|
in reply to Web moniitoring with Perl, comments and suggestions on general approach please
Three general points:
If you go with stevieb's libs, and switch to RPi:
* Working together is more fun. stevieb will get an incentive to mature his libraries (win for the community) and you will get good support.
* RPi's are more expensive. Think also about theft/raindamage/overheating of hardware and how to secure it (high pole, or inside a fake? hive or bury the hardware). Although, on second thought, I guess that bees guard the hive well. BACKUP your data!
* Graphics can also be made with javascript and because some have CDN links (which means the javascript code is downloaded from elsewhere) could potentially be lighter than generating images, as all you provide is the data with some initializing code.
I also like the fact you are going for RRD. Remember you can add data to more than 1 database, so if you collect each half an hour, and have 48 measurements, create another RRD that feeds on that first information, but summarizes the data into, say, 8 hour chunks (3 measurements per day). This way, you can have daily data, and weekly/monthly averages, easy to query and display on your webpage. And at fixed databasesizes. |
Re^2: Web moniitoring with Perl, comments and suggestions on general approach please
by predrag (Scribe) on Jan 28, 2017 at 19:00 UTC
|
FreeBeerReekingMonk thank you very much for your comments.
Yes, I think I will start my first project with RPi's. The reason why I was thinking about PICs solution for the hardware was because I would learn PICs and later may realize some other ideas that PICs are very suitable for. Also, I think power consumption will be less then with other solutions . And also, after years and decades of being not active in electronics, I am eager to do so many staf. A bit not so serious reason but…
I think I will not go with javacript, it will be too much for me.
RPi's are more expensive
For me as a beginner and an amatheur, the price is much more relative in comparison what it means to a professional. Even in the situation when I at this moment can't start project yet, because of the lack of the money and some other reasons. But I believe I will start this spring, till that I will preparing myself and learning. For me, it is the most important that I can finish that project and that I can do as much as possible of work myself, so I want to learn as much as possible.
Think also about theft/rain damage/overheating of hardware and how to secure it
Yes, I think about that. Till 5-10 years ago, that wasn't a serious problem here but now it became. I will maybe have monitoring hives close to someone's house…
I guess that bees guard the hive well
Yes, of course, but in general our race of bees is not much aggressive. But if I would have installed the hive to crash down in the attempt of theft, thefts will be give up.
Since I didn't begin anything yet, it is pretty childish of me to tell that I have many other ideas about web monitoring in beekeeping. For one, it will be important to have daily average temperature. I will not use RRD for my first project but for sure will use in some next, I see how powerful it is.
| [reply] |
|
|
Not childish at all. Remember that inventors have many ideas in their heads, but still go for one or two that seem the most interesting. Write them down though, and if one is good enough... you can even patent them!
Microcontrollers are booming with IoT around the corner. It pays to invert time in them to learn about it. And people that have cross-over skills (knows software and hardware (for example, how to protect a sensor with shrinkwrap so it lasts longer) AND has knowledge of, say, winddirection, so know how to place a box in the shade, away from rain) are valuable.
Also: the same set-up you make for monitoring your bees can be used in a tomato-greenhouse, where humidity (air AND soil) etc need to be measured too... so it pays to understand the full process (where to buy cheap hardware, how to set it up, how to read the data, software) and you can support other farmers. And you either get free tomatoes, or a bit of money for support/maintenance.
And it is not hard. I mean, you can go for the expensive humidity sensor or a very cheap sensor made of mostly recycled stuff.
See also:
www.gardenbot.org
On this website you learn about using bottles wrapped in white paper to keep the sun and rain out. Low tech does not mean bad tech...
| [reply] |
|
|
FreeBeerReekingMonk thank you for the answer and the encouragement for my work.
You are right that while working on my ideas, I may find something very interesting, who knows. For example, bee hive sound analysis is very interesting, becomes a bit popular now, and many years ago, I've got one idea from one Russian beekeeping journal about completely different and more serious approach to that. Now I see I can do all that with Perl, it is really amazing
Yesterday I've looked a little about RPi and it became obvious that it is more then a great tool for me. Especially because I am now enough prepared to start - after taking two Linux courses a year ago and since I've came into Perl. Before that, as I've already wrote, RPi wasn't enough attractive to me and maybe that also had to do with their way to represent it. Regarding PICs, I will need some learning, but I feel I am prepared for that too.
That small Serbian company that has been mentioned in previous posts produces some extensions for RRis, so with use of these it is possible to connect up to 100 different click boards that they produce, to RPi. On the other side, I can make some electronics myself too, since I have education in electronics, that just has to be renewed
Also: the same set-up you make for monitoring your bees can be used in a tomato-greenhouse
Yes, I really love gardening and thanks for nice links too
Low tech does not mean bad tech..
It is always so thrilling to me. Recently I've found an astronomer's website where they measure star brightness with Perl. And regardless of Perl, I am always delighted to see examples of simple solutions for something very complex.
| [reply] |
|
|
