This news page is now used mainly as a record of modifications made to the possums boxes and monitoring hardware/software.
For the latest news on the possums and other more interesting stuff, please see the Possum TV Blog .
3rd March 2021
A few bees have tried to reclaim Box 2 in a half-hearted sort of way, but this time propping the roof open for a bit and then dropping in some camphor blocks was enough to deter them.
2nd March 2021
Box 2 was put back in the tree at about 4 pm.
At about 5:30 pm a big storm with very heavy rain hit. A tree branch fell across the power lines and power was out for over 5 hours (5:39 pm to 10:52 pm to be precise). There was various damage done by the storm, but all of the possum boxes survived and no possums were hurt. On reviewing the footage afterwards, it appears that Box 4 was knocked around a little by tree branches during the storm (it's attached to the house and branches go very close to it), but no damage was done to it. I'm surprised that the footage was actually saved correctly, because the server must have been recording it when the power went off. The server is running Linux with an ext4 file system, so maybe it recovered the file automatically on power-up — I don't know much about how this works, but I've heard that ext4 is pretty good and it's coped flawlessly with quite a few power outages so far.
27th February 2021
Some bees recently took over Box 2.
Propping open the roof (by use of a very long pole) didn't seem to deter them, nor did the fumes of some camphor blocks, and I had to resort to flooding the box with water.
The box has now been removed from the tree and is drying out. There are traces of wax on the ceiling of the box where the bees had started building their honeycomb and I will remove this. There is some theory — I'm not sure how true it is — that the scent of the wax will attract more bees in future.
25th February 2021
Viewing streaming video over SSL (i.e. using https://possum.tv/boxes.html instead of http://possum.tv/boxes.html) may now not work on some browsers. It turns out that Chrome 81 (which is used by a number of browsers in addition to Google Chrome) now bans "mixed content". That is to say, if a page is accessed by HTTPS, anything on the page which is HTTP and not HTTPS gets blocked.
Unfortunately, the streaming video is provided by a program called "Motion", and the version I'm using does not support HTTPS.
I can possibly upgrade to a newer version of Motion, but that is likely to be a painful process which will involve down-time on the server. Worse, it might possibly (the documentation is unclear) force you to always uses HTTPS for the streaming video, which would create problems for people whose browsers don't support the right versions of SSL/TLS.
For the time being, if you find streaming video isn't displaying, your options are (in decreasing order of desirability):
- Access the site via HTTP and not HTTPS (i.e. use http://possum.tv)
- Switch to a different browser (Firefox does not currently have this problem)
- Switch to "1 Sec" update rate, which doesn't use streaming video.
15th December 2020
The temperature sensor in Box 3 was fixed about a week ago; the problem was a broken wire. I was unable to clean the camera lens any better than what you can see at the moment and I didn't consider the problem bad enough to replace the lens altogether.
I have also added some filter capacitors to the temperature sensors. If you have been looking at the box temperature graphs, you might have noticed that there has been noticeable noise on Boxes 2 and 7 lately (even worse than normal) and this seems to have been caused by recent wet weather. I think this problem was caused by me using unshielded Ethernet cable for the sensors and not shielded twisted pair. Anyway, replacing all of the cable is too difficult and expensive, so instead I added some capacitors to filter out the noise, and these seemed to have worked. Box 1 temperature was lost for several hours due to what appeared to be a defective capacitor, but this has now been fixed.
Also, I have changed some of the SSL settings for this site to improve security. This has included removing some weak ciphers and adding an intermediate certificate. The latter was to placate some browsers which were (incorrectly) complaining about the certificate not being issued by a trusted authority.
The main visible effect of this is that if you have been viewing this site using HTTPS (i.e. using the link https://possum.tv instead of http://possum.tv) and been getting a scary (but bogus) warning from your browser claiming that this site was "untrustworthy" — which I think was the case with at least some Android phones — this warning should now no longer be present.
It's also possible that if you are viewing this site using HTTPS with a very old browser, it might no longer be able to connect. If you are having any problems please let me know by , or leave a message on the Blog.
1st December 2020
There has been some very sad news; Rahmet has been found dead. I have written more about this on the Blog.
The temperature sensor in Box 3 has been playing up and also I think the camera lens in this box needs cleaning or possibly replacing, but this will have to wait a bit. Pax, who is now without her mother, has been using boxes 1, 3 and 4, and I don't want to interfere with any of these boxes in case she can tell I've been there and stays away. I'm probably being overly cautious, but Pax needs all of the help that she can get at the moment.
26th November 2020
The lens on Camera 4 had got some gecko poo (I think) stuck to it and had been giving an annoyingly fuzzy picture. Until today, I had been unable to get to it because Rahmet and Pax had been sleeping in the box every day. But today they slept somewhere else and at about 10 am I replaced the lens with a spare. I'm not sure how well the new lens is focused, but it's at least better than it was before.
11th November 2020
At about 3 pm today I swapped over the server from my old cable Internet service to the new NBN service, which uses the same physical cable but is more expensive and gives me half the speed. I was forced into this because my land-line phone has degraded to the point where it no longer works and will be disconnected entirely next year.
