Installing an 30 GB SSD on to a RiscPC

The message does say it’s been excluded. Not “forgotten”, not “missed”, but “excluded”. That saddens me.

You can write a header into server output to prevent archival. You can modify robots.txt to prevent the spider crawling the site. Our you can just request the site be forgotten. Our somebody else can if it’s a matter of copyright infringement.
That’s why sites are “excluded”. It’s a deliberate act.

It’s a deliberate act.

That’s the sad bit

Or somebody else can if it’s a matter of copyright infringement

A bit unlikely – if it was never spotted while it was active it seems odd that someone should get worried by it at this late stage!! Much more likely is something like what happened to the Virtual RiscPC pages – lots of useful data removed in what I assume was a bit of a sulk.

That’s the sad bit

A certain somebody cleaned their existence so throughly that it’s almost as if they never existed.
In terms of chucking toys out of prams, that was really batting for six…

A certain somebody cleaned their existence

While that’s probably a speculation rather than known fact it does fit with my thought line.¹
One of those things I suppose.

¹Each time I’ve got to the almost forgiven and largely forgotten stage another item comes up to cause me recall what type of behaviour triggered my previous (and previous and previous) feelings.
Oh well.

I have my A7000+ open in front of me with working CF boot from APDL (6 June 2013 £24.90 inc adaptor) It is configured to two Harddrives, no CD, one floppy as what it happens to be left at after failed attempts at another configuration.
Anyway, I have HForm on floppy. There is actually a CD drive in the first IDE position then the master CF adaptor and a floppy.
I boot then run Hform and it says 1683 cylinders, 16 heads, 63 sector detected, LBA compatible. But David Holden formatted to non-standard disc shape of 31272, 16, 63.

I purchased a while back a new HAMA 8GB CF and a new 8GB non-branded CF off Amazon. I was not expecting compatibility probs.
I know each computer system has its chracter and problems so maybe I can add mine to the record.
If I now hot swap to the cards Hform sometimes says cannot read disc identity disc error &8 or &21: HForm failed: unknown or missing variable.
I imagine that is what totally incompatible card looks like? I am a bit haphazard but I thought hot swapping lets it be read sometimes.

When it fails I then try booting with the unformatted disc. Boot gets to start-up choice menu where have to press key-X to desktop. I can then run HFORM and it does seem to see the unformatted CF as drive 4.
I now find it can read each card one as 15490, 16, 63 the other 1536, 16, 63. Are these cards then compatible or do they fail later in the format?
If I knew what configuration cylinders to set I could have ago at formatting them.
(I also have 6 old CF cards with audio recorder data I want to keep to play with.)
Anyone like to advise what settings to put in HFORM? and I will say how it goes?

errata: that should of course be 16383 (not 1683), 15636 (not 1536).

If I now hot swap to the cards

It probably won’t help but IDE drives (which CF cards really are) should never be hot swapped.

Anyone like to advise what settings to put in HFORM?

Using the cylinders, heads, sectors that HFORM suggests are best.

n.b. Some drives will error on first HFORMing but not on a second attempt. From experience these drives will be very unreliable, reporting ‘disc not formatted’ on the next (or after a few) reboot and all subsequent access attempts.

As Chris says, IDE drives should never be hot-swapped. The hardware (and OS) has no mechanism for dealing with such a thing. If you were to perform any accesses to the device, it is entirely possible that it will be garbling data cached from the (CF) disc it thinks is there, with data from the (CF) disc that is actually there.

Unless your IDE cable has a twist in it (rare these days), it should not matter whether your CD is in first or second place. What is important is whether it is Master, Slave, or Cable Select.
For now, perhaps best to eliminate the CD drive from the equation. Unplug it until the CF is working.

