Estimating size of RISC OS userbase

Does anyone have a semi accurate estimate on the current size of the active RISC OS user base?
Expect it would include minimal numbers where RISC OS is the main machine/used in a work environment and that the majority of users are hobbyists etc
Expect it will be quite hard to put any sort of accurate number on the user base especially with the emergence of Raspberry Pi etc.
Does any one have any semi- reliable numbers?
Are we talking hundreds, thousands or optimistically tens of thousands ?
Main reason for asking is curiousness with regards to new hardware platforms being announced – eg the titanium and CJE’s RapdiO Ti / RapidO Ig and RComps’s ARMX6- and wondering what the size of the possible market is. Surely the sales if split between the platforms must be quite low.

Where are you in Aus ???? East or west
I live in Perth

Actives users?
- tens of thousands: less.
- thousands : less.
- hundreds : probably less too.

Does anyone have a semi accurate estimate on the current size of the active RISC OS user base?

At Wakefield in 2014 Steve Revill from rool said £16k of nutpi had been sold, which would be 450ish copies of that alone, plus whatever has been sold in the 2 years since

According to raspberrypi.org 5 million Pi’s had been sold by 18 Feb 2015, so by now a conservative estimate of sales would be 6 million IMHO. If one tenth of a percent of those purchasers tried RISC OS (1/1000) that alone would bring 6000 new users. Many of those would probably decide RISC OS was not for them, but if 10% of those who experimented decided to persevere that would bring 600 new users, which tallies with Steve Revill’s nutpi sales. Then of course you have the emulators: it’s quite conceivable that the number of people running RISC OS under emulation equals or exceeds the number of native hardware users. My best guess as to the current user base would lie at the upper end of the range 1000-5000 (but I have no firm data). R-Comp and CJE would probably have the best handle on it, as the two main RO dealers, and data from Virtual Acorn and RPCEmu would be relevant too.

It’s easy to get into the mentality that if a user doesn’t frequent the newsgroups, mailing lists and forums they don’t exist.
I’m sure we all know of other users who do just the same. RComp and CJE exsist, and produce new products, because customers exist.

Everyone in this forum has a number. In a glimpse I saw +/- 3000 is the highest.
I don’t know many Dutch RISC OS users, but only some of them post here, I guess 1 out of 10. It’s probably the same in the UK and elsewhere.
That means approx. 30,000 RISC OS users.

I would say the biggest issue RISC OS has with respect to casual users (who then might get more interested) is its inability to coexist easily with other systems. (Unless something has changed recently)

RISC OS needs to understand regular partitions and work like other OSes in that respect, so multiple operating systems are tidy. The hack with the overlapping partitions, although clever, is horrible.

I would think even with the way RISC OS works at present, it would be possible to make a formatter that respects the existing partition scheme and formats(and types)the first physical partition to ADFS.

This should then produce a drive that co-exists smoothly with other OSes and still works with RISC OS.

Which will then enable casual users to have it alongside other systems, whether it be a media player or a full Linux distro.

Hi all
Thanks for the replies.

@ Mike Carter – yes i have been lurking in the shadows for probably 20+years with regards to RISC OS… (had some exposure in Beeb days – see below)….

@ Rob Andrews, I am based in Adelaide.
Used to be a member of Beebnet (local user group) back in the day and also ran Southern Been PD for a number of years.
Was also a member of 8-bit software for a while.
Still have a few beebs packed away in the spare room.
Used to use my main RISC PC up until 3-4 years ago to do invoicing/ spreadsheet accounts etc – running my computer hardware business – DISKS DELIVERED (used to support RISC OS machines as well as PCs – mainly Risc PC though as that is what i had)
Have a few RISC PCs left in various states of assembly….have sold off my A5000 and A4000 machines…. thinking about reducing my inventory of other items… have some expansion cards etc…..currently have some SIMMs on flea bay…. Have a few Raspberry Pi and a Beagleboard here somewhere as well….
Now mainly use windows (7 – don’t want to know about 10 and its big brother tendencies) as consolidated on same platform as my cad software.
Also dabble with Linux (Mint) for Folding@Home and Thea Render.
Finally re-assembled one of my RISC PCs and installed the 5.20 ROMs the other day (instead of working….)… now the 5.22 ROMs have arrived too !

@ George T.Greenfield – 1000 to 5000 seems realistic, but obviously hard to measure and also a very small market – but has its advantages – everyone seems pretty friendly (now that the spats I used to see on the newsgroups seem to have disappeared) – also have direct contact with developers – ROOL responded to 2 emails the same day this week! and have semi-recently dealt with Sprow with a TV card update…. took a long time…. but when saw his new board released at the end of last year realised what else he had been up to….

@ Peter Scheele – seems a bit optimistic, but if was in the order of 30,000 then would quite possibly be able to sustain some commercial ventures such as CJE and RComp (although I wonder if they also supplement their incomes from other activities).

@ Jess Hampshire – indeed – i think RPCEmu deals with this best.
I am about to put a Raspberry Pi B 512Mb into a ‘Fairywren’ mini ITX board and if it works OK with RISC OS would like to install it in My Phoebe case – of course would love to multiple boot the RPi though with say Raspian /RISC OS /OpenELEC etc

OK… that’s more than enough for now… back to my real work…
Cheers
Jock

For what it’s worth, I looked at the stats for my employer’s website a while ago and found that RISC OS users outnumber Amiga users by about 2:1. I’d excluded myself from those stats. I’ve also noticed that Otter Browser reports itself as being Unix, so anyone using that is skewing the stats.

I could take another look and get some actual numbers, but there’s probably little point.

Hi Chris,
Yes would be interested.
Presume that you’d be collating hits from Netsurf.
(although lots of people probably surf the net on PCs – like me using firefox)
Who is your employer ?

