The Icon Bar: News and features: What is the point of RISCOS Ltd?
Posted by Jeffrey Lee on 19:00, 7/11/2009
| A9, Flamewars, Open source, Opinion, RISC OS, RISCOS Ltd, RISC OS Open Ltd, Shows, Writing
 After hearing the news that videos from the recent RISC OS London Show are now available online, I decided to take a look at what RISCOS Ltd had to say in their presentation. The results I found to be quite shocking... Disclaimer: Although I've obviously been doing a lot of work for RISC OS Open recently, I am not a member of ROOL, nor am I speaking on behalf of ROOL (or The Icon Bar). The opinions expressed below are mine and mine alone, and any likeness or resemblence to any other person's opinions is entirely coincidental. Firstly, despite the concerns of a couple of the more "distant" members of the TIB staff, it looks like ROL are still alive and are still running the Select scheme. Rather unsurprisingly, they say (at 16:25) that their cashflow isn't stable enough for them to commit to any timetable/roadmap for what's going into future releases. They also cite (at around 23:40) commercial confidentiality reasons for not being able to announce the Vpod-related graphics improvements ahead of time. All fair enough so far. But then, at 38:50, when he talks about the lack of RISC OS 6 on the A9home, ROL big-wig Paul Middleton gives a rather shocking answer: "The A9home, one of the problems there is simply, it's hit the point where things have got to be done, we just don't have the people who have got the skills to actually solve the problem." And at 40:10: "All we can practically do is the areas that we've got people that have got interest in. And I think you can probably guess, the areas that people have got an interest in are [...] graphics" I.e. even though subscribing to the RISC OS Select scheme means that you'll be funding the development of RISC OS, there's no system in place to ensure that the money is being spent on areas that actually need development. If they come across a problem that's too complex for their skill set, or involves working on an area that doesn't hold any interest to the programmer in question, they'll just ignore it. Basically ROL have no direction and no commitments. They say that in the past they've always been sub-contracted to work on things (Omega, A9home, Vpod), but without an external company to give them a goal and a source of funding they're content with just sitting there free-wheeling and soaking up the cash of the Select scheme subscribers. At least Castle were smart enough to realise that they didn't have enough money left to fund the development of the OS, and as a result they open(ish)-sourced it to allow the remaining desktop users to look after it themselves. In fact, what is the point of ROL if they don't have any developers who are working on the code full-time? They can't claim that the money from the Select scheme is putting food on the table of any of their employees (43:25 - "We don't have any programmers whose full time job is doing RISC OS"). And they can't claim that they're using the money to fix any of the real issues with the OS. They can't even claim that they're using it to improve the desktop experience, because they don't produce any roadmaps detailing what they're working on for the next version. Even after a release has been made there's no guarantee that they'll disclose any useful information as to what's changed (see the Select 5i2 changelist, for example). They're just using their subscribers money to provide an extra bit of income to fund the whims of the developers. One thing is now clear to me - with Castle effectively dead and ROL lacking direction, the long-term future of the OS has been left entirely in the hands of its users. We've therefore got to ask ourselves how we want to proceed: Continue as things are nowThe ROOL supporters will continue to support ROOL, and the ROL supporters will continue to support ROL, even though neither can presently guarantee the long-term future of the OS. In the end one side will come out on top, but there's no telling how long that will take or how much the market will shrink in the meantime. Convince ROL to work on the ROOL codebaseThis would be a hard sell for ROL, as it would effectively involve them abandoning the last 10 years of their work. Their pride may also be a problem - could they bring themselves to work on "the enemy"'s code?
