New Dedicated hardware For RO64 - speculations

I have no intention of dying on any particular hill. Sooner or later I’ll maybe die on some hill or other, or more likely somewhere not on a hill.

As for the precise definition of planet or dwarf planet, I give not a shit.

Pluto is a substantial agglomeration of matter, substantial enough for gravity and rotation to make it pretty nearly an oblate spheroid. You could say that of several bodies in the asteroid belt, although being that much smaller they’re a little rougher in their shape. Like Pluto, they orbit the Sun, rather than being satellites of a larger object in solar orbit.

It could have been worse: the dumb asses at the IAU initially tried to designate such bodies as “plutons” (https://phys.org/news/2006-08-iau-definition-planet-plutons.html)

I’m sure Mr Google might have told them something, even back in the early 2000s.

Any half-decent dictionary would have told them, I think. I certainly learnt the word and its meaning in my teens, if not before that.

I have no intention of dying on any particular hill.

You must be new. You’re supposed to take a position and be unshakable even in the face of evidence to the contrary. Examples? Those who believe the earth is flat, and those who believe in any angry little sky fairy.

As for the precise definition of planet or dwarf planet, I give not a shit.

Me neither, to be honest. Pluto was a planet when I was young, it’s going to continue being that as far as I’m concerned. After all, the current definitions are the ones that people mostly agree on today. How long until the likes of Ceres are proposed as being “planets” and everything gets shuffled some more?

Pluto is a substantial agglomeration of matter

That could describe anything from snails to stars, depending on your interpretation of “substantial”.

initially tried to designate such bodies as “plutons”

If I remember correctly, they tried to appease people by making a special category for Pluto…

…that Pluto then failed to qualify for.
It’s hard being “the spare”. ;)

I could have understood this decision back in the day of heavy multi volume encyclopedias, when they didn’t want to have to reprint every 5 minutes when astronomers found the 10th, 11th, 12th… planets. But haven’t they heard you can just update the web page these days? In fact a bit of javascript and you could fetch the latest planet count fro the IAU website, and always be bang up to date!

That could describe anything from snails to stars, depending on your interpretation of “substantial”.

Indeed, but “substantial enough for gravity and rotation to make it pretty nearly an oblate spheroid” rather excludes the snails, albeit not the stars.

initially tried to designate such bodies as “plutons” […] that Pluto then failed to qualify for.

It’s surely the fact that “plutons” already had another, perfectly clear and different meaning in a discipline sufficiently close to astronomy that a second meaning could cause confusion, that’s the issue?

sufficiently close to astronomy that a second meaning could cause confusion, that’s the issue?

When has that ever stopped a standards group? Look at them using the “it doesn’t divide by a thousand” logic to retcon what kilobyte and megabyte mean ¹, so that if you see KB or MB, you don’t really know which interpretation is actually in use.

¹ Yes it was technically wrong, but it has been wrong for decades, they should have just left well alone. Apart from lying cheating storage manufacturers, everything was 1024/base 2 because that’s just how computers work.

When has that ever stopped a standards group?

Good point.

if you see KB

Then it is not kB.

Problem is, there’s what the standards say, and then there’s what loads of people do. Like KM, KPH, KG, etc.
Not to mention, kilo is contradictory. It’s supposed to be capitals for multiples and lower case for fractions, like Pm vs pm (petametre versus picometre).

So, KB might be intended to be kB…

So, KB might be intended to be kB

So like KB if your default units are bytes but kB if your default units are Megabytes?

Lots of people (me included) have moved on from pretending that K can mean 2^10, M 2^20, G 2^30 etc. with increasing error as the units go up.

For most ordinary people, the sizes most commonly seen are for non-volatile storage media, and their usable sizes are never an integral power of two anyway. Memory cards may have an internal integral power of two size, but the difference is the spare blocks needed for wear levelling.

I don’t feel cheated, and I don’t think anyone should.

I remember back in the 90s, boxes of 3.5" disks would say something like “Formatted for IBM 1.44 MB. Reformats to Mac 1.4 MB.” You might wonder where the extra 0.04 MB went, and the answer was “nowhere”.

The actual storage capacity was 1474560 bytes on both platforms. Divide by 1024 and you get 1440. The “PC industry” referred to that as “1.44 MB”, whereas Apple divided it by 1024 a second time and called it a “1.4 MB” disk.

