|
DMI News |
Previous Entry.. Next Entry.. Plans for the future
July 27, 2013 09:55
I have chosen in recent years to not announce too many of my plans, as I had practically no time to work on any of them anyway, and lack of progress tends to annoy people. Furthermore, many ideas which I actually made some progress on got stalled by other issues, many related to the age of the old server. Just getting fully upgraded to the new server took several months. Of course, it's difficult to worry much about new projects when the many projects that are complete and functional aren't even in service due to the state of clutter in the house. To make it more fun, efforts to reorganize the house to hopefully reduce the clutter have actually temporarily created MORE clutter. It just keeps getting better and better. In any event, the mess will be cleaned up in the next couple of weeks. There is also a chance, just a CHANCE, that I might be getting a new job. It will be a few weeks before I know anything about that, but when that happens, it should result in more available time to work on the site. Keep your fingers crossed. The first order of business is to get the functional things active again. This includes the critter and the other cams. At least the Living room and IPcam cams will be reactived. Hopefully in time I can also include the kitchen and hallway ones. I eventually want to add cams to cover the front porch and the backyard, partially for security, and partially for doorbell logging, motion capture, and sprinkler control. Next, I'm going to clear the projects list and start it over. I created it as a means to track my own projects, along with allowing other to submit ideas. Most of the ideas presented were either pointless, or rehashed ideas that I had already had or based on projects I had already done. So for now I plan to just use it to post my own projects and allow others to view the plans that are in the pipeline (a question that is very frequently asked and for years left mostly unanswered), and track my progress on them. It will also allow me to help focus some of my efforts. For quite a while now, I have not allowed people to add new cams. For a while, I shut down the ability to create new accounts as well. You can thank the spammers. I still have about 10,000 bogus accounts that I need to remove, and in the progress of removing them I have made a few database modifications and I need to make several more to speed things up. Nobody really notices when it takes 5 seconds to delete an account, but it's VERY noticable when you're trying to delete 100 at a time. I need to repeat some of that process for the cam database, as well as clean it up. Even of the legitimate cams, more than 75% of the cam accounts created have NEVER been used. As for cams, I'm looking into ways to improve that feature as well. As you well know, the popular cam portal sites are using flash applets to stream the cams (and audio). While this works great, the backend for that operation (flash media server) is quite expensive, and I'll have to purchase more bandwidth capacity on the server to serve any significant amount of streams, or streams of any decent quality. Not only that, my upstream here isn't horrible (1.5mbps), but it's certainly not great, and wouldn't support several decent quality streams. On top of that, I would have to upgrade most of the cams I'm currently using to something that can stream to that format. Sticking with motion jpeg like I'm using seems horribly dated, but it works well, it's low bandwidth, and it scales nicely. It also costs me nothing extra. Still, to improve in some of the ways I'm planning in the near future, some of these upgrades will be necessary. I've got plans for a new sign. One much larger and possibly in full color. The size I'm proposing would be 8' x 2' and would have 96 by 24 pixels. Right now it appears it would cost about $250 in materials and would take about 72 hours to build, not including R&D time. I'm investigating the option of installing a computer in my car permanantly (instead of using the laptop like I'm doing now). Ideally it would replace the stock radio in the dashboard and provide a 7" touchscreen. Since I need to run a real operating system on it (either windows or linux) in order to do my car broadcasting, available options in a ready state are very scarce, very expensive, and from very iffy sources. I will likely assemble my own which won't be much cheaper, but at least I'll be able to source the parts from more reputable suppliers. Interfacing it with the car's electronics might be a challenge. Interfacing with the power and audio won't be too much trouble, but it would be nice to also interface with the buttons on the steering wheel, the honda i-mid display on the dashboard, and possibly also the obd2 interface. I'll also need to have usb jacks set on the dash for the camera and usb to plug into, and something front mounted for my phone. Raspberry PI seems all the rage, and it IS a pretty powerful little computer for the price, and I can likely use it for a lot of projects, but it's lacking in a few features that the TS7200 I'm currently using makes available. It also runs a lot faster (but also uses quite a bit more power as a result). If I can reduce the power consumption down to what the TS7200 has, along with increasing the number of I/O and ADC ports available, I could use that on the critter, and also add audio capabilities to it. I may have to design stage 3 of the critter project to be in a larger weight class so it can haul a heavier battery. I've also looked into different types of batteries which might provide more energy per weight. They are, of course, far more expensive, but it would solve some problems. A docking station (or several) for the critter need to be built as well, which will allow 24/7 critter availability (with downtime while charging) and will not require human intervention to charge the batteries. Future versions of the critter may also have appendages or the ability to play lasertag type games with each other. Also working on SLAM algorithms to allow the critter to draw accurate maps of its environment and know where it is and how to pathfind to other locations on the map. This will allow people to click on a map to move the critter to that location, and allow the critter to find its charging stations. I'm also adding some other more "social" features to the site. I won't be creating a new facebook or anything like that, nor do I want to be another cam portal, but I have some ideas to make the site itself more entertaining beyond just lamp, critter, and other device control. I'll reveal those plans when they're closer to being ready.
|