Archive for March, 2010

Cost of the machine

Wednesday, March 31st, 2010

The first PC system I bought was a TRS-80 Model I for about $3k in 1979. That was a Z80 based 64k machine. Then there was the Gateway 386 with 1 MB of memory for about the same price in about 1990. I would be hard pressed to figure out how to spend that much money on a machine nowadays.

Technologizer takes a look at Classic PCs vs. New PCs: Their True Cost to demonstrate just what has happened to make these machines so much of a common appliance.

What is hidden in the look at personal computing systems is how the cost effectiveness of computing power has infiltrated so many established technologies and enabled so many new ones. Whether it is automobiles or cell phones, it is hard to look around at the many things in our lives without seeing some rather significant computing power that has made those things less expensive yet more capable.

And still some folks moan about the “good ol’ days” when life was rosy and all that.

Posted in Hardware | 1 Comment »

Finding the right place for the subsystems

Friday, March 26th, 2010

Test equipment has an input, a processing component, and a display. Modern computing technologies have complicated the question about how these should be integrated. From standalone devices and chart recorders the trend is towards segregating the functions of data collection, processing that data for measurement, and display of the measurement.

Electronics Lab describes an interesting example in Screenscope: a standalone oscilloscope – just connect a monitor. Oscilloscopes provide an interesting example because of the potential for complex processing and the need for a graphic display. Ancient technology just collected a signal, scaled it, and shot it onto a CRT for display. Modern equivalents digitize the signal, subject it to all sorts of analysis, and then create a graphical presentation. The range runs from the standalone oscilloscope to the software that uses a sound card on a common PC. In between are the data collectors that feed software on a PC and the Screenscope that only uses the display.

The key for the Screenscope is that it bundles its software with its input and only uses a video display as a ‘shared’ component. This heads towards the standalone format with only one expensive subsystem left for the user to define. This puts the software component as a commodity and reduces user complexity and constraints significantly.

The next step is one most often seen in video monitoring systems. Those present a video camera as a node on a network that can be used via a standard web browser interface. Geographic information systems are another example in this vein where personal navigation devices can provide not only a map display but also provide position information to an upstream computer. The possibilities are enabled by inexpensive processing capabilities that can not only provide the intended measurement but also present that via net services. They need to be able to provide both the manufacturer’s intended function with built in software but also to allow access to the collected input in near real time so that users or third parties can leverage the equipment via standard interfaces and protocols for additional or custom processing and use.

As with video, there are many ideas being tried to determine protocols and procedures as well as presentation and subsystems. The access to various subsystems is often obscured due to concerns about intellectual property protection. Progress is being made but there remains a long road to go. The goal is to make it easy for users to be able to choose their level of involvement and customization and to move the level of expertise for this from internal development engineer with access to proprietary information to that of the informed amatuer.

Posted in Tools | No Comments »

magic formulas for Firefox plugins

Friday, March 12th, 2010

First it was necessary to update to Firefox 3.6 to get the Flash plugin to work with the Ubuntu Karmic 9.10 64 bit system. That required adding a repository to apt’s list and applying a magic formula

In order to get that version of Firefox enabled, I had to remove the older version. Then, OK, Flash works.

But Java doesn’t.

The Java plugin got lost on the Firefox upgrade and a reinstall didn’t do the job. – search the net. find Allan Willems Joergensen who suggests
sudo update-alternatives --install /usr/lib/mozilla/plugins/libjavaplugin.so mozilla-javaplugin.so /usr/lib/jvm/java-6-sun/jre/lib/amd64/libnpjp2.so 50

These kinds of suggestions are one way to find out about some nifty things in the OS. Doing a ‘man update-alternatives’ brings up an interesting description about Debian’s method for handling multiple alternative resources with a common name for different needs. In this case, update-alternatives set up the proper plugin file as the Firefox 3.6 java plugin by putting links with the proper names in the proper places in a way that the system could keep track and maintain them.

Now it works. Flash and Java. That means time for an upgrade must be getting near as too many things are working as designed and it is time for a whole new set of puzzles.

Posted in Systems | 1 Comment »

GPX – the G isn’t for Garmin- link references

Sunday, March 7th, 2010

Topographic seems to be source for this common file format used by many geographic information system devices and applications. The current version has been around since 9 August 2004.

The GPX file format is a lightweight XML data format. This means it is a text file with special tags to identify a known set of fields common to GIS systems like latitude, longitude, elevation, description, and name. An example can be seen at Garmin Mass Storage Mode Devices which describes how Garmin uses the format for its GPS devices. GPSBabel describes some of the other extensions for the format.

This is the most capable and expressive of all the file formats supplied. It is described at topografix.com and is supported by EasyGPS, ExpertGPS, and many other programs described at topografix.com

GPSBabel’s reader of this module attempts to preserve tags it doesn’t really understand.

Garmin has a page that describes several of the GIS oriented file formats they support at Garmin Formats. These include AXM, Device XML, GPX, GPXX, KML, and TCX with links to the various schema definitions.

A nice history of the file format is at Travel By GPS The History of eXchanging GPs Data.

The backup is to use CSV files. These are also text but do not have explicit fields and are flat files whose data needs some interpretation. Each line is an individual record. A GPX file has defined fields made explicit and each field is of a know type. A GPX data record may be more complex than CSV formats would support.

The key issues at present is how to fill a GPX record from CSV sources and then see how the fields in the source data end up on the target device. Once that is figured out, it would be possible to figure out how to parse and clean the data source for desired results in the GPS device.

Posted in GPS-GIS | 1 Comment »

Cockpit at the front?

Thursday, March 4th, 2010

The picture Reynolds provided of the new Honda jet got me thinking. Have you noticed how sci-fi continues the trend of having the control center for a vehicle up front and exposed? Naval ships started with the control station exposed on the quarterdeck. In Hornblower’s time, the captain was expendable and expected to be right out there in the midst of the battle. Things have changed. Modern navies have control rooms in protected spaces below decks. Even cargo ships are creating secured spaces in engine control rooms as a means to deal with the pirates off Somalia.

But cars and planes still place the cockpit or control center up front so the pilot or driver can actively steer the vehicle along a safe route. The need for that is changing. Experiments have shown that an airplane can be safely flown when the pilots have no access to exterior windows. Automatic parking has started to show up in automobiles to facilitate hands-off parallel parking.

That view up front is prime real estate. Even in airplanes such as the 747 that have multiple decks, the passenger deck stops short of the nose to give way to radar and other instrumentation. This paradigm follows to sci-fi where rocket ships have the control cabin in the nose. It seems that the control center needs to move to a more protected area, especially in space ships. That can increase security as well as leaving prime real estate for passengers.

It is the passengers who want to see with their eyes. The pilot, navigator, captain, helmsman and other control officers can see better with other means. If fact, if they turn into robots, they may not even have their eyesight close to their brain. Then only thing you’ll need a human for is to hit the off switch if something goes awry.

Posted in Uncategorized | No Comments »