PiFace Problems

Hi Thomas

THAT FIXED IT!

Both PiFace came with the jumpers set in the other position - THANK YOU! (and why would they do that?)

I believe that Scratch needs an updated version to work with PiFace, is there anywhere I can download it from using my current version of Raspian?

Thanks

Keith

Ah-ha! That's brilliant news! The jumper switches tell PiFace what number it is. With both set to 0, you tell PiFace that it's board 0. By default the Python module addresses board 0 so that's why it wasn't working when you had it set to whatever-else.

I don't think you need an updated version of Scratch, you just need to enable Mesh. Here are some old instructions (I'll be honest, they need improving ). If you have any trouble then start a new thread and I'll try to help when I can.

Good luck, and have fun!

Hi Thomas,

I just installed the new Python3 version of pifacedigitalio following the directions on github: https://github.com/piface/pifacedigitalio. I'm using the latest version of Raspbian (2013-02-09) with all of the latest updates installed (via apt) and the latest firmware (via hexxeh's rpi-update script).

For some reason, my system is not setting up the permissions properly for the /dev/spidev* files. Output of 'ls -al /dev/spi*' is below:

pi@raspberrypi ~ $ sudo ls -al /dev/spi*

crw------- 1 root root 153, 0 Dec 31  1969 /dev/spidev0.0

crw------- 1 root root 153, 1 Dec 31  1969 /dev/spidev0.1

I'm getting ready to just add manual udev rules to set the permissions, but I wanted to check as to why they aren't being setup automatically. Does the new Python3 version of the Piface Python IO library setup not handle permission setup yet?

Thanks,

Andy

Ah, so I haven't! I'll add this to the install script tomorrow, it must have slipped my mind since I was already working on a machine that had the permisions set up.

For now, try running the spidev-setup script here.

Thanks for the heads-up!

Yea, I just manully ran the old script and got it working. Thanks of the confirmation. With that sorted out, both the io library and emulater seem to be working properly.

-Andy

I've updated the pifacedigitalio install script to update the spi permissions. Everything should work now, I might also add auto un-blacklisting the spi kernel module later.

Joe

Not sure if you are the man or not.

Your download and instruction has an issue at :-

It seems the python script subtracts 1 so you need to use 1 to 8

ie pfio.digital_write(1,1) # turn on

or pfio.digital_write(1,0) # turn off

cheers

Mark

Hi Thomas, I'm new to R-Pi and Pi-Face. I thus far have everything working quite well with your pifacedigitalio; but have an issue on the interrupts. Once I have called wait_for_input with loop=True I would like to be able to exit gracefully from the interrupt routine. Currently I try to set loop=false with one of the inputs but that does not work (I think for obvious reasons). Any help on how I can exit gracefully after receiving the interrupt I need?

Thank you

Sven

Hi Sven,

Currently the interrupt handler only supports looping forever or exiting after one interrupt. It should also handle a keyboard interrupt gracefully (Ctrl+C).

For some reason I never thought of the use-case where you might want to exit the loop only from a certain interrupt. I've updated the code so that instead of using a loop variable to determine whether the loop should continue waiting for inputs, it decides based on the return value of the callback. If a callback returns True, wait_for_input keeps waiting for interrupts. If False, wait_for_input should return.

I've updated this in a new branch called dev_interrupt_return but I haven't tested it yet (I'm in the middle of exams, I can test later in the week). The branch is here if you want to test the new version for me. Inspect the README for new interrupt instructions.

How are you finding the interrupts on pifacedigitalio by the way? I tried to make it as simple as possible but I've not had much feedback! In particular, how is the spartan README? I was going for brief, but informative.

Hi Thomas,

Firstly, apologies for disturbing your exams. I will most certainly test the new branch. At the moment the interrupts are doing the job I need them to do and when I hit input [1] I have simply put in a rather crude sys.exit(). It works for now. I kind of like the readme, it forces you to think about the situation.

Interestingly, the solution you proposed with the boolean return from the callback is exactly what I tried but I could not quite get it to work.

Thanks for your quick response, I'll let you know how it goes.

Cheers

Sven

Edit:

I got it working. I had to make some small changes to your code, I added the following in call_mapped_input_functions():

if int_bit == 0:

    keep_waiting = True

    continue

and I also added the following at the end of the function:

return keep_waiting.

These may not be the most elegant solutions but for now it is working. Thanks again for your help.