Anyway, there would have been an interruption while I was changing the modems over and while the new IP address was propagating, but everything should have been back to normal by 4 pm. It is possible there may be a few short outages over the next few days if I have to reconfigure something.
18th October 2020
Box 2 is now back up in the tree after refurbishment. The roof has been entirely replaced with a larger, stronger one which is reinforced with aluminium angle around the edges. The box has been cleaned and repainted and a new, larger, electrical junction box has been added to make it easier to deal with the cabling.
I haven't made any significant changes inside the box. I had thought to add some climbing branches like in Box 1, but these would have taken up too much room in this small box, and these are mainly for the benefit of baby possums - but mothers with babies probably won't be using this box anyway. I had also thought to reposition the camera further away from the entrance hole, but it was only possible to move it a small amount which probably doesn't make any difference. I had vaguely considered adding a fan to this box like in box 1, but this would have been difficult and I considered it wasn't worth the effort until the box 1 fan has actually shown itself to be useful.
The mounting board behind the box is in poor condition and in retrospect I probably should have been replaced when I had the box down, but this would be have been quite a bit of extra work and I felt the priority was to get the box back up in the tree. I did put a couple of extra screws through the mounting board into the tree to hold it more securely and I think it should hold up this relatively light box fine for a while yet.
At 12:30 pm the box camera was turned back on and the box is now ready for possums.
15th October 2020
At 10 am Box 2 was removed from the tree for refurbishment.
In July the camera in this box was replaced and in the process some rotten wood on the fascia shading the entrance hole was removed. After this was done, possums stopped using the box and my hypothesis is that wood I removed had in fact been performing a useful function.
I had considered simply replacing the missing wood, but other parts of the roof are also a bit rotted and the box itself needs repainting. The box is 15 years old and was last refurbished in 2013, and it's showing its age.
So I've made a new roof for the box and am fixing up other minor issues. The box will be repainted and I might move the camera position to make the camera lens a little less vulnerable to possums entering and leaving the box.
3rd October 2020
The temperature sensor in Box 3 has been playing up. It is possible that Rahmet's baby Pax has been interfering with it, but probably more likely that I disturbed the connections when I was trying to fix the camera after it too had been playing up. The connectors on this box are not in the best condition and are in a difficult location to access.
Currently, Rahmet and Pax are sleeping in this box every day, so I won't be able to look at the sensor for a while.
24th August 2020
I have finally replaced Box 1. This box was first installed in February 2005 and was in pretty sad shape. It was actually OK inside, but the roof was rotting all around the edges and the frame attaching it to the tree had been all but destroyed. The box was hanging from the tree by one corner from a single bolt and it seemed to be only a matter of time before it fell. Also, the camera in this box was an old one in poor condition and there was no temperature sensor.
As of midday today the old box has been replaced by a new one. The new one is octagonal inside like Box 7 and incorporates a number of (what I hope are) improvements over previous boxes.
I will put the details of the box on the blog soon.
15th August 2020
I have finally bowed to pressure and obtained a properly signed SSL certificate for this site rather than using a self-signed one. The problem is that modern browsers complain that a simple http site is "insecure", and want you to use https. But they also panic when they see a self-signed SSL certificate. Opera will even prevent you from connecting to an https site which uses a self-signed certificate.
The obvious solution in my case is to use a free Certifying Authority - Letsencrypt - to obtain a properly signed certificate. I had been hesitating doing this because the default Letsencrypt client not only requires root privileges (which is bad enough) but also automatically updates its own code, which violates the design philosophy of this site. But anyway, I've now managed to find a way to use Letsencrypt without requiring root privileges or using automatic remote updates. The system is not fully tested at the moment; I will need to wait another three months to see if automatic updating of the certificates works properly, but there is every indication that it will.
What this means in simple is terms is that, as of about 2 pm today, you can now connect to https://possum.tv or https://www.possum.tv (both go to the same site) - if you want to - without your browser getting its knickers in a knot.
29th July 2020
Today the camera in Box 2 was replaced by a new one. The previous camera, which was installed in 2013, was never any good and deteriorated over time. The replacement camera is an FPV (First Person View) Drone camera, modified by removing the IR filter and installing it in a box with the usual IR LEDs. I will post about this on the blog.
6th July 2020
The cable Internet was down from 9:47 am to 10:17 pm today. I'm not sure what the cause was, but soon the cable will be changed over to the NBN, so there is probably all sorts of work in progress.
13th Dec 2019
The cable Internet failed at 7:30 am today. ETA for repairs was about 2 days, so I changed the router over to one with a backup 4G SIM in it. As of about midday, this site has been running over the 4G backup. You can probably expect this site to be a little slower until the cable comes back. In addition, when the cable does come back, the IP address will change and the site will be off line until I can change it in the DNS server.
Also, I'm sorry about the fact that Box 1 is still out of service and the Box 2 camera is so bad, especially as Box 2 is the only one in use at the moment. I do intend to address the problems.
