Atrix Lapdock -> RISC OS Portable

Not being up with all the things to do to install RO, do I need to wait for the imminent release of the ROOL image, or would some kind soul guide me through the process with what is available now?

1. Download the 8 Aug alpha SD card image - note this will
   disappear once the beta image appears.
2. Write this onto an SD card using Win32DIskImager
3. Replace the following files in the FAT partition:

4. Add the file 'HardDisc4.util' to the FAT partition (from 'software')
5. Edit the file 'config.txt' - add '#' before the lines starting:
   'hdmi_group=' and
   'hdmi_mode='
6. Boot up RISC OS, shift double click on !Boot and double click on Loader
7. This exposes the FAT partition - copy the file HardDisc4/util onto
   the SD card, set its file type to Utility and expand it. Copy (only)
   the directory !Boot from the directory HardDisc4 onto the root directory
   of the SD card overwriting some of !Boot.
8. Reboot RISC OS
9. It's a bit of a hack as you should start with the new boot and add things
   to it, not the other way around. Copy a few files from 510Hook to PreDesk
   to get sound working.

I am struggling with this, as I am a bear of little brain.

1. I had to twig that it was your SD image, only linked to way back in the other thread, not a disc image on the ROOL wiki site.

3. The ‘riscos.img’ name does not refer to some other SD image, but the name of the rom, which is just ‘riscos’ in the ROOL zip.

3. I am not sure what to do about the other 3 files to replace. If I follow the given links I am told that binaries are not available, only source. I have saved those same files from the Linux distro I downloaded at the weekend. Are those sufficient?

4. Is ‘HardDisc4.util’ inside the tarball? What is the recommended program to read that file on XP?

I should be OK if I can get through to #5. ;-)

On the other hand, I think I may wait for the official distro after all.

@SD
HardDisc4.util – self extractor file that produces the whole of !Boot extracted into a directory labelled HardDisc4

Boot off the SD card ROM and create a RAMDisc to place the file into, ensure its filetype is “Utility” and doble click to run.
Use the Format utility in the extracted sub-directories to format a boot disc and copy the whole content of the HardDisc4 directory into that boot disc partition. Twiddle various ettings for personal taste.

Or just wait a week or so for the official distro release.

On the other hand, I think I may wait for the official distro after all.

More by luck than good judgement, I think, I have got RISC OS running. Many thanks for the help.

I have some further questions.

The screen starts in X640 Y480 C256. I can set X1360 Y768 C16M, and very fine it looks. I can do this in Configure, too, but nothing is saved. I read that currently the image only supports 1920 × 1080, so how is all this happening between RO and the Pi? I can also set 1920 × 1080, and it is small, but works ok.

When I drag an object there is a trail. Is this an artifact of the way the screen is being scaled, or is it a known problem of the rom?

CH said that some files needed to be transfered from 510Hook for sound to work, but did not specify. In any case, if I try to run Maestro I get a SWI not known error. Is this connected, and should I also be doing something with the HDMI configuration?

Who needs to know that Basalt and the HongKong demo game work on the Pi? I did enter the info in one place in the Wiki.

Many thanks

The screen starts in X640 Y480 C256. I can set X1360 Y768 C16M, and very fine it looks. I can do this in Configure, too, but nothing is saved.

check the state of the first line in !Boot.Choices.Boot.PreDesk.Configure.Monitor

If the first Line begins with a | then change it to a L so that it reads (for an Acorn monitor)
LoadModeFile BootResources:Configure.Monitors.Acorn.AKF60
save the file and then change the access to protected write.
Then reboot.

Items on these forums suggest that it is the flag for hard rest being set, but the code given here to clear that bit followed by a cmos save still didn’t clear the problem for me so I protected the file and sidestepped the problem.

@andym

I’m awaiting one of these cables to adapt it into a split cable to power a Pi from the internal battery, and use the keyboard and trackpad on the lapdock for input.

Did you have success with that cable? I got one, but the power leg did not appear to actually provide power as required by the RPi. On sacrificial investigation I found that the red wire might not have been connected in the main plug. Anyway, I found the legs a bit short and made up a lead in a different way.

All we need now is some form or wire-free networking to make it truly portable. I guess a WIFi adaptor or even working 3G would be great!!!

Did you get one of these working? I am waiting on an Edimax, but reading the Pi forums makes me doubt whether I will be able to use it under RO.

To make the system as portable as possible I am sorely tempted to lighten the Pi of its sound and composite sockets. Has anyone here successfully desoldered from this board? I have a desoldering iron pump, but I am aware of the damage that can be done to multi-layered boards.

I have a working lapdock now, except for the screen mode.

Rather than attack a cable with cutters, I put a sliver of sellotape over the 5V pin on the USB data cable.

I used keymap to set the second ctrl to menu.

I added hdmi_drive=2 to the config.txt

I use the lapdock with the xM at this time. I have a 3rdBotton-Modul that set the right ctrl to menu.
With today’s ROM ComCentre working stable again with my UMTS stick.

Now I have a working laptop, I am in a position to test comcentre, (especially if USB now works properly)

On a related topic, has anyone developed a driver for a WiFi dongle, specifically using the RealTech chips?

I have had mine working with Raspbian.

I seem to remember there might have been something using the SimTec stack once upon a time.

