Bounty proposal: Paint

I don’t think the shapes, like “circle” should be actual sprites, rather they should be internally generated sprites created when you select them.

This really sounds like two different tools: (1) a scalable brush tool with translucency and a range of basic shapes, and (2) a sprite ‘stamp’ tool. To keep the interface simple, it might be best to have two buttons on the toolbox, one for each.

At this point, though, I wonder if we’re putting quite a lot of features into a single bounty proposal? Unless there’s something really compelling we’ve omitted, I’d be in favour of freezing the feature list here and seeing if we can get it published to attract some interest.

At this point, though, I wonder if we’re putting quite a lot of features into a single bounty proposal? Unless there’s something really compelling we’ve omitted, I’d be in favour of freezing the feature list here and seeing if we can get it published to attract some interest.

Yes, agreed. Any comments from ROOL?

[ Paint bounty proposal ]

This looks like a great set of proposed updates to me. I’ll review the wiki page with the other ROOL guys – we’re long overdue a review of the current set of bounties anyway.

Warning Steve’s risking going off topic (sorry)…

We love all this activity on the forums and all the great suggestions for improvements and new bounties. Those exactly the sort of vibrant exchanges of ideas that we could only dream of when setting up ROOL seven long years ago. Thanks to everyone for contributing!

Having said that, we’ve been seeing very few actual donations coming in over the past year or so, which is disappointing. For example, there were well over 2,000 downloads of the 5.20-related work in the first couple of months and if everyone downloading that had donated just a fiver as a bit of a “thank you” you can see what a potential pot we could’ve had to throw towards improvements like this. As it happened, we actually received less than £100 (around 3p per download in the end) :(

This should be a warning to all those discussing new bounties; it’s all very well coming up with great ideas and suggestions but unless there’s actually money coming into those pots, the bounty will never be likely to translate into reality and we’ll have to reluctantly retire it in a year or so. Such a shame!

I’m giving you a bit of a preview here of what I will be focusing on for my talk at the London Show next weekend – our belief that without a significant change in the amount of money coming into RISC OS, we’re never going to be able to truly break free of being stuck in the niche we’re currently in. And I’m not knocking the great work done by our volunteer contributors (ROOL staff included) by saying that!

Maybe the links on the download page could take you to a “Your download will start in 5 seconds” page. (Easy to implement with a meta refresh.) The new download-pending page can have a “Please consider donating to one of our bounties, to aid the continuing improvement of RISC OS” message. “Just a fiver from enough people makes a difference”, etc. :) And it could list the open bounties.

we’ve been seeing very few actual donations coming in over the past year or so

Hmm. That is disappointing. Perhaps there’s scope for publicising the bounties a bit more – Iconbar, csa.*, etc. I could do the former; I’m sure there’s someone else who’s happy to do the latter.

I do wonder whether something like Paint bounty might do better than, say, the inline assembler target. The average user might well see the benefit from updated apps in a way they don’t for updates to the developer tools. Will be interesting to see.

Also are there enough sells of the RISC OS USB, CD, NutPi, etc to put part of the profit towards the bounty scheme.

I am not sure what the breakdown is. Though from my angle if there is enough to maintain the servers, office space with utilities and ISP service (combined total aprox $400USD per month) and pay any full time employs $700USD per month (A minimal living wage with enough left for some extras [more than I make]) then any remaining balance could be devided between the Bounties, and the continued internal development of software.

I am aware that most people are accustomed to making $2500USD or more per month, though I see no need for this level of waste when $700USD per month is enough to support two with all the extras that you could ask for reasonably, and add an additional $200USD per additional person. I do admit that supporting two people I would like to make $850USD for more extras, though it is not needed. So even if you adjust the wage of full time employs up to 900 (assume that any person supports three [self inclusive]) this should work.

The existing bounties do not really fit with my current experience of OS5 aggravations, such as they are.

OK so !Paint might usefully do with a couple of fixes otherwise the app seems OK to me. And I could not care less about any RO6 based wish list, other than Boot logging obviously, see csa.misc.

The real RISC OS issues here are not all ROOL matters. They are the slow upload speed from the Raspberry Pi, NOOBS out of date and it would be really nice to have a stable RPCEmu for my Mac.