24th March 2019
Yesterday, I replaced the failed temperature sensors in box 2, 3 and 6. I made some modifications to the circuitry to hopefully protect them a bit better against lightning surges, but without knowing the exact mechanism which caused them to fail I can't be sure this will do any good. Today I calibrated these sensors and the temperatures logged after 6 pm should be correct. I did not recalibrate existing sensors 4 and 7.
17th March 2019
A couple of days ago there was a nearby lightning strike which somehow managed to destroy the temperature sensors in boxes 2, 3 and 6. The server would also have been off the air for a short period around this time owing to the firewall locking up and needing to be restarted, however the possum TV server and cameras kept running. I can replace the temperature sensors, but it's been raining on and off since that time and it isn't urgent enough to justify working on the possum boxes in the rain.
27th October 2018
For the last couple of days, the cable Internet has been dropping out for extended periods, so if you have been unable to access this site, that is the reason. I do not know how long this situation will last
19th August 2018
The replacement camera for Box 3 has finally been installed. I had been ready to install it yesterday, but the was a possum in the box (probably Sasha and baby) at the time.
17th August 2018
On the 10th of July, box 3 camera failed. This box is difficult to get to; the top of the box is 5 m above ground level, there is an inconvenient amount of foliage underneath it and it's guarded by a nest of angry ants. In addition, this box has been in use by possums on several occasions (the temperature sensor is still working so I can tell when a possum enters). So it's taken me until now to look at it.
Normally when a box camera goes off line, there's either a bad connection somewhere or a blown fuse (I had used some cheap M205 fuses to protect the cameras and it turned out that some types of cheap low-current M205 fuses are prone to random failure after an extended period of use) and the problem is easily fixed, however in this case it looks like the problem is with the camera itself. I have removed the camera and will replace it with the one from the long-gone Box 5.
Unfortunately, this type of camera (Jaycar QC3474) is no longer sold and my only remaining spare is the one that was removed from Box 6 after it became embedded in honeycomb - I'm not sure it still works and it's unlikely to give a good image even if it does. Box 1 and 2 cameras are currently abysmal and well overdue for replacement, so I'm going to have find a new type of camera.
7th June 2018
This site was down from about 1 pm to 6 pm today due to more problems with the Cable Internet. This may be due to rain, however the cable came back with a vastly different IP address, so it's possible that Telstra are also making changes related to the NBN downgrade.
19th Feb 2018
There have been ongoing problems with the Cable Internet since the 16th. The Internet is currently down as I write this. If you're reading this it will obviously be up at that time, but don't be surprised if it goes down again soon. Telstra are not officially admitting to a problem, so I don't know how long this saga will continue.
17th Feb 2018
From about 6-8 pm yesterday there was a blackout here. Subsequently, the cable Internet was out until this afternoon. Since there isn't any possum action at the moment, this isn't a big deal, but is annoying nonetheless. I have been thinking about getting a UPS for the server, but that only partially solves the problem since there's not anything I can do if the cable Internet is down at the other end.
9th Jan 2018
This server would have been inaccessible to the outside world from about 8:30 am to 4 pm today. This is due to the Telstra cable dropping out. Given that these times correspond to normal business hours and given that the Telstra cable in this area is in the process of being "upgraded" to special NBN cable, it seems likely this was a planned outage. Any ISP that cared about its customers would notify them in advance of a planned outage, and wouldn't schedule one for business hours, but of course I have Telstra/Big Pond.
22nd Nov 2017
Earlier in the year, I decided to incorporate a radiation monitoring station into this server. This is currently on-line and working here.
"What," you ask, bewildered, "does radiation monitoring have to do with possum box webcams?"
The answer is nothing at all. I decided, for reasons I have explained on the FAQ Page (but can be summed up by two words, the first of which is "Donald"), that it would be prudent to start monitoring the local radiation levels. Piggy-backing on the possum TV server was the logical way to both log this data and display it on the web.
For the record, the history of this logging station is as follows (in reverse order of date):
11th Nov 2017 - I officially launched this monitoring station on the web, to the adulation of my many fans (actually, no one noticed).
5th Nov 2017 - I modified the software in the counters to automatically restart should it lock up. This uses the hardware watchdog timer in the microcontroller and also performs a hardware reset of the ethernet module, so it should be virtually foolproof.
14th Oct 2017 - The software in the counters stopped communicating to the server during a thunderstorm, probably due to a network glitch.
10th Oct 2017 - I installed the second Geiger counter.
22nd Sept 2017 - A radon washout event was detected from about 8 to 9 pm. It is encouraging that the station is capable of picking up the relatively small change in background level this involved.
19th Sept 2017 - I installed the first Geiger counter and started up logging.
8th July 2017
In addition to fixing the script mentioned below, I also added a backup method to alert me if the IP address changes and it fails to send an email. This uses a Server Side Include to display a warning message on the site.
3rd July 2017
The site was off-air from the 22nd of last month until today. The problem is that the script that's meant to alert me of an IP address change on the router failed to work. The site was still accessible from my local network, so I didn't notice anything amiss until I found out by accident today.
