• February 2012
• January 2012
• December 2011
• November 2011
• October 2011
• September 2011
• Complete archive

The NUIGroup Google Summer of Code students (I was lucky enough to become one of them for PyMT this year) are asked to summarize their weekly activities in blog format. Given that the first week has passed I figured I should just quickly outline what I have been working on up to now.

My proposal aims at developing more advanced text input methods for PyMT.

Work on PyMT

Some of the ideas I will realize draw heavily upon spelling correction and suggestion. It is therefore necessary that PyMT can interact with a spelling backend. Given that PyMT should be kept modular, I first implemented an abstract new core provider for spelling suggestions to become independent of a specific library. I then realized two concrete implementations of this provider:

• An enchant spelling backend. This uses the enchant spelling library which can itself be used with different kinds of dictionaries.
• A spelling backend based on OSX’s AppKit spell checker.

After the foundation was laid out I adapted a virtual keyboard with spelling support that Mathieu once developed to the new API and added it to the code base. All of this is not yet finished and needs some more love before I can merge it back into the master branch. You can check the branch I’m currently working on here.

Work on Movid

While spellchecking is important for some of my upcoming widgets, some other text input approaches make use of additional information provided by the tracking application. For example, one idea I had was to split the keyboard in half and dedicate one half to each hand. The halves would then automatically orient themselves following the respective hand’s position and orientation. Theoretically, further information such as properties of the user’s hands (length of fingers, etc.) could be taken into account to lay out the keyboards. For this I obviously need some kind of hand and fingertip tracking. Luckily I implemented that for Movid already:

However, since Movid is still not ready for end users due to a missing calibration utility and a proper (generic!) blob tracker (which means I can’t use it yet either), I continued my work on both of those. Again, both of which are not finished, but I can see the light at the end of the tunnel (or rather, the light below my fingers):

I hope that we can finish all of this and push out a first version of Movid for end users soon. And obviously, I want to test my text input widgets on my multitouch table and not in the mouse simulator.

This concludes my work for week one. If you have any questions or are interested in PyMT or Movid, feel free to join our IRC channel at #pymt and #movid on irc.freenode.net.

From 10th to 12th of June (next month!) there will be an awesome conference in Lille, France for those interested in Multitouch and other HCI topics such as gestural interaction. There is no entrance fee. The conference consists of three parts:

• Workshops, which are basically talks with discussion afterwards
• Showrooms, where people can exhibit their stuff (e.g. multitouch tables)
• BarCamp, a more community-driven place to gather, discuss and share.

We (the PyMT team) have submitted a paper and hope that we can give a talk about our nifty framework. Make sure to check out their website and register (Registration is free. It’s just for the organization to get an idea about the number of participants)!

I will be there to meet people, talk, get inspired and, of course, hack. I’ve recently been in Lille for a similar reason and I can assure you that it’s a nice city. If you’re interested in the topics, you should definitely consider attending!

Hi everyone, I am glad to announce the birth of the Movid project: movid.org

Movid is an acronym; it stands for ‘Modular Open Vision Interaction Daemon’. It’s a cross-platform and Open Source vision tracker, designed to be as modular as possible. Although the project is pretty young, it already features more than 20 modules, including blob and fiducial trackers as well as TUIO output. Movid is coded in C++, and use WOscLIB, cJSON, libevent, libfidtrack, jpeg-8 and XgetOpt.

Movid has several key characteristics:

• Cross-platform: It works under Windows, Linux and MacOSX.
• Daemon: You can run the program without a GUI and control it from another computer over the network.
• Threading: Each module can be run inside a thread. This means that you can finally fully utilize your multi-core processor!
• Remote API: The daemon can be controlled with a JSON API. This also means that you can write your own GUI, e.g. in Flash, and the daemon can be controlled from any application that can make http requests!
• Full HTML5 embedded administration: By default, the daemon acts as a HTTP server. You can control and modify the tracking pipeline in real-time and adjust many parameters.
• Image streaming: Most modules process images. For your application or GUI, you can get the output image via a stream. So your applications can show any image from the piepline or use it for advanced features
• Flexible pipeline: Unlike other applications, Movid allows you to fine-tune your image processing pipeline if you are an expert. You can create new pipelines, add modules/filters and change their parameters in real time.

However, Movid is not ready for users yet, since we are missing a few modules, like calibration. Right now, we are searching developers to support us with the further development.

More info:

• Licence: QPL (Open Source)
• Website: movid.org
• Documentation: movid.org/Documentation/Index
• How to contribute: movid.org/Documentation/HowToContribute
• Source Code: github.com/tito/Movid
• Developers ML: groups.google.com/group/movid-dev
• Movid on techsparked.com: techsparked.com/project/movid
• Vimeo videos: Vimeo group
• Fiducial tracking: vimeo.com/11047041
• Introduction to Movid: vimeo.com/11056184
• (If the videos are not working yet, you can check these mirrors: http://txzone.net/files/projects/movid/screencast-fiducial.ogv http://the-space-station.com/~dennda/gallery/mt/movid.m4v)

The core team currently consists of:

• Christopher Denter
• Mathieu Virbel
• Seth Sandlers
• Thomas Hansen

So, I’ve just handed in my bachelor’s thesis and got some time to blog again. The BSc thesis and the things I’ve been working on the last couple of months are for an extra blog post, though.

Right now I just want to bring a new website to your attention. In case you’re as much interested in emerging technologies such as multitouch as I am, you might definitely want to check out techsparked.com

See you there!

The awesome PyMT library has just been released in version 0.4.

This is a major release that brings a ton of cool new stuff, including a new animation framework, speed & stability improvements and much more. Take a look at the release notes to see what’s new in this release.