And also, if the ROL developers were to suddenly start getting paid to work on the ROOL codebase, how would the people currently doing work for free react? Would ROOL possibly lose momentum rather than gain it? Convince ROOL to work on the ROL codebaseOf course convincing ROOL themselves to work on the ROL code is probably impossible, but if all the ROOL supporters were to throw their time and money at ROL, what would happen? Would ROL gain the funding and skill they need to rescue the OS, or would they still lack the determination and direction that's preventing them from getting RISC OS 6 working on the A9home? Or is there simply not enough money left in the market to guarantee the success of a commercial branch of the OS? (The last time ROL tried asking for money, for the Select on Iyonix scheme, they fell far short of the target they believed they needed to reach for the venture to be commercially viable, and thus the project was scrapped).
And there's also the question of how many ROOL supporters could be convinced to support ROL in the first place, considering ROL's current lack of direction, and the fact that ROOL are so close to having the full OS source released under the shared source license.
To mis-quote Paul Middleton (at 33:05, when he's talking about the open/shared source threat): "If you're doing a job, and you've got the skill to do it, you should get paid for it" - which is exactly why I'm not paying ROL, because they're not doing the job!
Links
| |
What is the point of RISCOS Ltd? |
| |
This is a long thread. Click here to view the threaded list. |
| |
|
Jason Togneri |
Message #112091, posted by filecore at 21:53, 26/11/2009, in reply to message #112090 |
Posts: 3868
|
I have always been tempted to buy Virtual Acorn, hey, the guy who sells it lives ten minutes away from where I grew up, and that really is putting cash in the local economy!
However, even at the current offer price, I cannot justify the expense to play a couple of retro games, and maybe fire it up occasionally a feel sad for what could have been. http://www.redsquirrel.fsnet.co.uk/ - free emulator from the same guy who makes Virtual Acorn, although it's slightly more limited in what it can do. For occasional bouts of nostalgia, it'll be more than fine. VA really isn't worth shelling out hard cash for unless you really like supporting what's left of the 'community', or you're some sort of a developer on the platform. |
| |
 [ Log in to reply ] |
| |
|
Steffen Huber |
Message #112092, posted by hubersn at 11:20, 27/11/2009, in reply to message #112090 |
Member
Posts: 91
|
I have always been tempted to buy Virtual Acorn, hey, the guy who sells it lives ten minutes away from where I grew up, and that really is putting cash in the local economy!
However, even at the current offer price, I cannot justify the expense to play a couple of retro games, and maybe fire it up occasionally a feel sad for what could have been.
Use the "Tom Walker Emulators" - Arculator for the old games, and RPCemu for Risc PC-style work.
What I would love to see would be the RISC OS equivalent of a Linux USB stick. A lightning fast OS which can launch from an external device and give you fast access to a decent and serviceable OS.
It would be great to have a nicely packaged RPCemu with RISC OS 5 and our favourite freeware software.
However, for wider consumption, the big problem that RISC OS faces is the fact that most of the software that is still competitive (Techwriter, ArtWorks, Ovation Pro, Messenger Pro) is not freely available. I don't think you will convince many people to use RISC OS by showing them Paint, Draw, Zap or StrongEd. |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112093, posted by filecore at 13:44, 27/11/2009, in reply to message #112092 |
Posts: 3868
|
I don't think you will convince many people to use RISC OS by showing them Draw Show me a direct equivalent on PC or Mac that does is small, simple and lightweight, and yet allows you do draw such impressive vector designs with ease!!!1111? |
| |
[ Log in to reply ] |
| |
|
Gunnlaugur Jonsson |
Message #112094, posted by Gulli at 14:00, 27/11/2009, in reply to message #112093 |
Member
Posts: 138
|
I don't think you will convince many people to use RISC OS by showing them Draw Show me a direct equivalent on PC or Mac that does is small, simple and lightweight, and yet allows you do draw such impressive vector designs with ease!!!1111?
The time that this had any bigger effect than a "hmm, that's nice" is ancient history and completely irrelevant to just about everyone out there, most likely this would soon be followed with the question "what do you mean it can't show youtube?" |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112095, posted by solsburian at 14:03, 27/11/2009, in reply to message #112092 |
Member
Posts: 23
|
If there was ever a chance of RISC OS being attractive to the wider market, for now we should be less worried about 3rd party apps (though nethertheless having a good choice of apps is always a good thing) and more concerned in addressing the shortcomings in the operating system, like its lack of memory protection etc.