It turns out that Big Pond have changed the way their mail server works (presumably to combat spam), and this was preventing the alert emails from coming through.
5th May 2016
The site has been off-air for a few short periods over the past week. This was due to Telstra sending me a replacement cable modem/router and I also took the opportunity to upgrade the hardware firewall that I have in series with this modem. In principle, the modem/router supplied by Telstra (a Gateway Max) should act as an adequate firewall on its own, but if it gets hacked - and you can bet that there are a lot of people working on this as we speak - you are completely exposed to the world.
I find it sad, and indicative of the state of affairs in this backward country, that Telstra decided to send out these modems. Obviously it means that the cable network will be with us for a long time to come and the idea of an NBN has been pretty much abandoned.
30th Mar 2016
A replacement server has been built and was put into service at 1:30 pm today. All of the hardware (including video input cards) and software has been upgraded to the latest available versions. I will update the "About" pages soon to reflect this.
As a viewer you probably won't see much difference. In theory the movies should be marginally better quality although the limited bandwidth of the Internet connection is still the major impediment to streaming high quality video. One problem I have seen on the new system is that the box cameras on the "Box Overview" page were occasionally displaying the wrong boxes. This appears to be a problem with the latest version of Motion, which I think I've fixed.
Also, if you're using SSL to access the site, you might have your browser pop up a scary warning about the site being untrusted or having incorrectly configured security. You can ignore this. The security has in fact been significantly upgraded, and the warning is due to me having generated a new self-signed certificate. Browsers distrust self-signed certificates, however the alternative is to use a Certifying Authority, which is a) expensive and b) not really any more secure.
It's possible that more bugs will appear over time. If you find anything not working or have any questions, please email me ().
13th Mar 2016
At about 1 pm on the 8th of March there was a power outage which caused the motherboard of the Possum TV server to fail. Unfortunately I was away from home at the time and couldn't do anything about it until today. I have temporarily substituted a spare motherboard to get things working, but I intend to upgrade the server as soon as possible.
30th Nov 2015
Yesterday, box 3 was found to be almost ready to fall out of its tree. The four screws at the top of the frame had pulled out of the tree entirely, leaving the box hanging outwards at a severe angle, held only by two bolts at the bottom, one of which was broken. No wonder Farley recently abandoned this box for the less comfortable, but safer, box 2.
This box was fixed today. I used four 100 mm long x 8 mm dia stainless coach screws, and I put spacers behind them to allow for growth of the tree. I still haven't found the perfect way to attach possum boxes to fast-growing trees like Flooded Gums; as soon as you screw anything into the tree, the wood starts to swell outwards at this point (the tree tries to engulf the foreign object), and in a few years can push out as much as two inches from the original location. It doesn't help that bark, which would normally be shed, gets trapped behind the screws as well.
24th July 2015
I have added a Box Camera Overview page, which displays images from all six (working) cameras simultaneously. This is intended as a convenience for visitors; rather than having to tediously step through the individual box camera pages one at a time to check if any box is occupied, you can now just go to the Overview page.
This page is currently fixed at a 5 second update rate to conserve bandwidth and isn't really intended for watching possums in their boxes (although you can use it for this purpose if you want). If you want to watch an individual possum box, you are better off going to the individual box camera page, where you can set the update to a faster rate, as well as do other things, such as expand the image and take snapshots.
I was motivated to build this page after a visitor actually went to the trouble of creating a page like this for their own use.
25th Jan 2015
I have just put a new page on this site to display possum box temperatures as graphical trends. This code has had limited testing, so please bear with me if it has problems (and also feel free to email me if you find something wrong).
Currently, there are a couple of limitations deliberately built into the code: it won't display more than month's data at a time, and when you select more than a week's data, it only reads the temperatures at 10 minute intervals; therefore the maxima and minima won't be truly accurate in this situation. The intent of both of these limitations is to reduce the load on the server; this may be unnecessary, but I want to be cautious to begin with.
19th Jan 2015
I have just installed a temperature sensor in Box 4. Like the others, it's been calibrated to within 0.1°C of my reference; of course it's a question of how good my reference thermometer really is, but at least I can say the box temperatures are consistent with each other.
I still need to put a temperature probe in Box 1, but that's going to be a big effort because I want to replace the camera and cabling to this box as well.
Also, I've changed the box cameras to default to streaming mode (rather than a 2 second update as before) when you first access them. If you've already been looking at the box cameras, you won't see any difference because the site remembers (via a cookie) what update rate you were using before, however new users will be greeted with streaming video. The reasoning behind this is that Kiki might soon come into Box 7 with her latest baby and we might get new visitors as a result. A new user might not think to select streaming mode and might miss out on much of the fun.
6th Nov 2014
I'm not sure what's happening, but Telstra/Big Pond have been changing my IP address a lot more frequently this year than ever before. In the past, the IP address would often remain the same for over a year at a time, but so far this year it has changed five times already. It's been hopping back and forth across a number of wildly different subnets, so it would seem that Telstra have been making big changes to their network. I don't know what's behind this, but I suspect it may be related to the building, and then subsequent destruction, of the NBN.