I’m using PyMT for my thesis (see picture above) and I love it. Make sure to check the new website, too! (There’s also a new demo video in the works. I will update this posting as soon as it’s available.

I just want to share the following video. Seriously, how awesome is this? (Click this posting’s title if you’re reading via a planet to see the video.)

That thing runs Gentoo.

Source.

For those of you who don’t know it, if something provides Multi-Touch input methods, it means that you (and potentially an almost arbitrary number of other people) can interact with the same device using as many of your fingers as you like.

This technique is relatively new to most of us and I have been blown away when I first saw a video of someone interacting with a so-called Multi-Touch table:

(If you’re reading this posting via a planet or feed reader, please click this posting’s title to see the videos on my blog directly.)

In case you like the python programming language, you might be as excited as I was to know that there’s actually a library that allows you to write multi-touch software yourself. This library is PyMT. It’s based on top of OpenGL and allows you to deal with multi-touch input events in a nicely abstracted way. PyMT is cross-platform, open-source and actively developed. It comes with many examples, a mouse simulator (in case you don’t have such a table) and (in the development branch) support for the new touchpads found in recent macbooks as well as other types of hardware (HP touchsmart, etc.). Here’s an old demo video that shows what PyMT is capable of already:

I was so impressed by what is possible that I started diving into the matter quite some time ago. I’m even building my own table at the moment. The thesis I’m currently working on also relies on PyMT. If you got an appetite, feel free to join us in #pymt on irc.freenode.net or the mailing lists.

In order to show you how easy it can be, here’s a quick demo I just wrote.

If you are interested in building your own hardware (yes, you can), let me suggest you take a look at the excellent NUI community. They have software, forums and even a book available for free for you to learn and explore.

During 2008′s Google Summer of Code I introduced a new storage abstraction layer to the MoinMoin wiki engine. This allows you to run moin on a variety of backends (filesystem, mercurial, etc.).

This year (2009) I participated again. My application contained three major areas of work:

• Reintroducing ACLs for the storage branch.
• Adding fairly advanced ‘routing’ configuration capabilities for your storage backend(s).
• Adding a SQLAlchemy backend which, in theory, should support a variety of RDBMS.

The first two objectives were finished quite well. Two different pieces of middleware were introduced for both, the ACLs and the routing. You can now use several different backends in a simple or optionally fairly complex way. You could just use one single filesystem backend or potentially hundreds of different backends, each with different ACLs applied, and mount each one into its own namespace (you may be familiar with this concept from UNIX, where you can mount discs and such into arbitrary places of the filesystem tree).

In addition to that, ACLs have been refactored. ‘revert’ and ‘delete’ were removed (obviously you will still be able to perform the corresponding actions from the UI), ‘create’ and ‘destroy’ were added.

As for the SQLAlchemy backend, we do have a somewhat working version. Unfortunately it still suffers from severe performance problems. Originally I wanted to fix that after SoC, but real-life (bachelor’s thesis, etc.) caught me, so I will have to postpone that. If you have some SQLAlchemy experience and want to help out here (or with something else), you are very welcome to join us (#moin-dev on irc.freenode.net).

For an overview of what else has been done during GSoC 2009 on the storage side of things, take a look at the following resources:

• http://moinmo.in/Storage2009
• http://hg.moinmo.in/moin/2.0-storage/

Please keep in mind that this is still fairly alpha. Especially the UI is for developers and geeks only and will be redone properly (with Jinja2) before the release (help needed here as well).

Thanks to the whole moin crew and especially my mentor Thomas Waldmann for the nice and fun collaboration!

(Oh and – hopefully – hello planet python!)

We all know that one of the cool things that Python offers is it’s interactive shell that makes it possible to quickly test things. It’s included in every Python release and can be invoked by simply entering ‘python’ on the command-line. While this is great already, the official interactive shell lacks several features that would be convenient (such as code completion, syntax hilighting, pasteservice support…). IPython is an attempt to provide an interactive python shell with the same capabilities, but additional goodies. I was using it mainly because it featured code completion.

I came to the python support channel today to ask about some strange behavior I had observed. Consider the following snippet:

    In [1]: class Foo(object):
    ...:     def __getattr__(self, attr):
    ...:         print attr
    ...:         
    ...:         

    In [2]: foo = Foo()

    In [3]: foo.asd
    asd
    asd

    In [4]: Foo().asd
    asd

Now this should of course print ‘asd’ once in every case. This is IPython’s fault and seems to be related to this bug report. When asking about this, somebody suggested bpython. I headed over to their page and was quite pleased by what I saw in their video. bpython features syntax hilighting, code completion, expected parameters, rewind and pastebin support. See for yourself:

Update: The below no longer applies. A fix has been released. (So you can undo any ugly hack you used in order to make it work.)

Dear Python developers,

when upgrading to Ubuntu 9.04 (Jaunty Jackalope, which is to be released in 22 days from now) please keep in mind that python-virtualenv (the tool we all know and love) is broken. It still works with python2.5, though.

While trying to resolve this issue I had several ideas:

• Deinstall python2.6 so python2.5 is used instead. This turned out to be a bad idea.
• Some may be used to update-alternatives. This can’t be used for python.
• If you just change the symlink /usr/bin/python from /usr/bin/python2.6 to /usr/bin/python2.5 it will first seem to work, but some tasks will fail. E.g. when installing python2.5-dev, you will get the following error: ValueError: /usr/bin/python does not match the python default version. It must be reset to point to python2.6
• You then need to do another ugly workaround.

I am not sure whether this will wreak further havoc, so please be cautious. I hope that we can get this fixed until Jaunty is released. There is a bug report.

Just letting you know. Took me some time to realize.