This is a tough challenge, but solving that problem would be worth a lot.
There should be a ray-tracing equivalent for sound (wave tracing like that: http://blenderartists.org/forum/showthread.php?252708-E-A-R-Evaluation-of-Acoustics-using-Ray-tracin...)
that takes a scene graph and time as input and outputs sound. Something like that will be needed for VR products and would be great for the Entertainment Studio product.[/quote]
Usually I would try to pull off a lean startup with the Shared Intellectual Property World, but webserver administration is totally too much for me.
(Also see my proxying to node.js thread: Mod Rewrite - Node.js).
Another thing I would love to start is writing a series of programming tutorials that show the big picture knowledge (how all the topics are connected, what angles there are to approach the topics and what problems certain technologies and solutions solve).
Guess it would describe different approaches to realizing certain project ideas - especially in the web development and game programming department.
And a site: commonsfiles.com with shared intellectal property and files released under creative commons licenses would be great, too. For example with a subdomain http://ouya.commonsfiles.com that holds conceptart, bitmapfonts, spritesheets and 3D models of characters for the OUYA that anybody may use in their projects.
Eventually I would love to get access to the new media world in some way.