When I moved away from DynDNS.com, I decided to address the problem of changing IP addresses by the use of a small Perl script that periodically checks my IP address and sends me an email when it changes. [BTW, if anyone wants to know how this script works, feel free to contact me () and I can send you a copy.] This script has (unusually for me) worked flawlessly so far, but it has the disadvantage that it relies on me checking my email and then manually editing my DNS settings. This means that the site could go off-air for the best part of a day if the address changes at the wrong time.
So please bear this in mind. If the site becomes inaccessible, you might have to wait a few hours before the problem is resolved.
On a separate note, it was interesting to see that the day after the Shellshock bug was disclosed, I was already getting bots probing the server for this vulnerability. Fortunately, I don't use shell scripts to do CGI, so the server was never vulnerable, but it's a salutary warning that if you run a web server, you need to keep on top of these things. Every day I get referrer spam and bots probing for open proxies and vulnerable PHP scripts, as well as old favourites like the Romanian "ZmEu" (who has been around ever since I enabled logging). You don't have to be paranoid of these nasties; but you do need to realise that, like cockroaches and blowflies, they'll always be out there trying to get in and you need to keep your flyscreens secure.
18th Apr 2014
It looks like I wasn't a moment too soon in registering the possum.tv domain name, because DynDNS have just informed me they are closing down all of their free accounts. So in 30 days possum.dyndns.tv will no longer exist.
I have currently set up possum.dyndns.tv to redirect to possum.tv, but this is going to stop working after the 30 days is up and you will need to change any bookmarks to this site to http://possum.tv.
13th Apr 2014
I have added daily maximum and minimum box temperatures to the box camera displays. If you don't see "Box x Temperatures So Far Today" on the right of the screen, you might have to do a refresh. Please let me know if there are any problems with this (bearing in mind that there is no temperature monitoring on boxes 1, 4 and 5). Eventually, I'd like to display temperature data in the form of graphical trends, but don't expect anything soon.
29th Mar 2014
I've managed to obtain the domain name possum.tv. Previously it was being used by some Russian site, but it seems they shut down some time ago and let the registration lapse. So you can now get to this site via http://possum.tv and http://www.possum.tv, in addition to the original link http://possum.dyndns.tv.
The idea at this stage was to grab the domain while it was available and worry about the details later, so the setup isn't perfect.
In detail: The potential problem here is that the server is still on a dynamic IP address, and I'm not currently using a dynamic DNS for the new domain. So this means that I need to manually update the DNS entry whenever my router changes IP address. By contrast, the possum.dyndns.tv domain is hosted by DynDNS and theoretically updates the IP automatically. Unfortunately, using DynDNS has two problem: Firstly, about 5% of the time my router powers up, it updates the IP address wrongly (even if it hasn't changed), making it almost as bad as a manual system. Secondly, DynDNS are trying to eliminate free accounts like mine. I'll lose the domain name if I don't log in to their system once per month, and at any time they could decide to cancel it. I suppose I could transfer my new domain name to DynDNS, but this would require me to upgrade to a paid service. This would also eliminate having to log in every month, but OTOH I don't want to reward DynDNS for their (arguably) deceptive behaviour regarding free services, and there's still the problem of the address sometimes updating wrongly.
The ultimate solution is to move to a static IP address. For reasons that I can't fathom, it is not possible to get a static IP address for cable Internet in this part of the world. I had hoped things would change when the NBN arrived, but now that the NBN has been scrapped, it looks like I'm stuck with a dynamic IP address for the foreseeable future.
In other news, I've also found a typo in the script that takes snapshots of possum boxes. The effect of this bug was that if the system was a bit slow, rather than waiting for a new snapshot to be generated (as intended), it instead returned an older snapshot in its place. This has been fixed; it will now wait up to 1/2 sec for the snapshot to be generated (which should be more than enough given the system should be able to capture in excess of 50 frames/sec), and return the error message "Timed out taking photo" if this fails. I apologise to anyone who has lost snapshots due to this bug.
For the record, the temperature sensors in box 3 and 6 were calibrated on the 9th and 15th Jan respectively. So all box temperature probes are now calibrated to (theoretically) within +/- 0.1 °C.
5th Jan 2014
Temperature monitoring is now in place in boxes 2, 3, 6 and 7. Box 7 temperature sensor had been installed some time ago (22nd Oct 2013) and boxes 3 and 6 were hooked up today. These last two sensors haven't been calibrated yet, but should probably be accurate to the nearest degree.
13th Oct 2013
Temperature monitoring has now been hooked up for Box 2. This is a temporary arrangement for testing purposes; if it goes well, I'll put temperature sensors in some of the other boxes.
I'll document this in more detail some time later, but briefly: I've put an AD592 temperature sensor in the box and cabled it back to an ADAM-6017 Data Acquisition Module with Cat 5e ethernet cable. A custom C program on the server polls the ADAM-6017 using Modbus/TCP protocol, and then feeds the temperature data to Motion using http requests.
