Profile cover photo
Profile photo
Jörg Cassens
Lecturer & Senior Researcher at University of Hildesheim
Lecturer & Senior Researcher at University of Hildesheim


Post has attachment
Tip:  > settings > General > "Who can email you via your G+ profile?" > No one (or your circles). You are welcome.

Post has attachment
Seems lots of people like the new look here at Google+. I do not, and it is not only because I find that the aesthetics leave lots to be desired (or, more bluntly, it looks ugly). Usability has also been sacrificed (see link).

But you can restore at least some sanity, remove some visual clutter and enhance usability: go to settings and tick "Accessibility"-box for screen reader compatibility.

+Instagram has been released for Android, but it is no Android app.

I have played around with it for a bit, and have to say I am less than impressed. It looks and feels like an iOS application and does neither fit with emerging user interaction patterns, nor with the patterns which have been codified by Google with Ice Cream Sandwich.

First thing to notice is the use of a bottom tab bar. This feels just wrong on Android, and it is one of the things to avoid that are listed on Google's "Pure Android"-page on Android design, Tabs should go into the Action Bar on the top, which is most mostly empty besides some search or reload Action.

My Android phone has physical navigation keys, so when I wanted to check the settings, my first reaction was to press the menu button. But that does nothing at all. You will find the settings dialog when you go to the tab with my username and press on what it looks like an "action overflow" button in the Action bar. But it is no "action overflow" button, because then I would expect to be able to open it with the menu button. It only opens the settings page (and a button with the same appearance opens a page where you can block or report users when you are in a different view, which in itself is a major usability blunder).

So now I found the settings menu, but that is also not Android. It uses the iOS-style of having right-pointing carets on lines which dig deeper into the settings hierarchy. You do not use these on Android, it is another iOS-design which has carelessly been "ported" over to Android.

So it is no Android app, it is an iOS-app on Android.

In terms of functionality, well, that can be discussed. Instagram restricts what you can do - use the effects they have predefined, use only rectangular images. The choice to give users less choice can be motivated, and I think it was part of the appeal in the iOS-market. It gives a more consistent feel and makes it easier to produce photos that fit in.

But I am a KDE-user, not a GNOME-user because I value choice (among other reasons). So for me, Instagram feels too restricted.

These kinds of applications live from the user base, and Instagram hit over 1 million downloads in less than 24 hours. So it is probably more likely to encounter someone you know on Instagram than on other mobile photo sharing sites. But I prefer Lightbox, which is a better Android application (although it has some UI-problems as well) and gives me more options. It also lets me upload to flickr and facebook and tweet at the same time.

And if I look at what permissions Instagram and Lightbox ask for, I prefer the one which wants to know less about me and my phone. Instagram loses also in this category.

Post has attachment
Over at boing boing, you can find a very nice article by Glenn Fleishman on Roboto, the new font for Android 4.0 Ice Cream Sandwich:

"Roboto is a bespoke sans-serif font, created by a Google employee and used throughout Android’s user interface (UI) as part of the larger user experience (UX) overhaul. The intent is to make Android more intuitive, cohesive, and fluid, and work better on a variety of screen sizes, especially tablets."

Two more of my students "defended" their bachelor theses today. Some exciting stuff I hope other people will be be able to build upon.

One thesis was about a cloud storage solution where a local client can work both with synchronized data and non-synchronized remote content. Think of a more versatile Dropbox.

The other thesis proposed a solution for location-centric integration of mobile android devices with desktop or laptop computers. You can use your android smartphone or tablet to control a computer playing back media from for example the Network Environment for Multimedia Objects (NEMO) even if both machines are on different networks.

Congratulations again to both new Bachelors of Science.

Two of my students "defended" their theses today (bachelor and master, respectively). Very good work, was fun to work with both of them.
The bachelor thesis was about augmenting paper documents with multimedia elements, the master thesis was an extension to the MATe-backend which should make it easier to integrate new reasoners in the future.

Congratulations to both candidates. Well, I cannot call them candidates any more, I suppose:).

Post has shared content
My students often worry about the fact that the work they actually do for their thesis can not be easily transformed into a linear story as required by the form of a written thesis. I have used the diagram below to explain that this actually is a quite common problem.
In a way, it is not only about "what people think it looks like" and "what it really looks like", but also about "what it really looks like" and "what story I tell about it". Writing a good thesis is not only about designing and implementing cool stuff, it is also about telling a convincing story about the project.
When actually working on the thesis, you will usually have to backpaddle, you will end up in cul-de-sacs, you will have to reconsider decisions. If you take a user-centered perspective, you will have to try things out and modify them according to feedback.
Telling this story, the story of success and failure, is a very interesting one. But for the written thesis, you need to linearize that story, present an after the fact view on what you did. And if you reflect on it, you should see how what you found out in the analysis actually influenced your conceptual design, how the conceptual design is the foundation for your implementation.
The path might not have been straight, but in a well executed thesis, what you did earlier is important for the result. It is close to impossible to tell the whole story on 60-120 pages without making the reader get lost in details, so you have to rephrase it. In the end, part of what your written thesis has to achieve is selling your success story to the reader, and for that to succeed, you have to tell a convincing story.
So do not be worried if reality does not seem to match the format of written theses - it hardly ever did. But look at the process and identify those parts which form a coherent, somewhat linear story - that is the foundation of your written thesis.
The truth behind most success is often messier than it seems. We all do our best to draw a little harmony out of chaos.

Having a huge "What's hot on Google+" box at the top of the incoming stream is pretty annoying. Twitter trends are much easier to ignore for normal use and still discoverable (if one cares). The Google box is just adding noisy clutter to an else still somewhat clean design, too bad.

Post has attachment
Reminder: The online courses in Artificial Intelligence by Sebastian Thrun and Peter Norvig and in Machine Learning by Andrew Ng are still open for enrollment. Both courses look very exciting. Recommended not only for students of mine who have not yet have the chance to take a course in AI or ML.
Machine Learning
Machine Learning

CONTEXT 2011 and MRC 2011 were a success, looking forward to next incarnations already.
Wait while more posts are being loaded