FWIW, the number of votes in this year’s awards was just over 160 IIRC (the RISC OS machine’s not on ATM, so I can’t check the exact number.)

That exceeds last year by quite a bit, and is only marginally better than the year before, but I’m sure the number of active users is higher. As someone said above, Steve’s talk at Wakefield a couple of years ago suggest more, even if some of the Nut Pi sales were doubled up. (IIRC I’ve bought it twice, for example).

I could take another look and get some actual numbers, but there’s probably little point.

Extracting numbers from the logs is going to be a nightmare due to the format. Normally we just use Google Analytics but it doesn’t catch most RISC OS browsers due to its dependency on JavaScript. Interestingly last month the site had a visit from a “RISC OS-NC 5.13” user!

Who is your employer ?

If I tell you that then you’ll pounce on the site and skew all the stats :)

Normally we just use Google Analytics but it doesn’t catch most RISC OS browsers due to its dependency on JavaScript.

If anyone like me has ever visited your site, you’ll get no indication of it at all from Google Analytics – it’s blocked at my router and, when I’m elsewhere, it’s not whitelisted for Javascript.

Persuade the company to abandon Google, and invest in some software for analysing the actual site logs on the server.

My router isn’t smart enough to block anything, but on my PC it is in the hosts file as 127.0.0.1 and I run NoScript aggressively. On my phone, Firefox + Ghostery.

In short, no unknown or third-party scripting is permitted. In fact, pretty much the only third-party thing I permit is the avatars on this site.

and invest in some software for analysing the actual site logs on the server.

Analog is free, isn’t it?

I used to use it on my site’s logs, until it became boring realising that Yandex.ru was my only friend… :-)

No, seriously, I’m just not that interested in demographics. People will visit my site, or they won’t. The end.
(but then, I’m not a business)

Persuade the company to abandon Google, and invest in some software for analysing the actual site logs on the server.

The logs can’t give you all the information that companies are interested in, a lot of that information needs to be accessed through javascript. Using Google for everything is, in my opinion, a questionable thing at best and there are alternatives but good luck convincing the business side of the of that.

The logs can’t give you all the information that companies are interested in, a lot of that information needs to be accessed through javascript.

Which is better?

• Some information on all visitors
• No information on some visitors

And when considering that question, bear in mind that with increasing awareness of issues surrounding tracking – not to mention malvertising – the number of people blocking Javascript is only likely to increase.

FWIW, my hosts include log analysis (as well as access to the raw logs) as part of the package, and it seems pretty comprehensive.

Using Google for everything is, in my opinion, a questionable thing at best and there are alternatives but good luck convincing the business side of the of that.

Quite so.

Other ‘analytics’ services exist.

And can also be blocked. ;)

I don’t want to hijack the thread so one last answer.

Which is better?
Some information on all visitors
No information on some visitors
And when considering that question, bear in mind that with increasing awareness of issues surrounding tracking – not to mention malvertising – the number of people blocking Javascript is only likely to increase.

Nothing says you can’t have both, but in reality very few users turn off javascript so it’s pretty much a mute point. Plus more and more websites become completely useless without javascript.

FWIW, my hosts include log analysis (as well as access to the raw logs) as part of the package, and it seems pretty comprehensive.

For an e-commerce site these logs alone would be pretty much useless except for the most basic stuff, modern browsers can for example provide the user’s exact location that can be used to help with determining pick-up locations for parcels, matching IP addresses to location can in some cases give a failure margin of hundreds of kilometers.

Back on topic:
It would be really interesting to have an estimate on the number of RISC OS users, this question pops up every now and then but there doesn’t seem to be a way to determine that other than simply making a list, someone interested in building a database? :)

Yikes; I wish I’d never mentioned the “G word” now!

We used to use a standalone package but replaced it with GA a few years ago. The vast majority of users don’t have it blocked, and as Gulli says above, convincing me ≠ convincing the business.

there doesn’t seem to be a way to determine that other than simply making a list

Then you have the same problem of needing users to register on the list.

I think there are a lot of lurkers.

New developments, new hardware. Either they’re nuts, or the market is bigger than we think.

Exact size? Well, let me ask you, how many people used to watch BBC THREE before it went internet-only? Ratings take a very small subset of the population and extrapolate. Accurate? Flawed? Who knows for sure. Likewise with RISC OS. Okay, granted, this is the heart of RISC OS development, but how many of the lurkers consider RISC OS to be “a system they choose” with no interest in nerdy issues, beta testing, bleeding edge nightlies, etc etc? Just because we can’t hear their voice here doesn’t mean they don’t exist; and I’m tempted to say that the fact that new hardware is in the pipeline should be an indicator – and it isn’t the cheap end of the market being developed either. Why would this happen if nobody felt there wouldn’t be buyers?

Quantitative numbers? That’s harder.

data from Virtual Acorn and RPCEmu would be relevant too.

RPCEmu has approximately 4000 downloads a year. How that actually maps to a number of users or ‘active’ users I have no idea :)

How that actually maps to a number of users or ‘active’ users I have no idea :)

I think ~5 of those a year are me ;@)

Then you have the same problem of needing users to register on the list.

Not necessarily, you could have third parties register people they know of. Problem is that then you’d need some way of identifying the registered users to avoid duplicate entries and then you’ve opened up a whole new can of worms. Just to be sure, I’m not seriously suggesting anyone compile such a database!

I think there are a lot of lurkers.

I would probably count as a kind of semi-lurker, I read the forums and Iconbar with interest and an occasional comment but I haven’t used RISC OS for a while (my Pi got another job) and my RiscPC hasn’t been itself for many years.