There’s a great Slashdot article on host naming schemes today. It talks about the struggle of smart people who want to give their machines useful and fun names based on consistant naming schemes such as Soviet leaders (Leonid, Yuri, etc.) or French Wines (Pauillac, Pomerol, etc.) versus stupid managment types who want to give them meaningless and confusing names like xyzibm34211, xyzspc25502. It’s loads of fun to read about all the crazy naming systems people have come up with. And I’d never realized there were any RFC’s on creating host naming systems but there are two. RFC1178 gives a lot of useful and interesting guidelines for creating a naming system and RFC2100 is, well, worth reading anyway.
The host names at NCC are all creatures from Science Fiction. Triffid, Horta, Nimon, Rodan, Mothra, Vorta, Ferengi, Krynoid, Vorlon, Zanti, and Wirrin so far with more to come. Eventually we hope to give each host a web page of it’s own explaining the origin of the name. We only have a couple of them up so far as you can see from the links above but more will go up as time allows.
The one problem at NCC is that Randy still names all of his machines wrr[something] – in violation of multiple provisions of RFC1178 it seems. Maybe he’ll read the article and become enlightened.
The 2.3.24 Linux kernel was released yesterday and we definitely have a CPiA breakthrough. No kernel Oops!, no crashes, vidcat will grab multiple frames, we’ve got 320×240 images in color. There are still some bugs but overall it works!. For those who want to see the proof, here’s one of the good images I captured with it:
[[image:good.jpg:Good CPiA image:center:0]]
The biggest remaining bug is if you start capturing images continuously, the images start losing information.
[[image:bad.jpg:Bad CPiA image:center:0]]
It almost looks like the camera is trying to do some sort of motion compression where it’s only delivering pixel data that has changed since the last image. That would explain why most of the background went away but, on the other hand, notice the perl book on the table? It didn’t move at all and it still shows up. Hmmm… maybe it’s just a weird bug in the CPiA driver? If I get time I’ll try to take a look at the code and see what I can figure out.
We’ve been having continued troubles with our new T1. Verio blames Worldcom, Worldcom blames Southwestern Bell. And Southwestern Bell blames Verio. So far none of them have been able to find the actual cause of our problems. SWB switched out our Adtrans smartjack yesterday and Verio swapped out our Cisco 1700 router. So far no downtime but it’s gone for two to three days without problems before so we’ll just have to see. Meanwhile, I’m about to pay a visit to the Dentist so today is off to a wonderful start.
My 1992 Acura seems to be on its last legs (or would that be wheels?). The monthly repair costs are begining to get high enough that buying a new car would be less expensive. The current inventory of problems includes a leaking radiator and a timing belt that has to be replaced soon. I’m really hoping to get one more year out of it so I can wait for the 2001 model cars to come out. That way I won’t have to spend the next few years driving around in last millenium’s model. Besides, a 21st century vehicle has got to be a lot cooler than a 20th century model, don’t you think?
For those following the saga of USB support for the CPiA webcam in the Linux kernel, I just finished testing the latest kernel (2.3.22) with the camera and the only result was a toasted interrupt handler. Maybe 2.3.23 will do it.