7th Oct 2013
Box 2 camera has finally been replaced. The end result isn't quite as good as I hoped and gives a fairly murky picture during daylight, but at least the interference is gone.
New cabling was installed and some new conduit was added; including a piece of flexible conduit linking to a junction box on the outside of the possum box. This allows the camera to be unplugged and the box to be removed from the tree more quickly and easily. Also the roof of the box was given some additional protection by fixing some Hardiflex cladding to the top.
More details can be seen on the Blog.
17th Aug 2013
As an experiment, I have set up the system to allow full streaming video to the outside world. In theory, all you need to do is set "Update Rate" to "Streaming Mode" on the box camera page. In practice you might need to refresh the page, or possibly restart your browser, the first time you do this. If you're using Internet Explorer, you will probably also have to install the Cambozola Applet; your browser should prompt you to do this if needed.
Please bear in mind that this is an experiment. My uplink has barely enough capacity to support even a single user at a time, and I'm streaming using "multipart jpeg" format, which doesn't have sophisticated caching like you have on sites like YouTube. This means the video might become jumpy or stop altogether under some network conditions. If it gets annoying, you should be able to switch back to the old system.
Also, I've set it up to automatically stop streaming after 30 minutes of continuous viewing. This is to stop it chewing up bandwidth if you accidentally leave it running. Unfortunately, the way the system works is that it can't pop up a message to inform you of this; the picture will either go blank, or simply freeze, which isn't always easy to notice. To get it running again after it's stopped, just refresh the page.
Let me know if you have any problems/comments/suggestions ().
1st Nov 2012
Box 5 has been taken down; it was not attracting any possums and was pretty much just a waste of space. I intend to reallocate the camera to a new box eventually, however upgrading the cameras in boxes 1 and 2 will take priority over this.
18th Sept 2012
This site has been inaccessible for about a week and has changed its address from www.possum.dyndns.tv to possum.dyndns.tv . Since you're reading this this page now, you have obviously been able to find the site at its new address.
In case you're interested, a technical description of the problem follows...
This sequence of events was triggered by a power outage which caused the router to send the wrong IP address to DynDNS. There appears to be a race condition involving the cable modem and router; sometimes when both are powered up simultaneously, the router ends up thinking the WAN address is the default power-up address of the modem rather than the actual address it gets through DHCP. The router then sends this address on to DynDNS, which renders the site inaccessible until I find out what has happened and update the address manually.
This happens rarely (perhaps 1 in 20 power-ups) and is not easy for me to detect, because everything on the local network works correctly. I still even see web traffic on the Possum TV Live server; apparently bots are accessing it by IP address.
This time, I manually updated the IP address, then tested the site by feeding "possum.dyndns.tv" into the site http://www.downforeveryoneorjustme.com/ and also following a link in my browser to www.possum.dyndns.tv. Since this worked, I though I was alright. However, it turns out that only "possum.dyndns.tv" was resolving and not "www.possum.dyndns.tv"; my browser was actually failing to find the site at www.possum.dyndns.tv, but then automatically trying possum.dyndns.tv, which then worked.
Why was the www name not resolving? It turns out that some time ago (last year, I think), DynDNS made some changes to their service. Amongst these was that they no longer supported wildcard domains (i.e. xxx.possum.dyndns.tv) for free accounts. It seems they didn't go through and purge their DNS records at that time, and it was only when I went and manually updated my IP address that this was triggered.
10th July 2012
After being away from home again, I found the changes I made last time actually made the site worse. The problem was that a 1 second update rate is far too optimistic for a slow satellite link; it was trying to refresh the image before the previous one had even started to load. I had actually anticipated this, but the method I used (inspecting the "complete" property of the image object to detect whether the previous image had loaded) doesn't work for really slow links.
So I've made another change; you can now manually select the update rate. I recommend starting with the 1 Sec rate; if it doesn't update or sticks on the "Webcam Interface Initialising" message, try progressively slower rates. Under normal circumstances, it's unlikely you'll have to go slower than a 2 Sec update. As before, you might need to clear your cache and please email me if you have any problems.
20th June 2012
After being away from home for a while, and having to access Possum TV over the web (as opposed to the local network where I have real-time streaming video), I realised that it wasn't performing as well as it should. Out-of-date pictures were often being displayed on startup and sometimes only part of a picture would be shown. Previously I had put these problems down to caching issues and network delays respectively, but after some investigation it turned out that my scripts were primarily at fault.
18th Apr 2012
If you've been unable to get onto this site recently, it's because the cable Internet was down for almost a day. This is probably the worst outage I've had so far with Bigpond cable, which is normally very reliable. No idea what caused the problem and of course Bigpond have nothing about it on their Service Status page, which is unfortunately standard practice for ISPs in this part of the world.
The python which ate Gumdrop and/or Lychee has returned and is currently camped in box 4. Quite possibly it's looking for another meal. While I shouldn't pick sides, I do hope that it finds some rats to snack on rather than any of our possums.