I have done some penance for the above, and I am very appreciative of ROOL’s achievements.

Also are there enough sells of the RISC OS USB, CD, NutPi, etc to put part of the profit towards the bounty scheme.

Doubtful, the price of those probably covers the cost of putting them out to users and little more.

I am not sure what the breakdown is. Though from my angle if there is enough to maintain the servers, office space with utilities and ISP service (combined total aprox $400USD per month) and pay any full time employs $700USD per month

Pay? Probably counts as the most wishful thinking ever.
The ROOL guys have jobs doing other things. This is private time or hobby time.

If I’d got spare money I’d be putting in – unfortunately I’m a victim of the bankers and Tories¹ so my income currently equals the bills plus a few hundred.

¹ B****rds think rises lower than inflation are a good idea. Cobblers about annual increments don’t help either – I had one in 2007…
Meanwhile they have noses in the trough.

The existing bounties do not really fit with my current experience of OS5 aggravations, such as they are.

Well, everyone will have their own priorities.

Boot logging obviously … slow upload speed from the Raspberry Pi, NOOBS out of date and it would be really nice to have a stable RPCEmu for my Mac.

For example, none of those bother me but I think automatic monitor handling would be useful. Currently you either have to understand how to edit an MDF, or get hold of a working MDF from someone else.

By the way, when I said < £100, I was talking about donations to ROOL (the admin bounty) rather than to any of the other bounties…

Ok so the sales does not provide much, this I did not know.

Though if you have a couple of hundred left after bills you are living well compared to I.

I am aware that most people are accustomed to making $2500USD or more per month, though I see no need for this level of waste when $700USD per month is enough to support two with all the extras that you could ask for reasonably, and add an additional $200USD per additional person. I do admit that supporting two people I would like to make $850USD for more extras, though it is not needed. So even if you adjust the wage of full time employs up to 900 (assume that any person supports three [self inclusive]) this should work.

You’re joking, right?

<snip lengthy breakdown of how ~€1000 per month is barely enough for one, never mind two, arriving at around €18 per day and that’s only after the regular things are paid – everything else comes out of that sum; your claimed figure is half of this ($700=€512) which is below the poverty level of numerous European countries (except the UK as the basic state pension isn’t far off this figure…). In short, $700 for two people with all sorts of extras is only viable of you’re living in your mom’s basement, walk to work, and somebody else pays the utility bills……..?>

Your $200 per extra person comes out as a mite under €7/day. That wouldn’t buy a McDonald’s meal over here… And this is talking purely food, not extras – clothing, toilet paper, shampoo, and all the other rubbish that is necessary for day to day life that has magical properties of doubling the price of a trip to the supermarket.

On behalf of RISC OS, maybe they ought to get a PayPal account set up? I’d drop a few pennies their way from time to time. Cheques are extremely impractical for small donations because I rather suspect that anything I could send would be lost in random charges by the bank exclaiming “WTF is this? It isn’t even English!”.

I’ve made a contribution to ROOL’s administrative expenses this evening.

It took a while, but I finally managed to realise where we’re all going wrong in this discussion of how to make the brush tool more intuitive.

At the moment, the only way to copy part of one sprite into another is:

• Select the ‘Copy’ tool.
• Select the ‘Export’ option.
• Draw a bounding box around the area in the source sprite you want to copy (hoping that it doesn’t require the window to scroll).
• Give the result a unique name and save it somewhere.
• Switch to the destination sprite.
• Select the brush tool.
• Deselect ‘Shape’ (if this is your first time), enter the name of your copied section (probably “export”) and click ‘OK’.
• Pray that Paint doesn’t crash immediately due to general crapness.
• Pray that neither of the two sprites have custom palettes, because there used to be (and possibly still is) a bug in the brush tool which means the colours of the brush sprite will get completely messed up.
• Pray that you don’t trigger whatever bug it was to do with one or other of the sprites having masks which caused the brush sprite to be completely transparent. (I’m basing my experiences on RO3.7 Paint here, but I don’t think RO5 Paint has advanced that far.)
• Click once over the desired area.
• Remember to clean up your temporary sprite from wherever you put it afterwards.

