My Blog

If you really want, you can get an RSS feed of Chateau Gladstone news here:

For more information about what I'm interested in, see the Ponding pages.

Weblog installed  30 May 2001 I finally took the plunge and made PHP work. What a nightmare! Anyway, I'm now using weblog to maintain the diary.  more...

24 Nov 2000 We moved the anemometer onto the roof onto the same mast that has the analog anemometer (the Vigilant from Maximum). I also adjusted the rotor inside the anemometer assembly so that both magnets triggered the reed switch. This should (at least) double the reported wind speed, and bring it into line with reality.

2 Nov 2000 I finally got around to adding some new pictures to the website.

13 Aug 2000 I finally added the current temperature to the pondcam. It is all done in a very murky way and is not as general purpose as I would like. The trouble is that anything that happens on a per-frame basis had better be cheap (under 0.1 second of CPU) otherwise it starts to saturate the system. The final approach is that the data collector writes out a file containing the current temperature every 3 seconds, and then it is read in by the camserv code (the current time is substituted) and then added to the image (now uses imlib2).

7 Aug 2000 The problem appears to be in the rain guage connection. More investigation required.

5 Aug 2000 It now seems to be working fine (after routing the 1-wire bus through the surge suppressor). I don't really know why.

4 Aug 2000 The replacement DS9097U arrived from Dallas and the system is now recording temperature and rainfall again. However, it seems much less reliable than before -- it works for a minute or so, and then stops for a minute (with the circuit being reported as shorted). I wonder if some water has got into the wiring (especially the underwater temperature sensor). Tomorrow I will route the circuit through a regular telephone surge suppressor. I guess that I'll try and narrow down the circuit problems.

21 Jul 2000 Dallas Semiconductor are shipping me a new dongle (DS9097U) to drive the instrumentation bus. The general consensus was that the ESD protection did it's job, and absorbed the energy. However, it fails open, so there is now no access the the bus. The replacement will be protected behind a regular telephone line surge suppressor. Who knows if it will help?

19 Jul 2000 I realise that the reason that Linux will not boot is due to the CMOS setting of the graphics card shared aperture now being 8MB rather than the old 2MB. For some reason Linux doesn't detect the amount of memory on the motherboard (over 64MB) and so I had to hardwire it. The ethernet card still works (hooray). It appears that the video capture card is dead -- or maybe the camera is fried. It turns out that the video cable is not connected to the capture card -- I disconnected it when I was trying to figure out what was broken. Reconnecting the cable fixes that. The temperature and rain gauge sensors are not operational -- the instrumentation bus does not detect the presence of any devices. Having said that, it also doesn't detect the presence of a shorted out bus, so I suspect that the interface dongle may be fried. I now have surge suppressors for the video wiring and the instrumentation bus wiring.

18 Jul 2000 A big storm rolls through, and though the promised hail does not appear, there is a nearby lightning strike, and the pond server gets into a funny state. It reports that the CMOS battery is low (which is unlikely). I reset the CMOS, and it no longer will boot Linux. The system is attached to a UPS so I am hopeful that it isn't really fried.

Topic:  All Internet Radio Phone Pond Weather

Rolodex

ffmpeg
Open source audio/video streaming software -- this drives my pondcam