A few days ago, I finally started posting to my Possum TV Blog. I'll generally be using this blog from now on, so this News page is probably going to be updated even less often than previously.
11th Mar 2012
I've just put up a new page on this site with photos of our possums. Eventually I would like to expand this to more of a gallery like the one on Sneddo's Possums site. And this reminds me that I promised to link David Sneddon's possum site about six months ago, but never got around to it. I've been meaning to create a links page; maybe I'll do it some time this year.
18th Feb 2012
On 15/2/2012 a carpet python came into box 2 and attacked Gumdrop and her baby (either Lychee or Wasabi), killing and eating one or both of them. From the size of the bulge in the python, I would say that at least Gumdrop was eaten. It was very sad. Gumdrop was our most well known ringtail; she has been with us for several years and shown us many of her babies. She was one of the very few ringtails who would take food from hand.
Carpet pythons are native to the area and their natural diet includes possums, so it's not really fair to blame the python. A couple of plump little ringtails must have been a very inviting, and perfectly legitimate, meal. Still, it would please me a lot more if they could eat the cane toads instead.
The attack was caught on film. I think it will be popular if I upload it to YouTube, since many of the hits on my "Python Visits Possum Box" movie seem to be from people looking for footage of pythons attacking possums. Incidentally, the python in that movie is a different one to our current visitor.
In other news... well actually there's not really any other news. The cameras in boxes 1 and 2 still need replacing, as does box 6. There are more possum movies that need to be edited and uploaded, including an epic fight between Piranha and Marlon. And I need to start a proper blog some time.
17th April 2011
Still not much in the way of possum-box action this year. Švejk was using box 7 for a while, but was warned off by Kiki and hasn't been back. Gumdrop and her baby Chilli were using box 2 in turns, but since Chilli was scared out of the box by power company workers trimming trees nearby, neither have returned. I expect one of the ringtails (probably Gumdrop who has babies in the pouch) will return to box 2 in the near future, however you can never be sure.
All of the movie highlights (with the exception of one blurry possum fight which wasn't worth the effort), and all of the movies from the old Possum TV site, have been put on YouTube.
The ants were removed from box 3 by dousing with water, however no possums have so much as stuck their noses into this box since that time. Meanwhile, more ants have taken over box 1.
17th Dec 2010
I have to apologize on behalf of our possums to anyone following this site; there has been very little to see since about September. This might be related to the unusually wet spring/summer we've been having so far, although the ways of the possum remain a mystery. The wet weather has also deterred me from from doing work on the boxes (or at least that's my excuse). Boxes 1 and 2 should have their cabling and cameras upgraded, ants have moved in to box 3, box 5 should probably be relocated and box 6 is still out of service.
I'm currently in the process of putting some the possum movies onto YouTube. This seems to make sense; it's easier for people to use and there aren't any bandwidth problems.
On 20/11/2010, I upgraded the power supply for cameras 3–6. The old power supply was a linear bench supply which was woefully inefficient (about 30%) and the cabling had grown into a shameful rats nest. I replaced it with an SMPS unit in a dedicated enclosure with proper power distribution. As a bonus, the problem with crosstalk between cameras 3, 4 and 6 went away after I did this, possibly due to my cleaning up the mess of cabling and swapping out a few dodgy patch leads.
21st Oct 2010
On 26/9/2010, bees took over box 6 with frightening speed (I may put up a movie of this some time). This is apparently a common problem with sugar glider boxes and happened to the Melbourne Wildlife Sanctuary sugar glider box earlier this year. Anyway, the bees have now finally been removed but the box will require a bit of work before it can be returned to service. The wood of the box was soaked in beeswax, which has to be removed or it will attract more bees (or so I am told). The box camera was embedded in the honeycomb and may be difficult to salvage
I am going to have to work out how to deter bees from taking over this box again. The most effective method is said to be to attach a shelltox pest strip to the ceiling of the box, however these are no longer available anywhere, and I'd be cautious about using anything toxic, given that possums frequently lick the walls of their boxes. There is some suggestion that attaching shag-pile carpet to the ceiling of the box might work—the idea is that bees start building their honeycomb from the top down and need a solid base to attach it onto. However, from what I've read on the web, the effectiveness of this method is not well established.
24th Aug 2010
I've created a new page which contains a few of the more interesting movies, edited down small enough to download in a reasonable time frame. I might also upload some of these movies to YouTube some time.
I haven't looked into the camera problems yet. The brightness in some boxes pulses intermittently and sometimes there also seems to be crosstalk between the cameras. Possibly it's a video card problem.
6th Aug 2010
I've temporarily installed a new box (box 7) in the garage in the hope of enticing Yoshi out of the old tyre in which she sleeps. The motives for this are partly selfish; Yoshi has a baby in the pouch and I'd like to capture it on film.
The problem with box 1 camera mentioned below turned out to be interference caused by a VCR, which had been connected to monitor the camera in parallel with the server. This was fixed by placing a loop isolator in the circuit. Meanwhile, box 3 camera has started to play up.
