Small liist of usability mistakes. I like point 2, which in my experience is the same as going the (wrong) persona route. Potentially, personas are a self-fulfilling prophecy: designing for a persona and then testing by pretending to be the persona.

Never thought about the scaling problems of Twitter in such detail, but indeed, this is a challenging problem. Looking at the profile of the mentioned example ‘twitters’ they even have more followers…Of course, twitter now has $15M to worry about it…

Wladawsky-Berger hits an excellent point in this article, which investigates the role of R&D labs as ivory towers. His main point is that talented people innovate together across organisational borders in virtual teams on complex company problems.

Building teams is not just about building strong connections inward, but outwards. With the highest quality of information generally hidden somewhere in the world, it is necessary to go outside and find the information. This involves personal flexibility.

This report is interesting for the extensive description about what services are and their relation to science and technology innovation. Especially chapter 2 is worth reading. All other parts deal with the Australian situation.

I’ve seen this presentation at the QCon 2008, however it was not presented by Charles Simonyi then. It is an exciting idea, however there are still some holes in the validation, testing and evolution of the DSL scripts. But it is nice to closely involve domain experts.

Fascinating read about the development of the Connection Machine, one of the earliest parallel computers ever made. Reading history stories like this is really amazing, and to see some of the big names co-operate on the innovation of such a grand idea.

Amazon is not always the cool company it seems to be: They are trying to convert all the printing-on-demand (POD) business to their own service called BookSurge. Publishers that do their own POD will loose the ‘buy’ button from their books. Fishy…

For some personal projects I still use Django. This shows a number of interesting ways of debugging Django applications. Easy!

The NYTimes is at it again: now they used AWS for creating TimesMachine, a collection of full-page image scans of the newspaper. Again they only used a couple of hundred EC2 instances for about 36 hours. Imagine what an investment they saved.

I don’t agree with everything in this article but it provides a number of interesting pointers to different other information on cloud computing. Just realized: the AWS bandwidth can be bigger as it is often used as a backup service.

Microsoft is starting to see that even large enterprises are moving away from ‘roll-your-own’ datacenter to ’software in the cloud’. However, they’re assumption that they can charge more is off: they’re just providing a commodity.

Changing yourself is better than I already thought. It actually can save you from developing Alzheimer and gaining weight. The brain keeps changing when changing your own habits. Not to mention the benefit in creative thinking.

Amazing to see how fast Amazon Web Services bandwidth is growing compared to all other Amazon’s global websites. Too bad there are no real numbers to really see what this implies, but it is impressive even in a relative way.

I just finished reading the book with the same title (Ross Dawson). The book sums up techniques to set up knowledge based relationships with clients. In everything that is described avoiding commoditization is the main goal. Great as a reference guide.

Ok, the subject of his books are a bit weird, but even if he sells just a few (or one) copy of each book this can be a money machine. This is really about the enormously long tail of books, who would order a book about ‘washable scatter rugs in India’.

This article looks at cloud computing from the scalability perspective. Of course, massive scalability comes far from free, however if it can be pulled off, users will have a really high user experience from such systems.

Always cool to read about big corporations that are good at innovation. He states that sometimes it’s good to be frugal (creativity to move out the box) about innovation, where at other times you should go full ahead with as many resources as possible.

Slowly but surely science is moving towards a more open way of working. The main problem still being the way researchers make a promotion. ‘Open science’ can only take of if researchers get their credits differently. (good pointers to other information).

At first I didn’t understand why we would needs Google Gears, Adobe AIR and the likes, aren’t we moving into ‘the cloud’? This shows why: we’re transitioning. Desktop apps and cloudbased apps are not yet consolidated: offline support is a way to do so.

Nice term, a merge of Economics and Evolution, both part of complex systems theory. I bet that looking at it from this side, that something can be said about crowdsourcing and all the web2.0 buzz going around.

This is an amazing story about a man from Poland who totally devoted his life to his understanding of how people learn. The method involves relearning things right at the time you forget them. Of course, aided by computer programs (e.g. SuperMemo).

Unsuprisingly, the open source movement runs on corporate involvement: whenever there are large amounts of money involved, companies join in. Point to take home: open source does not put the corporate world upside down, it is incorporated.

Remember, stupid users do not exist…it might be stupid software they are using.

Quite a high level overview of some of the people who have ideas about security in the cloud and mashup arena. Not much detail, no concrete ideas.

An interview which touches on a variety of topics. The part about multi-core systems have been widely quoted around the web; Knuth has an interesting point of view on this case.

I wasn’t aware of the fact that the GPL didn’t include the notion of SaaS, it only was about distributing code through CD’s and floppies. AGPLv3 solves this loophole, and has been used by a company now.

Probably better to read the original article, but it seems like the author has some interesting insights in DSL development. The thing I do not like about it is that the elegant DSL script is transformed in heavy JBoss code.

Hardware hacking 2.0: an open source hardware platform with hot plugable modules that add interesting functionality. It’s more than just hardware or software, it’s all about the community around it.

This really shows where the world is going with ‘computing in the cloud’. Imagine what it would have costs (or how long it would have taken) to do this without Amazon (too bad that there aren’t many competitors yet…)