Are RISC OS folk meanies or just poor?

I am constantly surprised at the relatively small number of doners to each bounty compared with the size of the RISC OS user base. I see that the highest ‘code number’ for registered users of the ROOL forum is now over 5000 and realise that to take this figure as an estimate of the number of active accounts would be an over estimate; but, even if there were only 1000 active accounts, an average donation of say GBP10 would get many bounties over their target amount.

As an example, look at the USB stack (step 2 of 2) bounty which has been stuck at GBP3,100 for several months with only 24 donations, that is, an average of just over GBP125 per doner. Many (all?) of those doners have been generous and the bounty needs only another GBP800 to reach target.

I am aware that there are many RISC OS folk using old stone age machines who may not be interested in bringing the OS up to date in areas like TCP/IP stack or USB stack, just as long as their Archimedes A3010, or whatever, gets updates for applications that pre-date the internet.

Comments?

registered users of the ROOL forum is now over 5000 […] even if there were only 1000 active accounts,

If you look, there are four types of user:

• Spambots
• Lurkers
• People that used to take part but no longer do
• Active users

It’s the latter you’re talking about, and I rather suspect you’re an order of magnitude off in your estimation.

On a positive note, it looks like someone recently turned up with some (rather large) cardboard cheques and Filing system improvements (step 2)
is now over it’s estimated target of £6,750, RISCOS General has got a £4k boost and TCP/IP stack overhaul (Step 2 of 4) got a £1k boost.

Does anyone know who the developer was who tackled Filing system improvements (step 1)? I wonder if they know about the amount now available for the second part (£7190.72) and if can they be persuaded to take on part 2…?

Does anyone know who the developer was who tackled Filing system improvements (step 1)? I wonder if they know about the amount now available for the second part (£7190.72) and if can they be persuaded to take on part 2…?

I suspect some (most?) of the work is already done for the iMX6

Perhaps Steve Revill or another member of the ROOL team can give us some idea of the number of ROOL forum active users, excluding Spambots. As for non-active lurkers, what are they hoping to get out of this forum?

Personally, I don’t like the bounty system too much because of “if a bounty is closed (because it is not taken/does not reach a sensibe amount of money), the money collected is distributed onto all other bounties”. While I would really like to see filing system improvements, I don’t like the idea of financing “ARMv7 inline assembler in C compiler” and I strongly disagreed with “TCP/IP stack overhaul (Step 1 of 4)”.

My solution to this dilemma was to instead donate 20 UKP every month for “RISC OS General”, because that bounty is always worth the money.

The “code number” is just a unique integer, i.e. it can never be re-used even if the account is deleted. I have deleted hundreds of accounts of spammers, so the number of accounts is far less than 5700.

I suspect there are a fair few semi-active users like me who are impecunious pensioners, too – and for whom RISCOS is a fairly minor part of our lives these days. I use !Draw on a Pi connected to a big screen. It works, and I can convert drawfiles to SVG to put on other systems. The other stuff is interesting but (sadly) of no real consequence to me.

Yeah, you could probably tap me up for the odd tenner, but it really would be the odd one.

I suspect there are a fair few semi-active users like me who are impecunious pensioners, too

That’s my excuse for only pledging a mensual fiver – but if everyone did that …

That’s why I mentioned the “fairly minor part of our lives these days” bit. There are other calls on my definitely finite disposable income, and I’m sure that’s not just me.

Guilty, I’m afraid. Not a meanie. Well, not particularly so I hope. Really, I’m not sure what the opposite of disposable income is but I have that. Too many things vying for limited resources. Messing around with RO is about the closest thing I have to a hobby. Wish I could be more helpful :(

but, even if there were only 1000 active accounts

Hum, remove one 0, and I think you’ll be closer to reality. And remove one another for the number of active core developers :)

We’re simply not enough.

At the time of writing, the main forum index has “5,495 topics, 76,723 posts, 960 voices”.

While it is possible that people may have registered an account and not posted anything, I suspect they are outnumbered by the people who posted something in the past but have since moved away from RISC OS.

So, the number of active forum users is certainly below 960 and could be well below 500.

Now let’s look at the open bounties. They have raised £18,000 from 159 donations. Some people will have donated towards more than one bounty, so my guess is that 20% of the (presumed 500) user base have donated. While I am not an expert on raising money (as my bank balance all too readily attests) a 20% response rate doesn’t sound that bad to me.

