RISC OS on the Raspberry Pi

Thanks to Chris H and Theo M for replying on the Risc OS build for Pi. Sadly I don’t understand enough about compiling OS builds from sources to be able to follow. Can you advise how I can start off with a blank SD card and get up to the current state of the art – in baby steps please? An SD card image from Adrian would be great!

I built my RS232 breakout board, and made up a cable – then just at the moment I was about to power up to try it under Debian, I realised – this is a 3V3 level ARM chip!!! I could have fried it with my MAX232IN chip. Now have a MAX3232CSE chip on the way to swap in, meanwhile I may try using simple resistor networks on the serial TX/RX lines to make this 5V0 board safe for the 3V3 pins in the GPIO!

My goal is to be able to control Risc OS directly with a serial mouse and be able to test it without waiting for USB support.

Any chance of explaining how to do/get an RS 232 C level shifter board? There seems no information on this. My first step when I get my Raspberry Pi will be to get a serial terminal connected and try RISC OS in debug mode.

Would a SLIP connection to a machine sharing the internet and providing keyboard and mouse via VNC (and maybe even storage) be a possible future step, prior to the USB drivers appearing?

We don’t have access to any sources either yet, so we can’t compile at the moment.

On the subject of serial cables, if you plan to connect to a PC/etc with USB I suggest buying a USB-serial cable with support for 3.3V signalling, such as one of these:
http://www.ebay.co.uk/sch/i.html?_nkw=%28cp2102%2Cpl2303%29

Those should just clip straight on to the RPi board. If you want to drive from a real RS232 port there are some voltage converter boards available:
http://www.ebay.co.uk/sch/i.html?_nkw=rs232+ttl
Make sure you get one with 3.3V support.

For building the circuit yourself there are plenty of circuits around the net… the phrase you want to search for is ‘RS232 level shifter’. For example:
http://elinux.org/RS232_Level_Shifter
Many level shifters are aimed at 5V signalling, so you need to make sure you have the right chip for 3.3V. If you really have to, a resistor divider works – my router has 3.3V but I use a 5V mobile phone cable plus pair of resistors on the output.

SLIP/PPP is possible, but you would end up multiplexing an awful lot through one very slow serial connection. Storage is a more critical issue than internet access at the moment, and storage over serial would be even slower than BBC Micro Econet.

SLIP/PPP is possible, but you would end up multiplexing an awful lot through one very slow serial connection. Storage is a more critical issue than internet access at the moment, and storage over serial would be even slower than BBC Micro Econet.

Hopefully SD storage isn’t far off, so it would only be needed to get files in and out of the system, and until then I suppose working from RAM disk with a live image would be needed.

But obviously it’s more a testing scenario rather than a new main computer, until USB works.

I don’t think a USB to serial is much use as RISC OS won’t be driving the USB ports to start with – hence the need for a level shifting board connected to the GPIO header with a debug version of RISC OS (with input and output directed to the serial port).

Chris, unplug the USB to serial converter from the Pi and turn it round, connect to I/O header, Instant serial interface to another computer…;-)
That just leaves the software to write to talk to it …

Excellent thanks. The debug version of RISC OS does the necessary talking to the serial port already. Do you recommend any particular USB to serial convertor or will any work please?

Just had an email from Farnell. Mine is due the end of June.

So I got my serial board made up with the MAX3232 chip, wired it up using +5V, GND, TXD and RXD, but I am struggling to test with a serial mouse under Linux (Debian).

I can’t see a file called “ttyS0” in my /dev folder. Does this mean my serial board is bad, or do I need to do more to find the file? I already made sure I’m using root access, and for good measure made sure I’m added to group ‘dialout’.

Have run out of ideas from google searches!!!

wired it up using +5V, GND, TXD and RXD

Please tell me this is a typo and you’re not really using +5V?

The RPi is NOT a 5V chip (nothing has been 5V for some years now), and, if you try to apply 5V, you may damage something.

Noooo…… just using the 5v on the GPIO to power the MAX3232 chip only… whole point of getting that chip was to be careful with the 3.3V logic levels…

Sadly I am not having much luck with testing this under Linux – Debian image.

Story so far:

1. inputattach —microsoft /dev/ttyS0 OR
2. inputattach —mousesystems /dev/ttyS0
   inputattach: ‘/dev/ttyS0’ – No such file or directory

NOTE: Actually search revealed the RasPi uses /ttyAMA0 for serial line

1. inputattach —mousesystems /dev/ttyAMA0
   inputattach: can’t set line discipline

NOTE: Need to set line discipline… how? More googling:

1. ldattach MOUSE /dev/ttyAMA0
   ldattach: cannot set line discipline: Device or resource busy

NOTE: why is it busy? Googled the error. Need to use lsof to find out.