How do we overcome this? To me what made RISC OS special was its UI, perhaps the best approach would be to replace the Kernel and other low level components with another operating system such as Linux or Symbian. That way, RISC OS would have a more solid foundation then what it has now and have the benefit of having more developers (though indirectly) involved with it. It would also mean it could support a more border range of devices.
Other companies have done it, such as Apple replacing OS 9 with the Unix/OPENSTEP derived OS X and Microsoft replacing Windows 9x and under with the OS/2 /VMS/god_knows_what_else derived Windows NT. |
| |
[ Log in to reply ] |
| |
|
Gunnlaugur Jonsson |
Message #112096, posted by Gulli at 14:09, 27/11/2009, in reply to message #112095 |
Member
Posts: 138
|
perhaps the best approach would be to replace the Kernel and other low level components with another operating system such as Linux or Symbian Queue Peter Naulls and others that have strongly objected to this for various (mostly technical) reasons 
Even though I agree it would be the best way I think it's already too late for that. |
| |
[ Log in to reply ] |
| |
|
Andrew Flegg |
Message #112097, posted by Jaffa at 14:17, 27/11/2009, in reply to message #112093 |
Member
Posts: 53
|
Show me a direct equivalent on PC or Mac that does is small, simple and lightweight, and yet allows you do draw such impressive vector designs with ease!!!1111? Inkscape. |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112098, posted by solsburian at 14:25, 27/11/2009, in reply to message #112096 |
Member
Posts: 23
|
If the issues was about legacy applications then perhaps the best approach would be for the OS to have a Emulator built into it like Classic for PPC Macs or VDM's for Windows NT and OS/2.
The problem I could see with that is the performance penalty but to help counter that perhaps the new version should be only targeted for new ARM hardware. |
| |
[ Log in to reply ] |
| |
|
Simon Willcocks |
Message #112099, posted by Stoppers at 18:56, 27/11/2009, in reply to message #112095 |
Member
Posts: 302
|
How do we overcome this? To me what made RISC OS special was its UI, perhaps the best approach would be to replace the Kernel and other low level components with another operating system such as Linux or Symbian. That way, RISC OS would have a more solid foundation then what it has now and have the benefit of having more developers (though indirectly) involved with it. It would also mean it could support a more border range of devices.
Ahem.
http://ro-lookandfeel.blogspot.com/
http://ro-lf.sourceforge.net/ (just svn)
http://stoppers.drobe.co.uk/ (zombie site)
http://www.iconbar.com/forums/viewthread.php?threadid=3945&page=1#42225 (May 2003 Iconbar thread).
And, yes, it's doable but it's taken far too long. Still, it keeps me amused. |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112100, posted by solsburian at 20:53, 27/11/2009, in reply to message #112099 |
Member
Posts: 23
|
And, yes, it's doable but it's taken far too long. Still, it keeps me amused. I've seen your work before (just on the Drobe page though, I thought it was abandoned - I should have looked harder!).
Considering you have done it yourself, I'd say your progress has been very good. They very fact you have got something working does indeed show it can be done. |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112101, posted by filecore at 07:47, 28/11/2009, in reply to message #112100 |
Posts: 3868
|
I thought Peter Howkins was also working on something like this? Or perhaps just assisting. |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112102, posted by filecore at 08:22, 28/11/2009, in reply to message #112097 |
Posts: 3868
|
Show me a direct equivalent on PC or Mac that does is small, simple and lightweight, and yet allows you do draw such impressive vector designs with ease!!!1111? Inkscape.