The other explanation I heard surprisingly frequently was that the “missing” 0.04 MB on the Mac was taken up by the icons!

I remember back in the 90s, boxes of 3.5" disks would say something like “Formatted for IBM 1.44 MB. Reformats to Mac 1.4 MB.”

Are these the same 3.5" HD floppies that would format to 1.6MB on ADFS? So where did the extra 200K go on the Mac?

Are these the same 3.5" HD floppies that would format to 1.6MB on ADFS?

Yes.

So where did the extra 200K go on the Mac?

Dunno. Error checking with correction? (I don’t think so.)

ADFS used larger sectors from D format on.

ADFS 1.6MB floppies used 10x 1024 byte sectors, where as 1.44MB DOS used 18x 512 byte sectors. Floppies are soft sectored which means the track is a continuous magnetic surface and you can chose where you leave gaps between sectors, and with fewer larger sectors ADFS left less gaps and so could store more data. You could have a format with just one huge sector per track to squeeze a bit more out, but small writes would take a lot longer.

See https://en.wikipedia.org/wiki/List_of_floppy_disk_formats

So where did the extra 200K go on the Mac?

That was literally answered in the message that showed the differences. Mac told the truth, PCs didn’t.

ADFS used larger sectors from D format on.

Necessary. The older formats used 256 byte sectors. Imagine how many gaps…

You could have a format with just one huge sector per track

In order to keep the number of floppies down to something like 40, Windows 95 used a weird version of FAT that permitted something like 1.7MB on a disc.

In order to keep the number of floppies down to something like 40

IIRC the beta was available on 5.25” floppies too.

The release was fewer disks in the end.

In order to keep the number of floppies down to something like 40, Windows 95

My memory produces 37 for one of the installs, but

1. My numeric memory is awful
2. I had by that point built a boot floppy that brought up low level networking and then ran a batch file to install from the PC with the chunky HD in the corner of the workshop.

I started this thread to investigate the optimal price/performance combination using ARM CPUs and GPUs but it has gone way beyond offtopic. Support for reading legacy floppy formats, sure, but that’s hardly the heart of the system.

Assuming ROOL has come to the conclusion that continuing to use assembler to write RISC OS is a dead end, locking it to 1987-era technologies and instruction sets making it more obsolete than retro. A rewrite in C will both mke it maintainable and also portable to new instruction sets – a rather famous RISC OS programmer agrees with this.

Even if it’s over 20 years ago since RISC OS was my only operating system, I still miss the productivity and natural work flow using it. Using equivalent Linux software I cannot achieve what I did in RISC OS. That’s why I want to save it through a modernizing project.

Welcome to Aldershot.

Not going to happen – move along, nothing to see here.

A rewrite in C will both mke it maintainable and also portable to new instruction sets – a rather famous RISC OS programmer agrees with this.

I don’t think anyone, at least not here, disagrees, just almost all understand the practicalities of it.

Even if it’s over 20 years ago since RISC OS was my only operating system, I still miss the productivity and natural work flow using it.

Funny thing is, the exact same thing you used to do is still here, it hasn’t changed. Embrace it.

Even if it’s over 20 years ago since RISC OS was my only operating system, I still miss the productivity and natural work flow using it. Using equivalent Linux software I cannot achieve what I did in RISC OS. That’s why I want to save it through a modernizing project.

The problem is that 64 bit “RISC OS” wouldn’t run any RISC OS software without emulation. Why not just emulate RISC OS on Linux and update ROX so you’ve got a RISC OS style window manager? The key difference between that and designing an all new operating system is that you’d have access to a wide range of mature, up to date software in addition to native RISC OS stuff. Updating ROX and getting RO running nicely on QEMU seem like jobs that one or two people could tackle, whereas designing a new operating system is a task for dozens. The extra work to reinvent the wheel would result in an end product that would be much less useful and no closer to the RISC OS that we’re familiar with than an adapted version of Linux.

Not going to happen – move along, nothing to see here.

…followed by two people responding to the endless bleating. ;)

I don’t think there’s anybody around here that isn’t aware of the problems and limitations so they don’t need to be pointed out over and over and over again.
But, you know, there’s this annoying thing called reality that keeps interfering with cunning plans. If somebody wants a modernising plan, the code is all there and available. Take it, pay some programmers to get it do, voilà. Easy.