1. apt-get install lsof
2. lsof /dev/ttyAMA0

NOTE: This tells me 3x that command getty is using /dev/ttyAMA0, getty is PID 825

#kill 825

NOTE: getty restarted as process 1002

Followed these great notes at http://www.irrational.net/...

_I wanted to use the Raspberry Pi’s UART for my own purposes, not as a serial console. To achieve that, I did the

following.

First, I made a backup of the /boot/cmdline.txt file, which contains the kernel parameters:

sudo cp /boot/cmdline.txt /boot/cmdline_backup.txt

Then I edited it:

sudo vi /boot/cmdline.txt

Originally it contained:

dwc_otg.lpm_enable=0 rpitestmode=1 console=ttyAMA0,115200 kgdboc=ttyAMA0,115200 console=tty1 root=/dev/mmcblk0p2

rootfstype=ext4 rootwait

I deleted the two parameters involving the serial port (ttyAMA0) to get the following:

dwc_otg.lpm_enable=0 rpitestmode=1 console=tty1 root=/dev/mmcblk0p2 rootfstype=ext4 rootwait

I rebooted (sudo reboot) to confirm that kernel output was no longer going to the serial port. But the serial

console was still available. So I edited /etc/inittab:

sudo vi /etc/inittab

I commented out the following line:

2:23:respawn:/sbin/getty -L ttyAMA0 115200 vt100

Finally, I rebooted again and confirmed that nothing was touching the serial port anymore._

I then did ldattach and inputattach… all appears good, but no mouse movement.

I tried: cat /dev/ttyAMA0 -this came back with nothing (hoped for a stream of junk when I wiggled the mouse

I tried also connecting the RTS pin on the MAX3232 board to pin 11 of the GPIO (the third alternative function of this is UART0 RTS) but I don’t know how to enable this alternative functionality.

Help!!! I am out of my depth in Linux driver land…

Noooo…… just using the 5v on the GPIO to power the MAX3232 chip only… whole point of getting that chip was to be careful with the 3.3V logic levels…

If you’re powering the MAX3232 from +5V, the receiver’s logic high level is +5V. That’s the bit that can do the damage. Look on page 2 of the MAX3232’s data sheet.

Oh. Thanks Dave. I hadn’t processed the bit reading “Output voltages… R_OUT -0.3V to (Vcc + 0.3V)” so I could have output up to 5.3V over a long connection period…

have duly switched to powering MAX3232 from RPi 3.3V pin, hope no damage was done…

Now back to groping around in /dev/ to get the serial mouse working under linux

I’ve just been told by Farnell that my Pi should arrive in the ‘next two days’. As I have no interest in running Linux what do I need to do to get RISC OS running on the Pi please? Or shall I just leave the Pi in its box for the moment?

I’d recommend that you try Linux anyway. It doesn’t bite. Debian is currently the best choice.

I don’t think RISC OS will run under Linux.

It won’t – but you can make something similar under Linux – ROX

http://en.wikipedia.org/wiki/ROX_Desktop

It’s to Risc OS what OSX is to the original Mac OS… not tried it yet myself.

I should have the Pi before the Bristol RISC OS User group meeting on 9th May. Is there a RISC OS image yet please? Even one that has some bugs?

When there is one, it will be announced. There is currently not one, and nobody knows how long it is going to take. Can people stop asking that question yet please?

It bothers me that someone (and it sounds like only one) is holding on to the image to an unhealthy extent, preventing both demonstrations and development help by other people.

There was a binary in use at Wakefield. (Chris Evans, did you have access to a copy for use with David Heath’s board?)

My understanding from the brief conversations which have taken place (and perhaps Adrian Lees will be willing to clarify) is that the RPi source changes aren’t passed back to ROOL. So until they are, no one else can add to them.

If a binary image were to be made available, then perhaps that’d be some use (although limited, I guess) in the mean time. It sounds as if Adrian’s preference may be to not release his work at alpha or beta stage – but I guess we shouldn’t try to speak for him… Adrian, what do you say?

Bear in mind that some of the work Adrian is doing will require access to confidential details about the chip so there might be non technical reasons why he has not yet released his changes.

Although something else I haven’t seen mentioned is John Ballance was on the ROOL stand and also had a Pi running RISC OS so there is at least one other person working on it :-)

Perhaps there are commercial reasons why an image has not yet been released? After all a small licensing fee multiplied by a large number of schoolchildren could be a large number?