To put Stuart’s numbers into some perspective: https://www.crowdfunder.co.uk/greencard4europe

They are campaigning to have UK citizens in Europe able to have an EU green card to assure their rights continue after whatever Brexit actually comes to pass.
This is an issue that affects potentially millions of people ¹, and they made their target as follows:
On 17th April 2018 we successfully raised £4,830 with 212 supporters in 42 days

Many people could be affected (orders of magnitude more than have ever heard of RISC OS), and it ultimately took more people to raise about a third of the amount. By my reckoning, that makes RISC OS users more generous.

But – do not be misled by a simple statistic. I know there are certain people who buy stuff as their show of support, like, every show all the time. Some people buy things for others. And some get involved in work that doesn’t really have a sensible currency symbol attached (the redevelopment of the user guide, for instance). There is a lot more going on than is indicated by a simple bounty scheme, so it isn’t fair to call users mean or poor ².

Plus, as has been indicated above, there are some who don’t like the money redistribution of a bounty expires. At this moment there is £7,000 odd for filing system improvements, and no marked expiry date. 64 people have donated because it is something they want to see done. If it isn’t, the money is given to other open bounties and the desired improvements will never happen – rather like buying a Spectrum Vega+ then. :-p

¹ Far too many expats around here with the attitude “they won’t kick us out, we’re British”.

² I have pretty much zero disposable income, but I don’t consider myself “poor” as my metric for such an assignation depends on a lot more than the bank balance (kind of person you are, kind of live you live, etc etc).

TBH, I’m pretty sure I’m just a meanie.

But that isn’t to say that to my mind there are not problems with the ROOL bounty system, and after approximately 8 years I think it might be time to take a step back and have a look.

From my experience of other bounty systems on other open projects, here are some things that stand out;

1) Bounty schemes in general can be divisive on projects which accept voluntary contributions. For example what makes one volunteers work worth money whilst another is denied it?

2) Bounties work best when they are for small pieces of work that could be completed in a few hours (e.g 2 – 20 hours) by a potentially large pool of developers. There has to be the opportunity for someone to attempt the work, fail, and not be left significantly out of pocket, or feel they’ve wasted too much time.

2a) It appears the ROOL use bounties for large ‘roadmap’ items with enormous requirements on peoples time (the guide prices are based on minimum wage (7.83/hr * estimate of hours needed), at those prices the lowest listed is 380 hours (10 weeks of full-time work), the highest, 1800 hours (48 weeks of full-time work). ROOL is the only project that I’m aware of that uses bounties in this way, and it’s not really working.

2b) Even some of the enormous bounties are just step X of Y.

2c) Step X of Y bounties have led to some having ‘interesting’ Step 1 bounties, where the easiest work or most tangential work has been placed in it, with little of the future benefits of the work being available in the first drop. So far no Step 2 of Y bounties have been delivered. Step X of Y bounties realistically limit you to the one developer that completed Step 1.

2d) Bounties that contain simple small goals are aggregated into ‘super’ bounties instead of being listed individually. The perfect example of this is the ‘Paint improvements’ bounty, which realistically should have been ~20 bounties that people could individually choose to contribute to the features that they considered the most important.

3) What becomes a bounty on ROOL, who works on it, who sets the ‘guide’ price, who signs off on it being completed appear to be the same people. Leading to little community input into features that people actually want. And possibly leading to the argument that it’s not about getting community involvement in working on issues, but about the people running the bounty scheme wanting to get paid.

3a) In no way should ROOL ever get to close a bounty on which people have contributed money and redistribute it to other bounties. This is not ROOLs money to choose what to do with, but a potential payment from a user to a developer for a specific item.

4) There shouldn’t be a ‘RISC OS General’ bounty. I perfectly understand the wish of ROOL to raise money to cover costs and expenses, however bounties should be reserved for things with obvious and tangible results. The general bounty should be replaced with a separate way for people to contribute to these costs.

4a) In general ‘for profit’ companies such as ROOL (and RISC OS Developments for that matter) should be very open when they ask for donations, and should explain, in detail, where that money is actually being used. End of year accounts for small companies such as these ¹ are not detailed enough to explain it. ²

5) In 8 years only four code related bounties have completed. This is not a good record.

Whilst all of the above points are interesting it’s not really why I don’t contribute anything, realistically I just won’t benefit from any of the results of them. It’s such a long time since I had to use RISC OS for anything essential. I’ve also never been one to just hand over money to “Support the market”, there’s been 20+ years of that from people and it’s seriously warped the few remaining commercial players sense of market value.

¹ https://beta.companieshouse.gov.uk/company/05852060/filing-history
² Note, I’m not actually suggesting ROOL is misappropriating money here, just that there is no way to tell.

