Let's get started with a Pandora port

I’m not sure – RC8 shouldn’t have contained the bug that was stopping SDCMOS. When I was testing the fix everything seemed to be working fine on my Pi, but I’m using an old disc image so haven’t had to set up networking for a long time.

Sorry for this silly question… I am currently using the Raik’s live distro. I am wonsering which rom from the page https://www.riscosopen.org/content/downloads/other-zipfiles should I use to update the distro? OMAP3 or OMAP4?

Note: I have a 1 ghz pandora.

Thank you for your help.

François

OMAP3 for all Pandoras.

got again a disk error.
Seems the disk error appears as soon as
more as 2Gb is used on the SD-Card.
Shure the SD-Card os in ADFS-Format.

So, this prob is still present with the rom from 26.03.13

I have not a Idea what the problem is.
After I have rebuild my system on a fresh 32GB card only with my Pandora (format the new card and copy all the files (more then 10GB)from the old card to the new one) I have no errors. Before I have any errors but not “stable”. DiscKnight not found errors and I have no data lost…

i did the same, but, whenever the amount of the files exceeded 2Gb it came up with an unrecoverable disk error.
No matte, which type of sd-card i used.
but if i´m using a 2Gb card, no errors occurs.

I use a 32GB Transcend SD HC class 10 in the left slot. I have formatted it with HDForm in this slot … it never leave the slot since.
Before I use a scandisc ultra card. This was created on my xM. It works but from time to time I have had a “free map space” error. After restart all things fine and Discknight found no errors.

found out yesterday:
with ROM from 26.03.13, ethernet dorsn’t
work anymore. it doesnt recognise the lindy-
adaptor anymore.

is this the ROM or do i something wrong?

I’m not sure. The latest ROM works but it is possible that you must safe your unchanged networksettigs after changing the ROM and reboot. I don’t no why but ist helps often.

I forget… the Pandoras 5.20 Livedistro is aviable. ;-)

Is there a error in the Pandoras power management (under RISC OS)?
If I charge the battery to 100%, the icon is further to “charging” . Remove the power supply changes the icon to “rest time”. When I plug in the power supply e.g. 20% again, the icon changes back to “charging” but nothing will charged. The battery indicator will not change the “rest time” and the Pandora will restart with 0% battery (remember the power supply is inside).
Sorry, it’s difficult for me to describe the problem well.

My Pandora page is changed (But still under construction!). Maybe any screenshots are interesting ;-)

The Pandora TV Out Cable is aviable from stock now.

Mutter, mutter, mutter – no Pandora yet. No reply from Craig either, so might phone him next week. Mutter, mutter, mutter.

Unfortunately, there are the wildest rumors about Craig.
Maybe you should contact EvilDragon_at_dragonbox.de. Maybe he can help.

Thanks, but from what I’ve read EvilDragon has taken enough on already. Will contrinue to chase Craig first – I’m willing to consider that the rumours/speculation don’t cater for much explanation by him… which he’s withholding for some reason.

The rumors are from the English forum. I can not say anything bad about him. I keep my fingers crossed.

The TV cable is working fine but not under RISC OS ;-)

If you look at my homepage you can find a Multiboot-Image for the Pandora with RISC OS, Angström, Bodhi-Linux, Slackware and Ubuntu. Sorry the discription is only in German at this time but but why you should have it been easier than me here;-)

Hey,
I’m having a problem booting the RiscOS image from Raiks site. I have followed his instructions, but when I put the card with RiscOS on in the left slot and boot up, it sits there just showing the Pandora logo. I waited a minute or 2 to make sure.
Any ideas what I could have done wrong? If I boot to normal Pandora OS and open the SD card it shows 4 files, autoboot.txt, boot.scr, Cmos and riscos.
I have one of the original Pandoras, but I think thats what Raik uses isn’t it?
Any help would be great, really want to start mucking about with RiscOS

EDIT – I forgot to say, I’ve put the image on a 4GB card, but that shouldn’t be a problem should it?

EDIT 2 – Well, goddamn, this always happens. I’ve just tried it again by going to the boot menu and picking SD1 from there (which I had already tried, but admittedly with the older image from Raiks site) and it’s worked _
Sorry ^_

Sorry, I have not seen your post on weekend :-) To many things to do. It does not happen often;-)
You can also send me an info via my contact page. Then I see it on my smartphone without detours.
I’m not sure what the problem is. I have try the Image on my GHz-Pandora and on the First Batch (classic).
There should be no matter if the card is bigger. I will look at. I hope I find the problem. I not use this cards. I have a complete adfs-card in the right slot and the bootfiles on the left.

Does a driver for SPI on the PI involve more than accessing the Control & Status registers and reading/writing the FIFOs? I note that Geoffrey Lea wrote a driver in Basic, so presumably there is no time-critical software involved. What more would a driver have to do than setup the registers etc?

Mmm your question refers to the Pi, but this is a thread about the Pandora Port!
You mention that you read that Geoffrey Lea wrote an SPI driver in Basic.
Can you give a link to that?
I can’t find it using the Forum search.

Your question would probable best asked in that thread.

n.b. There is a very prolific ‘Jeffrey Lee’ but I don’t think there is a ‘Geoffrey Lea’

Is a hardware video scale aviable/possible comparable with the Pi e.g. if I will play Legends of magie on it. Resolution is 800×600 the Pandora works with 800×480 so only the upper screen is visible.

The OMAP chips can do hardware scaling for true colour modes, but I haven’t implemented support for it yet.

For palettised modes we could probably support it by rendering to an off-screen buffer and having the OS/video driver convert it to true colour.

For palettised modes we could probably support it by rendering to an off-screen buffer and having the OS/video driver convert it to true colour.

The route I took to solve this particular problem, was to allow most GraphicsV calls to drop through to the OS to get it rendering to DA2 (for game compatibility) and blit that on VSync to the GPU. setMode is then doubled up, so the original drops through to RISCOS and second goes to the GPU to set the same mode but in 32bit; with Update Pointer you can pass it through to the GPU to retain the hardware pointer.

In the past week I’ve updated the blitter to upscale MODE 15 and MODE 48 on the Pi and might do some tinkering with it on the Iyonix, at the London Show, to see if I can get MODE 0/15 working on a VGA monitor without any hardware upscaling. You’re welcome to use my code after it’s publicly released, if its of any use in RISCOS.

The blitter emulates a VIDC20, you probably don’t need that level of accuracy in RISCOS so can probably get away with a standard graphicsV driver and 1:1 blitter, if hardware scaling is a definite and you’re not bothered about aspect ratio correction.