Looks good. Back when I were a child, anything with a version number of less than 1 was a beta, and anything with a version more than 1 was an upgrade. Now it seems to be the trend (in the *nix world particularly) that version 1 is well-nigh unattainable (look at WINE, which got there after how many years?). Why is that? |
| |
[ Log in to reply ] |
| |
|
Simon Willcocks |
Message #112103, posted by Stoppers at 09:24, 28/11/2009, in reply to message #112102 |
Member
Posts: 302
|
Inkscape.
Looks good. Back when I were a child, anything with a version number of less than 1 was a beta, and anything with a version more than 1 was an upgrade. Now it seems to be the trend (in the *nix world particularly) that version 1 is well-nigh unattainable (look at WINE, which got there after how many years?). Why is that?
I think it's because the technical people are in charge of version numbering, not marketing/sales.
Everybody on the project knows its shortcomings and wont be satisfied until there are no more major features missing or broken. Every new version of Windows would probably have been called a beta by the engineers. If there's no money to be made by rushing it out, there's no pressure (especially if anyone who wants can use the beta). |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112104, posted by filecore at 12:22, 28/11/2009, in reply to message #112103 |
Posts: 3868
|
It just doesn't particularly inspire confidence, when pretty much EVERYTHING is in perpetual beta - if the developers don't feel confident even making it as far as version 1, even after many years' development, why should I as a consumer be confident in a product which is quite clearly labelled as being less than 100% ready? Or are you trying to imply that no software project is *ever* fully ready, and it's all some sort of marketing conspiracy? |
| |
[ Log in to reply ] |
| |
|
Peter Howkins |
Message #112106, posted by flibble at 13:24, 28/11/2009, in reply to message #112101 |
Posts: 892
|
I thought Peter Howkins was also working on something like this? Or perhaps just assisting. A simple iconbar app for 'standard' desktop linux written in GTK for X. And a set of gtk, window manager and iconset themes. I've not worked on it for about a year.
http://marutan.net/pics/rowm-20081031.jpg |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112107, posted by solsburian at 14:19, 28/11/2009, in reply to message #112104 |
Member
Posts: 23
|
It just doesn't particularly inspire confidence, when pretty much EVERYTHING is in perpetual beta - if the developers don't feel confident even making it as far as version 1, even after many years' development, why should I as a consumer be confident in a product which is quite clearly labelled as being less than 100% ready? Or are you trying to imply that no software project is *ever* fully ready, and it's all some sort of marketing conspiracy? Software will always have bugs, glitches, vulnerabilities etc. no matter the version number or how may fixes are made to it. In fact, many fixes will in fact introduce new issues.
If I'm not mistaken, some of the modules in RISC OS have not hit 1.0 yet ether. In reality, version numbers are indeed often a marketing tool, just look at the jump in version numbers for Microsoft Word, done to synchronize the versions for different platforms and to match WordPerfect.
One of the reasons why Inkscape hasn't hit 1.0 yet is because its missing some SVG support but given its stability and features its already got, it if was a commercial product it would have probably have passed the 1.0 version mark a long time ago! |
| |
[ Log in to reply ] |
| |
|
Simon Willcocks |
Message #112108, posted by Stoppers at 14:19, 28/11/2009, in reply to message #112104 |
Member
Posts: 302
|
It just doesn't particularly inspire confidence, when pretty much EVERYTHING is in perpetual beta - if the developers don't feel confident even making it as far as version 1, even after many years' development, why should I as a consumer be confident in a product which is quite clearly labelled as being less than 100% ready? Are you, as a customer, confident that Windows 7 won´t have any issues before service pack 2? None of their other versions have.
There´s an Acer catalogue over there that says both that they recommend Windows 7 and that their laptop includes a downgrade to XP option in the package. That really exudes confidence. 
Or are you trying to imply that no software project is *ever* fully ready, and it's all some sort of marketing conspiracy? It feels a bit that way when you´re working for a company trying to sell software and other people keep promising the well-nigh impossible to customers.
