frethop | # 1 | 2009-02-17, 15:02 | Report

I'm a university professor. I want to engage some students in research questions that look at topics surrounding tablets and Maemo. I would be very interested in ideas of what this research could be.

I'm defining "research" as investigation into unanswered questions or innovative solutions to problems involving, in this case, Internet tablets and/or Maemo development.

I don't think that software development by itself qualifies as research, especially if it's an implementation of a developed idea. So things like Maemo projects in the Google Summer of Code don't qualify in my view (at the moment). But maybe this is too narrow a view.

Any ideas here would help me out a lot. Thanks!

timoph | # 2 | 2009-02-17, 15:31 | Report

How about researching runtime effiency of GTK+/Hildon software compared to applications written in Qt, python, etc.

At least I'm interested in how much it costs to write SW with an alternative toolkit/language.

frethop | # 3 | 2009-02-17, 15:48 | Report

This is a cool idea. Also interesting might be the relative efficiency of scripting languages vs compiled applications. Or in how much the various speed-up methods enhance Python performance.

paulkoan | # 4 | 2009-02-17, 23:37 | Report

Input is the one.

Pretty must all small form factor devices suffer from the same problem - not being able to get data into the device at the same speed and comfort that you would a laptop or desktop device.

The success of the keyboard (even in the inefficient qwerty configuration) is unparalleled.

The N810 and similar get around data entry problems by actually adding a keyboard. This is a reasonable workaround, even if you are essentially limited by only realistically being able to use your thumbs. But it is still slow going and not exactly comfortable for long periods.

For devices that don't have a keyboard, you have various approaches that boil down to the same type of entry. You have an on-screen keyboard, or handwriting recognition. The keyboards may be enhanced (such as fitaly) or include various swipe guestures, and both types would include some predictive features to guess the word you are using.

But all of them - to one degree or another - fail to manage the separation of "how you enter data" from "what data you are entering" in the same way a keyboard does.

With only a basic level of proficiency with a keyboard, you can quickly get to the point where you are only peripherally aware that the keyboard is the thing making the words form on your screen.

With non-keyboard - screen based - input methods, the thought process of "what you are entering" is regularly interrupted with the thought process of "how you are entering". This may be to choose a word from a list of suggested words, or to correct a handwriting recognition errors for example. They all break flow.

Conversely, a keying error with a keyboard will appear to correct itself because you can maintain the separation, and not break flow.

I think there should be a shed-load more research into the area of small form-factor data input.

gpmartinson | # 5 | 2009-02-18, 01:45 | Report

Perhaps a study of how people use their tablets versus computers. What sorts of activities do people engage in, duration of theser activities, average time involved in engaging with the tablet, etc.

If what you are looking for is a more quantifiable study then research into the efficiency of the internal memory versus the memory card as a location of the operating system. Perhaps a study of the value of swap in the system the value of increasing it. There was a thread about a gentoo based distro for the tablet, maybe a study of the eficiency, speed and overall performance of the gentoo based distro versus a debian based system.

I am suggesting research that might be of value to the Maemo community, much as Timoph did.

lardman | # 6 | 2009-02-18, 11:33 | Report

Investigate the man-hours required to reverse engineer closed source components, such as PowerVR drivers

Stskeeps | # 7 | 2009-02-18, 11:55 | Report

Study on repurposability of tablets:

The ability to change software and firmware, even run other Linux OS'es such as Ubuntu, Gentoo and Android - because the tablet is not locked like iPhone, G1, etc.

To be able to re-purpose the tablets when they are technologically obsolete and unsupported by vendor, to have other purpose than being a mobile device (think secondary screen to your PC; picture frame; etc) - is it more environmentally friendly? Comparison to communities where you have to jailbreak or get developer key (OLPC, G1, iPhone, phones) to alter your operating system.. and ones you do not.

---

Last edited by Stskeeps; 2009-02-18 at 12:00.

frethop | # 8 | 2009-02-18, 12:05 | Report

These are all great ideas. Keep 'em coming.

Obviously, I want the Maemo/ITT community to benefit. I'm wondering if work should benefit Linux-at-large and just be focused on the Maemo platform as a case study.

iamthewalrus | # 9 | 2009-02-18, 19:09 | Report

How about look at the community and open source aspects of Maemo compared to Moblin/Android/ and how well these development strategies/communities work. Although that is more about sociology and economics than computer science.

---

Last edited by iamthewalrus; 2009-02-18 at 19:25.