Words to work by: Weasel words and the absurdity of corporate speak
...a reminder to give anything you write a decent bullshit test before sending it out. Would you ever talk to your mother or your friend that way? If not, why is it ok to talk to a customer that way? - Matt Puchlerz
By way of Dan Sully.
Now available, the 1.1.4 release of log4r (See: http://log4r.rubyforge.org or gem install -r log4r)
In this release I added code to implement Nested, Mapped and Global Diagnostic Contexts that work the same as they do in Log4j, see NDC/MDC on http://onjava.com/pub/a/onjava/2002/08/07/log4j.html?page=3 for a deeper explanation. Copied concept of GDC (global name for a program) from an older branch of log4r.
Also fixed in 1.1.4:
Logger.new was not thread safe, it is now
RollingFileOutputter created spurious non-numbered files, now fixed
Errors about null at end of yaml configuration files, now fixed
In the interest of having snappy plotting software for use on the ATA software, I'm experimenting with jfreechart and how it updates a dynamic plot, and jfreechart's admitted performance issue:
From section 10.2.2 of the jfreechart 1.0.13 Developer Guide:
Regarding performance, you need to be aware that JFreeChart wasn’t designed speciﬁcally for generating real-time charts. Each time a dataset is updated, the ChartPanel reacts by redrawing the entire chart. Optimisations, such as only drawing the most recently added data point, are
diﬃcult to implement in the general case, even more so given the Graphics2D abstraction (in the Java2D API) employed by JFreeChart. This limits the number of “frames per second” you will be able to achieve with JFreeChart. Whether this will be an issue for you depends on your data, the requirements of your application, and your operating environment.
I was wondering if the drawing was offloaded to opengl (and therefore more on the graphics card), that this updating process might be improved for this admittedly 2D focused work. I'm doing this on my core 2 duo mac pro, ~2007, which has 2 GeForce 7300's in it (not top of the line, but no slouch).
Indirect evidence is the life blood of astronomy as a science. When gauging the distances between our star---the sun---and objects we see in the sky, direct methods only work out to about 300 light years.
By direct measurement, I mean the use of a physical or geometric change (such as the Earth orbiting around the sun once a year) to measure a change in the position of stars as seen from the Earth. Take a picture of the stars in June, then another picture of the same stars in December, compare the results; knowing the distance the Earth has travelled, you can compute distances to stars, gas clouds or dust, directly. The primary limitations are how small are the changes in apparent position of objects in the sky that can be measured (in arc-seconds) and the distance the Earth has traveled (in millions of meters). If you've heard the term, parsec, it is named after this direct method of measurement, and is short for the parallax of one arcsecond.
300 light years is not very far on the scale of what we see in the universe. Our galaxy, The Milky Way, is estimated to be over 100,000 light years across. The nearest large galaxy outside of our own, 2,000,000 light years or over 600,000 parsecs away.