@Jess

The revision 2 boards, and late revision 1, do not have have the two polyfuses, F1 and F2, that protect the USB power lines. This allows the Pi to be powered from a hub, ie the Lapdock, without the need for a Y cable or cutting lines. If you are handy with a soldering iron it is very simple to short the fuses with bridge wire to effect the same.

I set Alt Gr as my Menu key.

The USB hub connection appears to be powered all the time, so there would be no off switch. At least on mine. How do you do it?

That sounds as if you are taking power from one of the Lapdock’s USB ports. With a Y lead you have to unplug the power connection. With mine, I can either unlug the USB at the Pi or unplug the HDMI, which powers down the Lapdock. The latter is more convenient the way I have it.

Yes, I hadn’t looked into Y-leads, i had assumed they plug into a USB port and send power only to the second device (i.e the Pi).

I don’t wish to be unpluging stuff, and I’ll put up with the extra cable and loss of USB port (I’m still 1 up using a mouse.)

Look, no wires!

Front view showing NetSurf at work.

Rear view showing the Pi and accesories.

The HDMI adaptor is one of those from China and is joined to the Pi with a flexible link, rather than a cable.

The Pi in a ModMyPi box with a thin magnetic sheet on the bottom; the back of the Lapdock is covered in a sheet of the same so the Pi can slide up and down as the screen is adjusted without becoming detached.

The Pi has been modified to bridge the two polyfuses, so the power and USB connection is just a single lead. The lead started with a female USB, but this was changed to a USB male plug.

The red box is an Edimax Nano Router, linked to the existing LAN socket on the Pi and powered from the second USB socket. It is configured in AP client mode, but it is extremely flexible and I have not yet explored all its possibilities.

The current cables are temporary and the arrangement will become a little neater.

Should someone write a driver for a network USB dongle, I think the upcoming Model A could be about a quarter the size of this arrangement.

Nice one. Bodgery at its very finest! :)

Wait. So does this mean we have bona fide WiFi support in RISC OS now?

AIU the Edimax can be put to use as an ethernet-Wifi bridge, same as all those “gaming adapters”.
Nice job btw.

No such luck.

Technically this does no more than we have been able to do for many years – attach an access port in ‘station infrastructure’ mode to the ethernet port of a device that does not have WiFi.

The delight of the Nano Router is that it is so small, relatively cheap and can be powered from a USB port.

The current cables are temporary and the arrangement will become a little neater.

Rear view showing the new neater arrangements.

The Nano router has had the matt-black treatment. It is linked to the Pi with a back-to-back RJ45 adapter that I made (tricky).

Looks good, if you put a pi on the back of a monitor would that be an iPi ;-)

I ordered an Edimax BR-6258 on the strength of this thread. The Edimax nano router allows many different interesting configurations.

Just to confirm that it can connect via DHCP Ethernet to the RPi, and route via WiFi to a router (in my case a battery powered MiFI). This wee beastie allows me to use my RPi connected to the ’net (via my WiFi) on my canal boat ;-)

Data is via the supplied short Ethernet cable. Power is via the supplied USB type A to male Micro USB connector (i.e. the same as the MiFi and the RPi power connectors). The Edimax powers from the Lapdock’s in-built USB connectors.

Not particularly elegant, but functional :-)

Ralph

My Atrix LapDock is supposed to arrive tomorrow and I am trying to figure out the cables needed from differing versions in this thread and elsewhere.
For the HDMI connection one needs to connect the male micro HDMI port on the LapDock to the female HDMI port on the Raspberry-Pi. One way is to use a micro female HDMI to micro female HDMI at the Lapdock end plus a HDMI micro male to HDMI male cable, but is there a cable that has micro female HDMI at one end and male HDMI at the other end?
For the USB connections, some are using a splitter cable and then having to cut into one of the forks to stop the 5v line. The single end (F micro) seems to go into the Lapdock and at the ‘double’ end (both M) one goes into the micro USB power socket (F) and the other into one of the two USB data sockets (F), both on the Raspberry-Pi. What is the virtue of doing this as opposed to using 2 separate USB cables one for the power and one for data? Is it to save a USB socket on the LapDock?
I hope I have got this right, but I am trying to work from pictures on webpages and descriptions in posts. The two male USB ends both seem to be of the same size, yet on the Raspberry-Pi the female port for the power is micro and the female one for the data is ‘ordinary’ size!! Or is that where the ‘DeLock’ micro USB female to USB A male adaptor comes in?
Oh, what a mess!
I wonder whether we have a Wiki on this?

is there a cable that has micro female HDMI at one end and male HDMI at the other end?

I have not found one and I have not seen one mentioned anywhere. The alternative is a micro-HDMI female to male HDMI adaptor and a fairly standard HDMI cable. These adaptors are moderately expensive from this country, but cheap from China if you can wait 10 days or so.

Is it to save a USB socket on the LapDock?

Essentially, yes. However, if you have a revison 2 (or late revision 1) Pi you may not have to use two connections. Look at the board next to the USB socket and see if there are F1 and F2 (polyfuse) components. If not, you can power the Pi from just the one USB connection from the Lapdock.

is that where the ‘DeLock’ micro USB female to USB A male adaptor comes in?

If you go that route, yes. If you have any experience with soldering I would suggest replacing the micro plug with a full-sized (A) plug – it is not very fiddly.

Do contact me direct at steve [at] kappa [dot] me [dot] uk. I can give you pointers to some suppliers, or I may have some spares.