There shouldn’t be a ‘RISC OS General’ bounty. I perfectly understand the wish of ROOL to raise money to cover costs and expenses, however bounties should be reserved for things with obvious and tangible results. The general bounty should be replaced with a separate way for people to contribute to these costs.

I don’t think this is a true bounty per se, but rather just re-use of the existing system, to cut down on development time.

I’m in the same boat as Peter on this, though I have donated to one bounty. I wonder how development of the OS will coexist with RISC OS Developments, especially in the light of the recent announcement given that some development tends to be done in the open and other development behind closed doors.

I’m sure it’ll all be revealed eventually. Something so major doesn’t happen without a plan.

It’s great to have a bit of debate on this topic. Please come and ask these questions and make suggestions at the London Show. Here are some of my quick observations:

• For example what makes one volunteers work worth money whilst another is denied it?

We typically select the first developer who volunteers. If they later pull out (which happens more than you’d think) then we have to find a way to complete or restart the work. This is easier said than done in such a small community. We have been known to do it ourselves just to get the bounty over the line.

• Bounties work best when they are for small pieces of work that could be completed in a few hours (e.g 2 – 20 hours)

A few things here:

The small things tend to happen anyway, one way or another. The bounty scheme exists to try to move mid-sized chunks along that are simply too big for any developer to reasonably contemplate doing without at least some compensation for their time/effort. We recognise that we cannot address the really big ticket items (and even these mid-sized ones struggle, in all honesty) because the community simply isn’t big enough to fund them. That’s basically why RISC OS Developments formed.

In addition, the benefit of a ‘few hours’ task is usually so minor, that the burden of setting up a bounty, collecting a few tens of pounds, negotiating with the claimant, vetting the results and sorting out the payments would make the scheme unworkably burdensome.

Anyway, the most important fact is this: if we only focus on the small things, RISC OS will die. End of story. You can keep fettling with minor niceities but that won’t help you when all SD cards are too large for FileCore or ARM retires the 32 bit architecture, or websites all stop using SSL or they finally switch off IPv4, etc.

• It appears the ROOL use bounties for large ‘roadmap’ items

Our bounties are the mid-sized items that we believe are valueable to RISC OS or otherwise interesting to the community. There are many reasons for us doing this, as I’ve attempted to explain in my other answers.

• Even some of the enormous bounties are just step X of Y

Correct – because if we said “here’s a thing that needs £50,000” we all know it’s simply never going to happen. So we break it down into (hopefully) more manageable chunks.

• What becomes a bounty on ROOL, who works on it, who sets the ‘guide’ price, who signs off on it being completed appear to be the same people. Leading to little community input into features that people actually want

Who works on a bounty is typically the first developer who volunteers. Given the number of active developers in our community, is it any surprise the same people come around again and again?

Yes, we do decide what bounties we publish and we make a judgement on their scope; I think we’re in a reasonable position to do so, given our background. This may sound pessimistic, but if we relied upon community input into everything it would: a) take ages b) be a lot of work to manage and c) probably not get done at all. We do consult the community for what should become a bounty, both at shows and in our wish lists forum.

And we sign-off completed bounties in the same way we vet code submissions into the repository – this is one of our primary roles, after all. We want to ensure that things are of sufficient quality, consistent with what’s gone before and maintainable in future.

• No Step 2 of Y bounties have been delivered

The step 2 filing system development bounty was marked as underway today. This work is of fundamental importance to RISC OS continuing to be viable in future and I can say with some certainty it would not have happened if we hadn’t raised a significant amount of money to support the work.

• Bounties that contain simple small goals are aggregated into ‘super’ bounties

Yes, because we need to balance the administrative burden of managing each bounty against the value it delivers. If you guys are volunteering to help run the bounty scheme, we’d be more than happy to increase the number of smaller bounties on offer.

We are wary of having too many open bounties at any time; that could have the effect of diluting the donations so much that none of them ever reach their targets.

• In no way should ROOL ever get to close a bounty on which people have contributed money

First of all, we don’t expire bounties at the drop of a hat. We typically consult with the community long after the point where it’s become clear it will never reach its target.

Secondly, we made a very deliberate decision that it was far more manageable to re-distribute a bounty across the others (excluding the General bounty) when it’s expired than attempting to re-pay everyone who contributed. Anyone who donates should be aware of this.

If you’re wondering why we made that decision, think about what happens to people – they move on, change email address, change name, change bank account (we don’t store those details anyway), die or otherwise become unreachable. We’re not overly interested in spending all of our (unpaid) time dealing with that sort of mess. Sorry.

