moonlight42, a web media development company based in Century Square, Cape Town, is looking for a developer to work on their MMORPG games currently in production and on the planning board.

Minimum experience:

• Java programming

Recommended experience:

• J2EE
• Java Applications Servers (Glassfish in particular)
• C#
• JavaScript / jQuery

If you’re interested, you can contact them via their website.

Our postgraduate coordinater, Tanya Ficker, put together a bumper edition of the Faculty of Engineering’s regular newsletter — you can read it here. This is a great overview of our faculty’s research activities, plus some excellent articles on what some of our alumni are up to.

An interesting discussion on Slashdot recently caught my eye: There are allegations that the FBI planted backdoors into the encryption software used by the OpenBSD operating system. Although the lead developer doubts that such backdoors eventually ended up in OpenBSD, code review is still underway.

But hang on, OpenBSD is an open source operating system. Surely if there were back doors built into the encryption software, somebody would have noticed it by now? After all, OpenBSD has been in active open-source development for 15 years, and the claims are that the backdoors were added a decade ago. Furthermore, OpenBSD is highly reputed for its security and correctness of code.

Well, sometimes the best place to hide something is in plain sight. Although no such backdoor has been found yet, that doesn’t mean that it cannot be there. As a very insightful comment in the Slashdot discussion pointed out, hiding nasty stuff in innocuous-looking source code is a bit of a hobby to some, as can be seen in the yearly Underhanded C Contest.

Have a look, for example, at the 2007 winning entries, where the challenge was to “write a short, simple C program that encrypts/decrypts a file, given a password on the command line.” A small fraction of the time, the program should dramatically compromise the strength of the encryption, and make the ciphertext simple or trivial to crack. However, the source code itself must look absolutely innocent.

The winning entries are fascinating: they exploit subtle programming errors that are almost impossible to pick up, and are highly likely to pass more casual review.

Open source is a brilliant model for improving trust in software, because as Eric Raymond put it, “given enough eyeballs, all bugs are shallow” — and, by extension, sneakiness too. But it’s useful to be reminded that this trust should never be absolute.

Although structured academic databases such as Inspec and ISI Web of Knowledge still play an important role in research literature study, over the past few years, Google Scholar has certainly become my first choice for finding academic papers on topics in my field of research. It’s Googlishly quick, cuts across many repositories and databases, and its advanced search features make it easy to filter on date range, specific journals or magazines, and specific authors.

If you do your Google Scholar search from a computer on the Stellenbosch University network, you’ll notice that many of the search results carry a sidebar link entitled “Full-Text @ Stellenbosch”. Clicking on this link takes you directly to the university’s E-journal subscription service, from where you can download the PDF. Gone are the days of poking through the library bookshelves for the right journal, and lugging a heap of bound journals to the photocopier before you can go and read the papers.

Many of us also spend much of our research time on our own computers, away from the university network. Did you know that you can also easily access your institution’s E-journal subscription from your own study? Setting it up is simple enough:

1. Visit Google Scholar, and make sure you are logged in to your Google Account.
2. Click on the Scholar Preferences link on the top right.
3. At the “Library Links” section, enter “Stellenbosch” (our your institution’s name, if you study or work elsewhere) and hit the “Find” button.
4. Tick the box.

You should now see the “Full-Text @ Stellenbosch” link whenever you browse Google Scholar results. When you’re not on the university website, you will have to enter your university login name and password to access the PDFs.

VASTech in Technopark, Stellenbosch has positions open for Software Engineers. Here’s the full ad:

Wanted:  Software Development Engineers

To join VASTech’s Next team

Computer Science , Electronic Engineering or Mathematics graduates with at least a 4 year qualification.

Continue Reading

On 14 September 2010, InnovUS (Stellenbosch University’s tech transfer company) is hosting an Entrepreneur Evening at the Faculty of Engineering’s Reitz Hall. They’ve invited me to give a short talk on academia and entrepreneurship, and why your time at varsity is the perfect time to hatch business ideas — that, after all, is one of the founding principles of the MIH Media Lab, where I spend most of my research time.

The other speakers are Philip Marais, business  developer at InnovUS; David Murray, CEO of Elprom; and Retief Gerber, CEO of NioCAD. It should be pretty interesting, and since it will be flypaper to entrepreneurial-types, it’s probably also a good networking opportunity. Also, there’s free drinks and snacks, which is pretty much what convinced me to participate

More information is available here. Please note that you should RSVP by 3 September.