That is a lot of convoluted, non-intuitive steps, combined with dull housekeeping tasks which should never be the user’s job.

What it should be is this:

• Select the ‘Select’ tool.
• Draw a bounding box.
• Copy to clipboard.
• Switch to other sprite.
• Paste from clipboard, moving the copied area to where you want it.

And once Paint supports the clipboard…

• The ‘Local’ and ‘Export’ options will lose all meaning.
• The destination doesn’t have to be another sprite edited in Paint. It could be Draw, for example (once that supports the clipboard).
• You could still mimic the ‘Export’ behaviour by pasting into a sprite file window (although some consideration of naming would be necessary here).
• It will all make some goddamn sense.

Now, that’s a UI change I really could get behind.

By the way, when I said < £100, I was talking about donations to ROOL (the admin bounty) rather than to any of the other bounties…

I presume the figure shown is the total amount ever donated. It’s more than the other categories. Trouble is it looks like that’s the amount that’s in there, which can’t be true. You must have spent some/all. Perhaps as it’s a special category, being permanent, an extra box showing what’s left would encourage people to top it up?

• Copy to clipboard.
• Switch to other sprite.
• Paste from clipboard
  

I don’t really see the value in replacing RO-standard drag&drop behaviour with clipboard usage, especially when the apps in question don’t support it. Does anything support the clipboard for anything but text?

Much better if the export sprite (or any other sprite you drag in) could be dropped into a sprite editor window, and have a “Use as brush?” query window open. Or even just jump to the brush tool with the sprite ready to go, without the query.

Much better if the export sprite (or any other sprite you drag in) could be dropped into a sprite editor window, and have a “Use as brush?” query window open. Or even just jump to the brush tool with the sprite ready to go, without the query.

Wouldn’t that be easier to understand as a user if a specific area at the edge of the sprite editor window had a brush symbol and anything dragged to that area (and nowhere else) prompted a “use as brush” response.

Save the global clipboard use for other things.

You may be right that Martins simplification was still not simple enough.

Wouldn’t that be easier to understand as a user if a specific area at the edge of the sprite editor window had a brush symbol and anything dragged to that area (and nowhere else) prompted a “use as brush” response.

Chris already suggested that the Brush tool could have a drop box, for dropping sprites. That makes the behaviour clear.

I’ve made a contribution to ROOL’s administrative expenses this evening.

Likewise: I’d meant to do it when I had a spare five minutes after I’d downloaded 5.20 (in lieu of buying the CD at Kenilworth), but never remembered to find the time.

I presume the figure shown is the total amount ever donated. It’s more than the other categories. Trouble is it looks like that’s the amount that’s in there, which can’t be true. You must have spent some/all. Perhaps as it’s a special category, being permanent, an extra box showing what’s left would encourage people to top it up?

A similar thought occurred to me when looking at the Bounties page this afternoon. The implication as it stands is that there’s a lot of spare cash in the administrative bounty, which there probably isn’t?

I don’t really see the value in replacing RO-standard drag&drop behaviour with clipboard usage, especially when the apps in question don’t support it. Does anything support the clipboard for anything but text?

Yes, most sane apps. I don’t really use graphics software, but I’ve found DigitalCD’s integration of global clipboard support with its playlist data very useful on several occasions.

I didn’t suggest copy and paste should be the be-all and end-all of it. Indeed, another thing most sane apps do (StrongED, DigitalCD, and I think NetSurf) is to allow you to drag and drop a selected area out of the application, which is equally useful.

A change in RO5 Paint which I’m very fond of is the ability to select, and drag, sprites around, into and out of the file windows. In-sprite selection should allow the selected area to be dragged as well, which would both match what other apps do and permit your ‘use as brush’ draggy-droppy idea.

A large part of my suggestion had to do with making Paint friendlier to new users, who are extremely unlikely to be able to figure the many steps of the existing idiosyncratic process out for themselves.

You’re joking, right?

lengthy breakdown of how ~€1000 per month is barely enough for one, never mind two, arriving at around €18 per day and that’s only after the regular things are paid – everything else comes out of that sum; your claimed figure is half of this ($700=€512) which is below the poverty level of numerous European countries (except the UK as the basic state pension isn’t far off this figure…). In short, $700 for two people with all sorts of extras is only viable of you’re living in your mom’s basement, walk to work, and somebody else pays the utility bills……..?