On the other hand, the company´s existence depends on being able to sell things and having non-beta releases are a necessity, because that´s what the customers demand. It´s always going to be a trade-off, there´s just less pressure on free software. |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112109, posted by filecore at 08:59, 29/11/2009, in reply to message #112108 |
Posts: 3868
|
There´s an Acer catalogue over there that says both that they recommend Windows 7 and that their laptop includes a downgrade to XP option in the package. That really exudes confidence. That's a common misconception - 'the customer is always right', etc. Bullshit. Sometimes the customer is so fueled up over anti-Microsoft bandwagonism that even when a subsequent OS is clearly better, on balance, than XP, the customer will still demand downgrade rights for the devil they know. Familiarity is better than progress, Microsoft are evil and must be doing this to punish us, etc. XP is clearly inferior to what's out now, so the fact that Acer is still offering downgrades is simply due to the pressure of the ignorant masses, who simply demand without reason or research most of the time. Pack of deluded sheep, the lot of 'em. |
| |
[ Log in to reply ] |
| |
|
Simon Willcocks |
Message #112110, posted by Stoppers at 10:50, 29/11/2009, in reply to message #112109 |
Member
Posts: 302
|
There´s an Acer catalogue over there that says both that they recommend Windows 7 and that their laptop includes a downgrade to XP option in the package. That really exudes confidence. That's a common misconception - 'the customer is always right', etc. Bullshit. Sometimes the customer is so fueled up over anti-Microsoft bandwagonism that even when a subsequent OS is clearly better, on balance, than XP, the customer will still demand downgrade rights for the devil they know. Familiarity is better than progress, Microsoft are evil and must be doing this to punish us, etc. XP is clearly inferior to what's out now, so the fact that Acer is still offering downgrades is simply due to the pressure of the ignorant masses, who simply demand without reason or research most of the time.
I don't think it's entirely about anti-MS feeling, but more a realisation that companies would have to spend a lot of money to re-train employees for no significant benefit. From their point of view, stability (rather than familiarity) is better than progress.
In what way is Windows 7 better than XP, anyway? (I honestly don't know, any more than I know the advantages of RO6 over RO4; the OS I have runs the programs I have and how can an upgrade improve on that? From experience, I'd only expect things to stop working, rather than work better.) Are there any programs that require a newer version of Windows than XP in order to run?
And what happened to Vista?
To be fair, I doubt a personal purchaser would downgrade their computer (I didn't downgrade from Vista on this one, and I was plesantly surprised to find that I could re-partition the drive from Vista). Not that I actually use it, mind.
Pack of deluded sheep, the lot of 'em. No comment, I still use RO for a lot of things! |
| |
[ Log in to reply ] |
| |
|
Gunnlaugur Jonsson |
Message #112111, posted by Gulli at 11:15, 29/11/2009, in reply to message #112109 |
Member
Posts: 138
|
Sometimes the customer is so fueled up over anti-Microsoft bandwagonism that even when a subsequent OS is clearly better, on balance, than XP, the customer will still demand downgrade rights for the devil they know. Familiarity is better than progress, Microsoft are evil and must be doing this to punish us, etc. XP is clearly inferior to what's out now, so the fact that Acer is still offering downgrades is simply due to the pressure of the ignorant masses, who simply demand without reason or research most of the time. Pack of deluded sheep, the lot of 'em. The poor design of earlier versions of Windows is one of the main reasons for people's need to keep running older versions of Windows because people need, using a term quite common in almost all discussions about RISC OS, to run legacy apps.
Quite a few applications were and still are roaming around freely in places they shouldn't have been touching, such as the registry, the Windows folder and other system folders. Why were the programmers so insanely stupid? Because that's how Microsoft not only led the way by example but even told developers to work like that.
