Overall System Design Specification

[hamishmb]

Hi,

I can get something ready to deploy now basically I have it measuring every 5 seconds at the moment, but I can change that. What I'm not sure of is how I can convert frequency into height of water, but I guess we can figure that out.

Yeah, shall we switch to the Progress forum?

Hamish

[TerryJC]

I have written a small application to test Penri's capacitive sensor, and it is attached here. I seem to get ~790Hz from the test board, is that correct?

It seems pretty close to me.

Even though I have my Pi running at 100% CPU all the time it works fine, see the comments in the code for more details.

I assume that the other processes taking up CPU time wouldn't be there in the deployed system?

Terry, would you mind peer-reviewing this for me please? I know it's a bit messy, but I'm thinking about releasing it as an easy way to measure frequency of a signal for people who are new to RPi.GPIO, like me XD

For someone who 'doesn't know Python', you seem to have made a pretty good effort

I'm not going to have time to do much today, but I can make some initial comments.

1. It is unimportant for this experimental code, but for the deployed code in the Lighting and the Bells software, I assigned the copyright to 'Wimborne Model Town' and released it under the GNU General Public Licence (the GPL). If you look at the example of the Lighting software at http://www.hadrian-way.co.uk/WMR/Code/ you will see how I did it.
2. I wouldn't bother with the user input of filename or the number of readings as an argument. The deployed code won't need it. It's good to see that you know how to do it and that you should check for errors.

That's all for the moment, but I'll have another look over the weekend. The fact that it works goes a long way towards the goal, so any comments I make at this time are not particularly useful.

By the way. Are you aware of the 'pylint' tool? This is a Linux command line tool which analyses your code and flags up syntax errors and so on. It's a bit picky and will complain about things that actually do not prevent correct execution, but simply violate the coding standards. At the end of the process it allocates a score (6 out of 10 for this code on this machine which hasn't got RPi.GPIO installed). Most of the things that it flags up are comments and warnings, such as invalid constant name (use of caps lower case). I usually ignore most of what it tells me, but it can be useful.

[TerryJC]

What I'm not sure of is how I can convert frequency into height of water, but I guess we can figure that out.

The deployed system will need to be calibrated. If the results turn out to be linear, then a simple calibration factor will suffice. If not, then we will have to use a look-up table and interpolate the answer from that.

[hamishmb]

Okay,

Yeah, I was thinking of releasing this as a separate module eventually, so may be easier to do that if I own the copyright XD

C++ is the one I don't know (though learning is going well), I've written tons in Python, including 3 publically available programs

2) In the end they'll go, but they might be helpful for the moment for testing and short deployments with Penri's sensor.

The other processes are my distributed computing things, so they won't be running, just an interesting point I thought I'd make.

Am I meant to be using my RPi for the test deployment? Might be tricky if so cos I'm using it for NAS media serving XD

Pylint sounds good, will have a look.

Hamish

[TerryJC]

Yeah, I was thinking of releasing this as a separate module eventually, so may be easier to do that if I own the copyright XD

I don't think it makes any difference at this stage.

C++ is the one I don't know (though learning is going well), I've written tons in Python, including 3 publically available programs

Oh. I could have sworn that you said that you preferred C++.

2) In the end they'll go, but they might be helpful for the moment for testing and short deployments with Penri's sensor.

OK.

The other processes are my distributed computing things, so they won't be running, just an interesting point I thought I'd make.

OK.

Am I meant to be using my RPi for the test deployment? Might be tricky if so cos I'm using it for NAS media serving XD

I can provide a Pi for the job.

[TerryJC]

I’ve attached the modified layout document herewith, can you please check through it for me to see if you think the document is consistent with itself and the diagram used in the design spec. is good enough.

Penri,

Some questions about this diagram:

1. You have changed Page 1, the Outline Schematic. I'm assuming that you want me to include this in the Design Spec along with the updated Page 2?
2. I don't think you've changed any of the other pages. Would you like anything from these included?

Once we've finalised which pages we will use, I'll split this drawing into 5 individual files, to avoid the problem with layers that Clive encountered. It will also be easier to manage the configuration of them once they are separated.

[Penri]

Terry

Apologies for the tardy response, the notifications are not getting to my e-mail and I've had a busy weekend and didn't check the forum, sorry.

I was happy with just Page 2 in the Design Spec., it provides a good overview to put the topics in the spec. in context. I did update the other diagrams in the document to try to make them all consistent with Page 2 so if anyone was looking for a bit more detail it would be contained there. I think splitting the document is a good idea, in doing so it would be worth reviewing each diagram to see if adds value and discarding / archiving those that do not.

Penri

[TerryJC]

All,

I have just posted V 2.3 of the Design Specification to the Development server (see http://www.hadrian-way.co.uk/WMT_River_ ... nt/Design/).

This version includes the agreed changes from our review which we held at the WMT on 1st June, plus some updates to incorporate some of the ideas being trialled at present. All earlier agreed changes have now been accepted in the document, so the new stuff should be pretty clear.

There are still a lot of what ifs and maybes in there, plus yellow highlights where we will definitely have to revisit the statements later.

I also created one new drawing for the document and extracted the Plan View (Illustration 3.1) from the composite drawing (Outline water system V2.4.odg) to make the whole thing easier to manage. I can do the same for any of the other drawings in the composite if we choose to maintain them. You can find the updated drawings at http://www.hadrian-way.co.uk/WMT_River_ ... /Drawings/.

[hamishmb]

Hi,

Thanks, I will take a look soon, and make comments if needed

Hamish

[hamishmb]

Right,

I have made some changes, and attached the file to this post, take a look and see what you think

Hamish