No joke. The analisys you give just shows excess.

On ~$700USD per month I support two. The monthly expenses are:
Mortgage = $192USD
Gasoline = Aprox $50USD
Utils ttl = Aprox $100USD
Food TTL = Aprox $150USD
Necessities (Diapers, TP, Soap, etc) = Aprox $60USD

Everything else is gravy so to speak. I could not do this if I had to pay for medical expences, though otherwise I am living well on aproximately $700USD per month with more than $100USD for extras each month. Sometimes extra needs dip into this extra money, though for the most part that is it.

You see I do not need excess, and the figure is relistic at an aditional $200USD per person as many of these bills are little changed or unchanged for extra people. I could cut the utility bills by not using the TV, or downgrading my Internet service, or taking only one shower per person per day.

Most will have many excesses in there list, and not think about them. Not I.

Also some still rent a place because they like spending way to much, buying a house + land is cheaper.

Your $200 per extra person comes out as a mite under €7/day. That wouldn’t buy a McDonald’s meal over here… And this is talking purely food, not extras – clothing, toilet paper, shampoo, and all the other rubbish that is necessary for day to day life that has magical properties of doubling the price of a trip to the supermarket

McDonadles is expencive, out of my budget. Clothing buy one garment per month and you will keep up with what whares out. I listed necessities, that includes Diapers, soap, Shampoo, toothpaste. Just be sencible.

It only costs a couple of dolars per day to feed a person, for those things that do not grow in the garden, or you are out of canned in the non growing season.

On behalf of RISC OS, maybe they ought to get a PayPal account set up? I’d drop a few pennies their way from time to time. Cheques are extremely impractical for small donations because I rather suspect that anything I could send would be lost in random charges by the bank exclaiming “WTF is this? It isn’t even English!”.

Now I would second that. PayPal is a good service.

I would go with keeping the standard drag and drop (I never use the Clipboard), though simplifying the task.

As regards the reason for little bounty this year. I can’t speak for other people but I donated to the USB bounty and it’s still open, and no update as to what is actually happening from a ROOL perspective.
Was it a complete waste of money? Is it still ongoing? has it been done?
The answers are largely irrelevant now, I’ve given up on bounties.

As regards the reason for little bounty this year. I can’t speak for other people but I donated to the USB bounty and it’s still open, and no update as to what is actually happening from a ROOL perspective.

This has been a lesson to us, certainly. I think we might need to clarify some of the rules for claiming bounties but I also think it’d be a good idea for each claimed bounty to gain a forum thread dedicated to it so the claimant (possibly anonymised through us) can post regular updates – we should probably stipulate at least one post per month – and people can ask questions.

Was it a complete waste of money? Is it still ongoing? has it been done?

Definitely not. I’m just about to re-open the bounty and offer someone else the chance to claim it (I’ll do another post to explain the reasons).

If the worst comes to the worst, we will just have to retire the USB bounty as unclaimed. In which case, all the donations will be redistributed across the other open bounties (apart from the ROOL Admin, which is a different thing).

The answers are largely irrelevant now, I’ve given up on bounties.

Please don’t give up on it – these are partly teething troubles because we’ve only been running the scheme for a bit more than a year. We’re still learning lessons.

@Steve Revill:
bq. Please don’t give up on it – these are partly teething troubles because we’ve only been running the scheme for a bit more than a year. We’re still learning lessons.

Me to, in claiming anyway. I will in the future only claim something if I have already almost completed it, and have an initial codebase to submit along with the request to claim the bounty. As well I will only claim a bounty that provides a simple way to upload code on a regular bases, keeping the Hardcopies seperate (as I learned my lesson on that one), and NOT USING EMAIL. Not using Email as it apears as if my code sent that way has not been yet opened by ROOL.

I think that these things should be taken into consideration by ROOL as well in the maintanence of bounties. I think that a view of what I have learned (as stated above) from the opisite side could help you to better deal with bounties in a way that works for everyone in the future.