Okay, so here's one idea BruceL and I discussed briefly.

For accessing books, you could have a 3D bookcase that looks something like my low-res rendering below:


then, (I quote Bruce):

Definitely caught my interest when you suggested separating models/views. Are there any mock ups/pseudo code of the implementation that you're thinking of?

I'd like to see more about the separation of code from view, what you think of describing models with enough info to have generic views, views using standard toolkits (custom, GTK, QT, Windows, MacOS), and a million other things

I tried clicking a link you posted earlier, but I think it said the document was $25.

When I think of models, I consider them as a combination of (a) the data that they store and (b) functions that apply to the data. My ideas are heavily biased by Rails and I've never really done GUI app development.

Hopefully I am allowed to post my own paper.

Try http://slypstreem.com/PortableGridEngine.pdf


Its a little out of date and it doesnt go into detail about the modeling language.

Obviously, if I want this to catch on I plan on explaining how to use the language. Im hoping to do that piece wise by example rather than in a technical book.

If anyone has an idea of how I can do that here let me know. After we get a slightly better example in code I will post the software and some examples.

Until then...

The modeling technique involves moving a cursor around an information structure and reading or writing to it. Because it involves moving a cursor complete structures need not be defined, For example, we can have a word that just moves to a color field and then is used for seting, reading or searching over it.

Because of that we shouldnt define words like MountainBike but we should define mountain and bike. The engine can scan around inside the definitions to see if anything can be made to match. So we could also use Big Red Bike to define, search, or describe.


It makes function calls look a little strange because references to the arguments are done by descriptions of them or by their positions (e.g. the first red bike), not by their names. Because these can be resolved early it wont slow processing down much.

If that makes sense to you then you are a better reader than I am! I will try to speed this along.

One last teaser which I hope will sound cool to some people but not scare the others: The operators for moving the cursor around information are similar to the arithmetic =, +, *.

Keep working on ideas people! I did the software but I think people here could make a better UI than I could.

That link worked, thanks I will also openly admit that it's a bit over my head so I skimmed a bit. I'm open to reading more and participating a bit if it all makes sense later.

If you think tablets will be one of your first targets for testing, why don't you create a wiki page about it? I'd suggest (for my own good) you give a few use cases of popular goals for the tablet, like getting directions and mapping a route, reading, playing games, listening to audio, writing a doc. You can include use cases you don't plan on supporting as examples as well. I know that the examples above are complex and might require quite a bit of modeling, so any description about any of them would be useful in getting people interested

Didn't read the paper (yet), but from your post... shrdlu?

I'm withholding serious comment until I do get to read it... Your post makes decent sense from here, though. Not hard to read.

OK, Tomorrow I plan on giving code and software to some people. (I need to finish packaging things.)

At some point I am going to ask for two or three developers who are willing to be a part of this project. If you think you can do that, let me know. The code is in C++ and it is short but dense. (Josiah, you are already on the list.)

But I am also going to give a copy of the executable program to a few people to play with to get the hang of the language and even begin to build UI components.

If you are interested in one of these two things please reply or PM me.

I've decided to GPL the code earlier. It is available at http://slypstreem.com.

Here is what I wrote to Josiah:

Just save it somewhere (on your device or in linux or maybe even Windows), then run

tar -zxvf slypcode.tar.gz

If you then go into quanta/slyp there will be ./slyp.exe which you can run. It's rather simple. The source code is slyp.cpp and there are two files "world.qnt" and "display.qnt" that define what is displayed.

In quanta/quanta there is a file qtest.exe which you can run with ./qtest.exe < tmptst.qnt or ./qtest.exe <tests.qnt. The latter one is some regression tests.
tmptst.qnt is what I use for debugging. More on the format of these later, but you might be able to figure it out.

If you have gcc installed on your tablet (and maybe g++) and make then you should be able to just run "make" in either folder and have it build. You also need to have openGL, sdl, and openGL-dev, sdl-dev, and possible others installed. If you have a problem building, please let me know. If you try to run make and it says something like "device out of space" there is an easy fix.

----
To avoid more delays I have gone back to a slightly older version of the code that will build and run.

Here are some things that need to be done:

Look in slyp.cpp; near the top is a switch statement that attaches the engine to OpenGL. It would be a great help if someone could add a few commands to the switch statement that will call cairo, pango, and glu. I suggest adding just enough commands to will allow basic usage so that improvements will be easier. When that is stable then new ones can be added with more confidence.

This will also be a good introduction to how the thing works. Don't expect too much just yet!

You wrote your own book?! How prolific!

Good on you, mate!

I will be following this thread with great interest!

}:^)~
YARR!

Gapt'n Gorrupt

Okay, so after some struggle I was able to compile and run Bruce's code, which is really rather impressive

Here's the steps I took, hopefully they'll save you a couple hours:
1) install the SDK repository
Add a new catalog with the following information:

2) install gcc from said repository
You should be able to do this with
I accidentally did it the hard manual way, tracking down every individual dependency

3) install libsdl1.2, libsdl1.2-dev from this repository
should work.

However, libsdl1.2-dev complained about xlibs-dev. You can get the binary here
download and dpgk -i it... it complains about a bunch of dependencies but will fetch them from the SDK repo

4) add another repository:
install libgl1-mesa-swx11, libglu1-mesa, libgl1-mesa-swx11-dev, and libglu1-mesa-dev.

5) After this I tried to compile it, but gcc complained about not being able to find SDL_net.h. Strangely enough, all the previous iterations of maemo contain libsdl-net1.2 but not maemo 4. Anyway, I grabbed them from the maemo3.2 repo.

libsdl-net1.2 is here
and libsdl-net1.2-dev is here

download them, then install libsdl-net1.2 first, then libsdl-net1.2-dev

then, you should be able to unpack Bruce's code, then:
And there you go Hopefully this is helpful to you guys! and thanks Bruce!

Some news and a status update:

The news is that I have created a web-site to make the project more open. It's http://infomage.com. I'm really hoping to make this a very big deal so any ideas, feedback, brain-power and so on are much appreciated. I would love to have a couple C++ devs on the project team.

Also, the code-name of the modeling language was "Quanta", but I never really liked it as the final name. I'm announcing the new name is "Proteus" which was suggested by Josiah.

The current status: There are four more things to do on the main engine. My guess is that each will take several weeks to several months depending on how much help I get. After that, there are a few optimizations to do but we should be able to start creating GUI themes right away. If you want more detail, see http://infomage.com/PlanIdeas

Lastly, here is a screen shot from my N800. It's not a great photo, and this is only a test of the code, not a real theme, but here it is.


I'm keeping up with this thread, but I hope to see some people sign up at the new site to be a part of the project, or just to keep in touch!

Bruce