Disc error 21 seems to be common when trying to get CF cards working. Error 8 means the drive isn’t ready, and 21 means the drive said it was busy when given a command – http://www.riscos.org/csafaq/part3.html#3_1

Compatibility problems are to be expected when trying to get modern devices to talk to ancient hardware. I understand that some of the later generation of PATA harddiscs expect to do UDMA transfers and just don’t work in PIO mode. This is unlikely to be noticed by the majority, but I expect the combi-chip Acorn used only does PIO mode.
In a similar way, I have to try to find out what the largest size of drive my Pentium 4 box can accept. From what I have read, it isn’t possible to format a large SATA drive smaller (as we used to do with IDE drives and Acorn’s 512MB limit) because the firmware simply won’t see the drive at all if it is too big. Basically, this means seeing if 128GB or 256GB is the limit. Same deal. Older hardware meets newer hardware, and it’s like French vs Quebecois. ;-)

Unless your IDE cable has a twist in it (rare these days)

There were IDE cables with a twist? Never seen those – are you sure you are not confusing that with floppy cables? Or ST-506 (aka MFM/RLL)?

There were IDE cables with a twist?

Nope. You’re right. It’s only whether or not pub 28 is grounded that determines which drive is which in cable select mode (so you might come across a ribbon cable with a notch cut out of one of the wires).

I already knew about floppy twist, so I did some googling and I think I was mixing it up with ESDI, in which a cable twist can make two drives with the same ID behave as different drives (you can have drive #0 and drive #1 on the bus, but if they’re both set as drive #0 then it’s twisty time).

The joys of being an old fart – you can remember using technology that most people haven’t heard of. ;-)
Oh, and the use of the vinyl scratch noise in modern movies makes me smile. I wonder what millennials think that sound is supposed to be?

The joys of being an old fart – you can remember using technology that most people haven’t heard of. ;-)

Or working in production test and working on equipment with ferrite core memory mats…

It’s only whether or not pub 28 is grounded that determines which drive is which in cable select mode.

Which may have a bearing on the working (or not) of some CF cards where pin 28 not being grounded causes problems for the IDE interface in the Acorn era kit.
IIRC the RPC interface didn’t understand cable select.
A bog standard 4k7 resistor from pin 28 to ground might cure things. There is a link fitting and what may be a surface mount resistor on the CF interface David sold.

Oh, and if I recall there was a set of info by David that appeared in (I think) RiscUser mag.
The doc I have stashed with the extract starts:

Using Solid State drives
Dave Holden shows how to use the latest technology in a RiscPC…

I Have now tried to format the 8 GB HAMA and no-brand CF cards using HFORM on the A7000+ as described above. I used the settings HFORM offered for bootable long files including LFA 1024 bytes.
I said no to add to defect list and to a soak test.
They behaved similarly except at Boot the HAMA hung the machine for quite a time after pressing key-X to get to desktop whereas the no-brand was instant to desktop.
The HAMA formatted successfully but I only have data on floppy to test it with quickly. Like to try and boot from it sometime. After formatting it the still bootless HAMA also lets go instant to desktop.
The no-brand did not format: gave HFORM failed disc error 20 at address at line 1900 twice.
If I had said yes to add to defect list would it have avoided that type of error?

I had also ordered a double CF reader master-slave visually similar version of the one APDL sold but it just green screens all the time so I cannot use it to backup CF card as my hope.
Is there a knack to placing those double CF readers on the IDE bus or is the card itself not compatible? Anything I can try to make it work to backup cards?

if I recall there was a set of info by David that appeared in (I think) RiscUser mag.

It’s in “RISC World,” here:
http://www.apdl.org.uk/riscworld/volumes/volume9/issue6/ssdrives/index.htm

The joys of being an old fart – you can remember using technology that most people haven’t heard of. ;-)
Oh, and the use of the vinyl scratch noise …

Vinyl? Wax cylinders, 78s.

I remember the 11 plus question “What is the next number in this sequence? 16 33 45 ?”.

But then, I can remember Roy Rogers and Trigger on-stage at the Liverpool Empire. Why this is so, I have no idea!

It’s in “RISC World,” here:

That’s the one.
The APDL technical information was more detailed as I recall, but the RiscWorld version does have the advantage of being on CD’s in peoples homes and as safe as information on a CD can be.
Interesting that the RiscWorld back issues are on the web on the site that seems to have asked for the redaction of the other.