A recent article which Dr HA Engelbrecht and I wrote for EngineerIT (and subsequently carried by MyBroadband.co.za) led to a few interesting comments in the MyBroadband forums. I thought one of the comments was particularly interesting, since it touched upon the issue of intellectual property (IP) in a research environment — specifically in the MIH Media Lab, where I spend most of my research time. The Media Lab is an interesting case with regard to IP, since we operate under a so-called “full-cost” research contract. In such an arrangement, the University agrees to transfer certain IP that would otherwise have belonged to it, to a sponsor who agrees to cover the full cost of the research. And by full cost, we mean the whole caboodle: equipment, bursaries, staff time, floor space; even “indirect” costs such as sanitation, cleaning and access to library facilities. What’s important to note is that the IP is usually well specified (typically before the actual research starts) and that the University in the first place only has claim to very specific IP.

For any creative industry to flourish, it must be possible for clients to commission creative outputs, with the knowledge that they will be able to own the output that they have paid to be created. If you commission a portrait from an artist, you would like to know beforehand that you will own the painting, once finished. Similarly, such IP transfer arrangements exist between universities and their students, and between sponsors and universities. Similar arrangements are found between consulting engineers and their clients, or software companies, their employees and their clients. I therefore believe it is very important for anyone involved in a creative industry (including engineering and programming) to be aware of the IP agreements that affect them.

The following post of mine in the MyBroadband forum touches on this point, and I thought it might be worthwhile reposting it here, because it is an issue that touches most people working/studying at a university:

Continue Reading

A colleague points out this quite excellent and accurate illustrated guide to a PhD. This is definitely worth contemplating if you consider doing (or have just started) your doctorate research. And if you have a doctorate, the last picture should help you keep things in perspective

An alumnus of Stellenbosch University’s DSP laboratory, Dr Charl Botha, who is currently at TU Delft, has a PhD position in Medical Visualisation (Surgical Planning with Adaptive Instruments) under his supervision. The project summary states:

The goal of the research project is to develop a pre-operative planning system for shoulder replacement that is tightly coupled with adaptive, patient-specific instruments for intra-operative guidance. This accurate and easy-to-use system will improve the outcome of total shoulder replacement, thereby improving the quality of life of patients as well as reducing costs by decreasing the number of revision operations.

The work will be done in collaboration with the hospital in Leiden. If you’re interested in applying, visit the vacancy posting, or mail Dr Botha for more information. Of course, you need an MSc in an appropriate discipline to be considered.

@albert_swart pointed me to this interesting claim that Apple is faking the visuals in one of its ads promoting the video experience on the new iPad. The gist of the claim is that the ad shows a shot from the latest Star Trek movie on the iPad screen (which has a 4:3 aspect ratio, similar to older TVs and computer monitors). The post shows how visual details differ on the promotional visuals, compared to the original widescreen shot from the film. At first glance, it seems as if the shot from the original movie was “photoshopped” in order to fit onto the device’s 4:3 screen and still look visually pleasing.

It’s certainly not impossible that the ad’s visuals were produced by talented graphic designers, rather than by depicting actual shots of an iPad playing Star Trek. But the claim’s argument is not convincing — a lot of research is currently going into the field of smart video resizing. I first came across this field at SIGGRAPH Asia last year. The objective is to find ways of post-processing widescreen videos so that the relevant visuals (usually the foreground) are visible in the entire screen (preserving their original shapes), while parts of the less relevant visual content (usually the background) are removed. The papers that I’ve seen opt to modify the background by compressing parts of the image horizontally, thus keeping the entire background in the frame (albeit somewhat “squashed”). A good example of this is demonstrated in the YouTube video of the paper “Motion-Aware Temporal Coherence for Video Resizing“. @albert_swart also subsequently pointed out this paper (with a video) that documents something similar. An alternative approach would be to preserve the aspect ratio of the background (not “squashing” it), but cutting off some of the edges, while compressing empty space in the foreground. This would produce the effect seen in the iPad advertisement.

This type of processing is not something that happens on the viewing device itself (i.e. it’s most likely not an iPad “feature”) — it’s a form of post-production editing that automatically converts the original movie to a movie file with an aspect ratio that allows it to be viewed on other devices. If the ad is not fake, it’s probably just depicting a movie that was processed in this way some time in the past. This does not prove that the ad is not a bit of clever photoshopping, but one should take any such claims with a pinch of salt. In signal processing, seeing does not necessarily imply believing any more.