The "about" page has been changed to break it into sections, which makes it easier to add stuff without the page getting too long. I've also added data about the boxes.
10th Apr 2010
I've hooked up camera 5 with a temporary cable. I haven't run a permanent cable yet because (to the best of my knowledge) no possum has ever slept in this box and I might have to relocate it. Meanwhile, box 1 camera has been playing up. Boxes 1 and 2 use old cheap CMOS cameras which were never as good as the others to begin with and seem to work even worse with the new video card/driver. I will be looking at replacing these cameras, however this requires significant recabling and it might be a while before I get around to it.
13th Mar 2010
I've recently upgraded the server.
The full story is this. When I put up boxes 5 and 6 in January, there were more boxes to be monitored than there were video inputs on the server. I therefore purchased another 4-input video input card (giving me up to 9 usable video inputs). This new card uses Techwell TW6805A chips. I had assumed my old TW68k driver would work with this card since the driver works fine with TW6802s and I would have thought that the TW6805A would be backwards compatible - but of course Murphy had something to say about that. The good news was that the guy who wrote the original driver had just produced a new, and much more polished, version that would work with the new chipset. The bad news was that it required a minimum kernel version 2.6.28 and I had 2.6.27. Upgrading the kernel pretty much means rebuilding the whole system, so I put it off for about a month while I concentrated on other (paid) work.
Anyway, I've now finished the upgrade. The server has had a new 2TB hard disk installed (not essential but it made the upgrade easier) and is running on kernel 220.127.116.11. A lot of other software was upgraded as well. Everything now works, however there's still some tweaking going on. Box 5 has not yet been connected to the server, and I'm debating what to do about this box since the possums seem to ignore it; possibly I'll move it somewhere else.
8th Jan 2010
Some new boxes have been installed. On 20/12/2009 I put up possum box 5 and on 2/1/2010 I put up possum box 6. Box 5 was put up in a hurry after a neighbour evicted some possums (Kiki and probably Leena and Yoshi) from his ceiling space, and I was worried there might be a housing shortage. This box is a horizontal design which was intended to be like the ceiling spaces the possums had been living in. Box 5 hasn't been painted nor had a camera installed at this point. Box 6 is a sugar glider box which I was prompted to put up after a sugar glider briefly visited box 2 late last year.
I'm also experimenting with a facility where you can click a button to a take a snapshot of a possum box, which can then be downloaded as a jpeg file. As currently implemented, this facility works and you are welcome to use it, however I'm considering some improvements.
9th June 2009
On 22-23/5/2009 box 2 was cleaned and repainted. On 24/5/2009 I put up box 4. Due to a wiring error on my part, the infra-red illuminator failed soon after, rendering the camera useless after dark. Unfortunately, this happened while I was away from home and I was only able to fix it on 6/6/2009. I've also fixed a few issues with the remote webcam interface, although it's still not perfect.
18th Feb 2009
I've just got hold of a preliminary driver for the IVC-268G video card, and we're currently running on all three cameras. If there's actually anyone out there looking at the site, you can expect a bit of downtime for the server and/or cameras while I test the new driver and fiddle with stuff generally.
7th Feb 2009
Over the past few weeks, this web site has undergone a lot of tweaking and improvements. The biggest issue at the moment is that we're still only running on one camera. I'm in contact with a guy who's working on a driver for my IVC-268G video card and I'm hoping that some time soon I'll have something usable. Once all three cameras are up and running, I'll look at opening the site up to the world.
Below this point, all "news items" refer to events which pre-date this section of the site. They have been included to give an idea of how things developed.
During January 2009 I built a new possum box (box 3). This box incorporated a CCD camera in place of the CMOS Swann Day/Night cameras used in the other boxes. I did this mainly to get better quality pictures, but also because the Swann Day/Night cameras are now apparently obsolete and difficult to find. The new box was installed on 26/1/2009, and at that point I found that it wasn't practical to run the new camera in parallel with the other two. It appeared that the signal levels between the two camera types were so different that the video chip couldn't keep up with switching between the two. The end result was that I ended up disabling the other two cameras because I particularly wanted to know how the possums were liking the new box.
At the start of the month, box 1 camera was hooked up to the box and I began tweaking and tuning the settings. During this period there are some pretty poor quality movies recorded as I tried out various options. On 11/12/2008 there was a slight setback when box 1 fell from the tree. This event was caught on film, but only shows about a second of confused tumbling before the camera cable broke. Four days later the box was replaced and fitted with a new camera.
On 24/12/2008 I hooked in box 2 camera. This required more tweaking of settings. Owing to the fact that I was using the Osprey 100 card which uses a single multiplexed video chip, I needed to drop the frame rate to be able to monitor two boxes at the same time.
During this month the server computer was built and all major items of software installed and configured. By the end of the month, the computer could stream and record video and was running an operational web server. At this point the computer had not yet been connected to any possum box cameras and only the most rudimentary web site had been built.