• In general ‘for profit’ companies such as ROOL (and RISC OS Developments for that matter) should be very open when they ask for donations, and should explain, in detail, where that money is actually being used

RISC OS Open is a not for profit organisation (if we ignore the side of the business that provides professional software development services to people). We are fully transparent about where the bounty money goes – it says in the descriptions and you (generally) get to see the entire source code when it’s delivered!

Our accounts are prepared by our accountants exactly in accordance with English law and what’s required of us by HMRC. If you want more detail than that, then it would cost a lot of (expensive) accountant time to produce – who exactly is going to pay for that?

• In 8 years only four code related bounties have completed. This is not a good record

Correct – it’s a disappointing record and not a good sign for RISC OS’s future if that’s how things continue.

There are probably only a handful of active developers and RISC OS is complicated so not all of them have the skills to work in all areas. Typically, these developers have actual lives and families to worry about.

If RISC OS users want to see RISC OS survive, then it’s rather unfair to expect that to happen by effectively free-loading off the generosity of those few developers. Either get active in supporting RISC OS in some way, or reach into your pocket and show some support for the developers, or sit back and watch RISC OS slowly die.

If RISC OS users want to see RISC OS survive, then it’s rather unfair to expect that to happen by effectively free-loading off the generosity of those few developers. Either get active in supporting RISC OS in some way, or reach into your pocket and show some support for the developers, or sit back and watch RISC OS slowly die.

Well said.

From next month I will do a monthly transfer! However RISC OS is almost a fulltime job for me, and for certain reasons I have to do it for free. I hope my code reaches quality enough to be part of RISC OS at some point. That would be reward enough!

Anyway, the most important fact is this: if we only focus on the small things, RISC OS will die. End of story. You can keep fettling with minor niceities but that won’t help you when all SD cards are too large for FileCore or ARM retires the 32 bit architecture, or websites all stop using SSL or they finally switch off IPv4, etc.

So, someone has to focus on the big things. Who is going to do that, or who is going to coordinate that?

If RISC OS users want to see RISC OS survive, then it’s rather unfair to expect that to happen by effectively free-loading off the generosity of those few developers. Either get active in supporting RISC OS in some way, or reach into your pocket and show some support for the developers, or sit back and watch RISC OS slowly die.

This isn’t really a bounty issue, I know, but I think it would help if there was a roadmap people could look at. It’s fine some people saying that they don’t want to make promises about things that may or may not appear. However, that’s a direction that ends in someone saying something like, “Support our subscription scheme and you might get something you like.” It’s harder to justify supporting something if you can’t see the benefits.

There’s this roadmap document from June 2017 and it feels like there’s been quite a bit of progress since then. What with recent “Developments”, maybe we might see an update soon?

So, someone has to focus on the big things. Who is going to do that, or who is going to coordinate that?

The impression that I got from Steve’s comments was that he thought that ROOL were doing that, leaving the little stuff to end users.

If RISC OS users want to see RISC OS survive, then it’s rather unfair to expect that to happen by effectively free-loading off the generosity of those few developers.

I might have said this before, but as the London Show approaches, I’ve resigned myself to the fact that I’ll be selling the same CDs with the same versions of the same software that I had at Wakefield – and justifying that on the basis that it’s at the other end of the country.

I think Steve’s point is potentially wider than just the ROOL bounties. It’s true that I’ve been busy away from RISC OS since April, but when I have had time to spend on the platform, it’s been spent organising shows or doing admin-type work for WROCC. These are all things that non-programmers could do, but it seems that those non-programmers are mostly (and with some notable exceptions) happy to sit back and watch a potential software developer do them instead of writing software.

when I have had time to spend on the platform, it’s been spent organising shows or doing admin-type work

This is a real issue and not just confined to hard pressed developers.

I used to organise and help run shows and know the effort that is required and in the end we had to agree not to do it as the numbers coming to the show went down and we couldn’t justify asking developers/companies etc to attend and waste their time for little gain.

I think we can all agree that small numbers in the community means pressures on all sides and sometimes hard decisions have to be made.

I try to do my bit by assisting in any feedback on software issues where I can , though I know I could do more but getting things organised for the MUG club and life in general intervine.

So to sum up thanks to all those who put themseleves out it is really appreciated and even if you only attend a show , make sure you buy something, or go to a user meeting and make sure you say thanks to the guest speaker.

But keep asking yourself can I do more.

But keep asking yourself can I do more.

Can you?

Seriously though, it’s hard to say what would help. I can’t really put any money toward a cause. What else can one do? Being on a different continent rules out a few things.

I’d love to write some helpful apps and utilities, however I’m pretty lost when it comes to implementing a front end. What can an average person do?