After a very long time, we finally have our Airmar Weather instrument collecting data. It’s not working 100 percent yet, but we are getting wind speed, wind relative bearing, temperature and barometric pressure–all the fundamentals. Of course, we got it all working at a point in time where the weather here is incredibly calm–calm being a relative term, because as you can see, we are still getting periodic winds over 30 knots, but nothing like a few weeks ago.
We are not lacking in mobile internet capabilities–or so we thought. The last few days have been unexpectedly challenging. I think we have been utterly spoiled by having DSL for the last few months. Our neighbor’s boat had to leave before the one year period of their DSL, so it was convenient that we routed the wire over to our boat and kept the line in use for the remaining months of their year. Before that, the same neighbor spoiled us with wifi access. I can’t really blame our neighbor for us being spoiled, but so it is.
Our arsenal for mobile access has since 2006 included a Treo with truly unlimited EVDO 3g that is tetherable to a computer. We have also recently purchased and tested a Virgin Mobile Overdrive Pro 4g hotspot. Our computers include the new multiseat Linux Shuttle computer, an eMachines laptop, a Raspberry Pi, and a Nokia N810. The N810 is a great always on navigation machine because it includes NOAA charts and has a built in GPS that can run my custom anchor watch program which alarms if our anchor drags. It’s getting a little long in the tooth and we would like to use the Raspberry Pi as its replacement, but we don’t have a functioning chart/navigation application running on the Pi yet. Nevertheless, days before getting underway, the SD card with the entire Pi operating system fragged itself. I spent the last days and hours with the DSL connection rebuilding the OS and installing all of the programs it had previously had. We also have not yet found a good DC powered monitor for the Raspberry Pi.
As we got ready to get underway, we rediscovered that the DC charger for the Nokia was intermittent and unreliable. That meant that we needed to run an inverter to keep the N810 running–especially since its battery is getting very old and doesn’t hold the charge it once did. So, the N810 was sort of marginal. Then, while connecting Brenda’s Emachine laptop to the Overdrive Pro hotspot, the laptop power supply went out–along with the DC 12V-5V power supply for the Overdrive Pro. The Emachine power supply is built onto the mother board of the computer. That left us with one really functional computer–the Shuttle which is energy efficient for a full sized computer, but not exactly a miser either. The Shuttle runs off a 500W inverter that also powers the two monitors for the two user logons.
But there is more. First, in pre-launch tests, the Treo hooked up great to the Shuttle and the Overdrive Pro worked with the Shuttle hooked up directly through the USB cable. Once underway, we discovered that those preliminary successes were statistical flukes.
The Overdrive Pro routinely locked up the computer when connected via the USB cord. The problem seems to be related to the sierra kernel module. We went through dozens of roulette reboots hoping for another success. After hours wasted in reboots, we dug out an old D-link USB dongle. I discovered that I can’t do anything via Network Manager with wifi devices while using Xmonad. I can’t find my notes on how to connect manually. Brenda finds out that it is easy in her Gnome desktop even though just about everything else looks foreign with the update to Gnome3. The connection with a plug in USB Wifi dongle works for about 30 seconds at a go. So Brenda must do the reconnects. That works for me–not her!
We then switched to the Treo and found it impossible to connect. Soon, the Treo would not connect even on the phone. We did dozens of module reloads and reboots and were blocked at one of three points: Treo won’t accept a USB device number, Network Manager won’t recognize Treo even though it has a USB address, and when the first two problems are randomly overcome, the Treo won’t connect with Sprint. We are back to reboot roulette and I hate that! The Sprint help desk worked on the third problem while Brenda was on our back-up pre-pay cell phone and now she can’t make any outgoing phone calls at all and of course no data either. Thanks, Sprint. We have to go to a Sprint store for a hard reset of the phone now. So, the Treo is now only usable for incoming calls and no internet. The Overdrive Pro requires a network restart every 30 seconds. Hardly enough time to Google what the heck is going on, let alone how to fix it. The N810 works and has a great internet connection via the Overdrive Pro, but searching on the web really taxes its tiny brain.
So today, it was off to Fry’s to buy a new laptop for Brenda, a new USB Wifi dongle for the Shuttle, a DC charger for the Nokia N810, and a DC power supply for the Overdrive Pro (and a new 12V battery so that we can raise our anchor without having to start up the Honda EU2000 genset–that’s another story) I feel like we have been through an electrical storm that blew out nearly all of our electronics. I have empathy for cruisers that find themselves in a similar electrical situation while anchored off of some south pacific island and are thousand of miles from a Fry’s store.
The new USB Wifi dongle doesn’t fix the problem with the Shuttle computer. I suspect that there is some power saving feature that is to blame. But, iwconfig shows that the USB device is not using power saving. The next possibility is that the computer OS is responsible for the power saving. I add a blacklist file in /etc/pm/config.d/blacklist with the line: HOOK_BLACKLIST=”wireless” But that doesn’t fix the problem either. Then I found a site that suggests configuring the Overdrive Pro to only accept 802.11g connections. That made a huge improvement. I don’t know why, but allowing the 802.11n mode was a problem. On a roll now, I then found a site that confirms that the direct USB mode with the Overdrive Pro only works when certain modules are loaded in a specific order–hence the randomness. But, they find that the sierra module that I noticed was really malfunctioning isn’t even needed for the Overdrive Pro. So, I blacklisted that module and may try connecting via the USB cord tomorrow because the Wifi connection is not as robust as I would like. And tomorrow it’s off to a Sprint store to see if they can fix the Treo, as well as installing that new windlass battery so that we are ready for re-anchoring during the storm that may hit over the weekend.
These are the creaky Internet Woes…next story will be the creaky body woes that go with getting underway again.
We were thinking that the last couple of months have been colder than last year. Fortunately, we have our wonderful Shipmate stove burning offcuts from our 2007/8 planking project. But is it really colder? I decided to find out. Here is the daily data from the SQL airport which is about 1 mile away.
I have plotted the daily mean temperature reported at SQL and shaded the area between the two lines red if 2012 was warmer or blue if 2012 was colder than 2011. As we can see from the large patches of red, this past Summer was indeed warmer than the previous year. We knew it felt warm, but since we hadn’t been around during the Summer of 2011, we couldn’t compare. Starting in November, however, this year has been much colder as can be seen by the blue shading. But what about daily high temperatures.?
This plot shows the difference in the daily high temperatures. The yellow-reds are where 2012 was warmer and the blue-greens are where 2012 was colder. The line plot has the monthly high temperatures. As we can see, the data looks pretty balanced between higher and lower. So what about the temperature lows?
In this plot, we can see that the summer temperature lows were warmer than in 2011. The monthly lows during the summer were a little warmer, but over the past two months, the monthly low hasn’t been much different from 2011. So, although the monthly low is similar, over the past couple of months, we are having day after day of those lows. That can be seen in all the daily counts that are blue since November. That explains why it feels colder this year.
We can see that this past spring had lower lows, but higher mean temperatures than in 2011. Again, we weren’t here in 2011 and will soon be leaving the mild climate of Redwood City. So we probably won’t get to experience a second Spring here. It will be fun to learn more about other places in the SFBay as well as elsewhere along the Pacific coast.
