aporem.net

A very good article about how to loose all your intelligent engineers.
The article is rather short but the comments add many more environmental effects that make geeks leave.

The main reasons the author cites are tradition (“we have done this for years, hence it must be right”) and authority (“if someone from above decided this, then it must be sound”).

The comments add two more to the top no-nos: noisy working conditions and interrupting the work (e.g. by meetings).

Bacteria have quite limited possibilities to communicate over distance or processing complex inputs. Nevertheless they achieve complex tasks such as population control or specialisation within a cluster of cells.

The means to overcome the limitations of a single individual is called quorum sensing. Although very little is understood by now, this is usually done by emitting a chemical which can pass through the cell borders, different concentrations trigger different actions in the individuals.

There is a striking similarity with quorums in clusters. But there is more about it than high availability! Quorum sensing in bacteria work much more subtle and do not depend on out of band communications. The fallback rule for bacteria is to do something different, continuing their work and waiting for an opportunity of a new vote. Computer programs on the other hand usually do not have fall back solutions, they just die.

Thus there is a vast field of opportunity in a networked ecosystems: fail gracefully and do/try something different instead. Think of a typical web-shop with a database backend. Under normal circumstances the failure of a single component can be compensated by the other nodes.
But these systems tend to be allergic to deteriorating situations where one node after another goes out of service: suddenly just one database backend server has to handle the load of all front end machines and will probably die while trying.
Now think of a system where one of the web-servers reconfigures itself to help the database server.

This would be real quorum sensing!

Again an article (german) about the future of programming. Though this time a sensible and sound approach to predict the future: no prediction at all.

The author concentrates on the gains of productivity made possible through ever higher levels of abstraction. Both parts of his statement are well understood and agreed upon. Higher abstraction means less repetition (by humans) and thus less room for errors. This results in more free capacity to solve more complex tasks and thus higher productivity.
The process started with programming languages (acutally already before with the use of assembler instructions) and goes on to the heir goal of software factories, programs that build programs suiting the needs of the user without any human intervention.

But this, as the author concludes, is a unlikely future. Software development seldom goes this far. For example, there are still a few COBOL programs in production and most code written nowadays for these systems is generated automatically by programs but these programs are written by software developers.

Thus the future of software development will lie in handling the layers of abstraction, to communicate the algorithms to the computers which is not very different from today’s work.

Nevertheless, most time is, and probably will be, spent on digital housekeeping and not in pondering abstract problems.

Friends of old games, rejoice! There is now an initiative to preserve computer games in the grand archives of the Library of Congress.

Actually, this initiative exists since September 2006, but it only came to my attention via a speech given by Henry Lowood, Curator for History of Science & Technology Collections and Film & Media Collections in the Stanford University Libraries at the Game Developers Conference in San Francisco.
There he proclaimed the ten most important computer games from the past:

• Spacewar! (PDP-1)
• Star Raiders (Atari)
• Zork
• Tetris
• SimCity
• Super Mario Bros. 3
• Civilization I und II
• Doom
• Warcraft-Series
• Sensible World of Soccer

You can agree to this list or not, but as long as it creates enough attention, it serves its purpose: computer games die if no one is playing them. And they die especially fast, if no one can play them. Computer games, like (almost) every software, depend on an ecosystem to be executed properly.

There are two ways to protect the games from digital decay. One is to preserve the originals, including the systems required to run them, which is a task museums naturally take over. The second is to tweak the programs into running on the ever changing bed of modern hardware and operating systems.

The latter method is now more than ever endangered. Not by technology or neglect, but by law. Copyright laws set up to protect the intellectual property of their creators now work against the preservation of those protected ideas. Legislation is, till now, blind to see this, but will eventually have to decide which way to turn. Is the representation of an idea really more important than the idea itself?

For some games, the decision has already been made: there are open source versions of Civilization and Wing Commander – Privateer and uncountable many clones of Tetris.
Especially Tetris will probably never die!

IBM announced the start of a public beta for a virtual Linux environment which allows IBM System p servers to run x86 applications without modifications.

The main incentive is to widen the range of supported applications for the Power platform. Surely, IBM is hoping to gain back some market share for his AIX operating system, once customers are willing to move to System p. (Recently, Qantas ditched Linux for AIX, so things can’t be so bad for IBM’s Unix.)

There is not much information available, even the AVE page is very mysterious about the technologies used. But since the company behind this is Transitive, who has created the Rosetta emulation engine for Apple, this should be more than vapourware.