When Microsoft finally started getting their act together and encouraging people to not to run the OS as a user with almost full access to everything, all these applications "broke" because they could no longer change just about anything they saw fit. These applications may be mission critical in businesses or vital to the regular home user but developers weren't exactly given a happy ride from Microsoft with help to fix these issues.
A case in point would be a company no smaller than Logitech, a friend of mine bought a web cam a couple of years back but when she bought a new laptop running Vista the webcam never worked. After a lot of googling and reading at Logitech's site I found that Logitech had given up trying to fix drivers for a range of webcams and other hardware because of the stupid requirements by Microsoft. In order for Logitech to test their drivers on Vista they had to ship the drivers to Microsoft which would then test the driver for compliance for a Vista signed driver. If the driver failed it could not get a signature, leaving Logitech with no possible means to test their drivers on Vista because an unsigned driver couldn't be installed! Fortunately Microsoft changed that but not before Logitech had abandoned all work on Vista drivers for much of their old hardware.
Stupidity like this makes people sceptical towards new versions of Windows, they have no idea whether or not their applications will work on the new OS and most people don't really feel like being guinea pigs.
[Edited by Gulli at 11:16, 29/11/2009] |
| |
[ Log in to reply ] |
| |
|
Gunnlaugur Jonsson |
Message #112112, posted by Gulli at 11:26, 29/11/2009, in reply to message #112098 |
Member
Posts: 138
|
If the issues was about legacy applications then perhaps the best approach would be for the OS to have a Emulator built into it like Classic for PPC Macs or VDM's for Windows NT and OS/2.
That is in fact the route many, including myself, suggested but was dismissed entirely as "too much work".
The problem I could see with that is the performance penalty but to help counter that perhaps the new version should be only targeted for new ARM hardware.
The performance penalty would probably be a lot bigger on current ARM hardware than current x86/x64 hardware. If I recall correctly most fairly new PCs were already capable of running RiscPC emulation at higher speeds than the actual RiscPC itself more than two years ago.
[Edited by Gulli at 11:30, 29/11/2009] |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112113, posted by solsburian at 15:44, 29/11/2009, in reply to message #112112 |
Member
Posts: 23
|
If I recall correctly most fairly new PCs were already capable of running RiscPC emulation at higher speeds than the actual RiscPC itself more than two years ago. Looking at the bechmarks on the Virtual Acorn Website: http://www.virtualacorn.co.uk/faq.htm I can see that the Iyonix holds up well untill you consider the faster Dual and Quad core designs.
Since OMAP runs faster then the Iyonix: http://www.iconbar.com/comments/rss/news1224.html it would be intresting to see how running RISC OS on a Cortex compares to running it on x86 hardware. |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112114, posted by filecore at 18:09, 29/11/2009, in reply to message #112110 |
Posts: 3868
|
In what way is Windows 7 better than XP, anyway? Where to start? Memory management in both Vista and 7, for one - the fact that it has any to speak of. XP just wasted RAM; you could have loads installed and it was almost never touched. Later versions are much more aggressive.
And what happened to Vista? It's actually fairly decent, post-SP2 - and before you complain that it took so long, remember that everyone who whines about how XP is better than Vista has conveniently forgotten that XP was pretty bad until post-SP2 as well. Doesn't say much about Microsoft, true, but in an XP vs Vista comparison, it's usually XPSP3 vs Vista SP0, which is comparing apples with oranges (insert Mac-based pun here). Compare Vista SP2 with XPSP2, and for many things, Vista might actually win.
As for RO4 vs RO6, that's half marketing as much as anything else. RO6 = RO5 but produced by a different company. Throw in RO4 Select/Adjust, too? The RISC OS world isn't any better than the Wintel bunch. |
| |
[ Log in to reply ] |
| |
|
Peter Naulls |
Message #112118, posted by pnaulls at 17:37, 30/11/2009, in reply to message #112096 |
Member
Posts: 317
|
perhaps the best approach would be to replace the Kernel and other low level components with another operating system such as Linux or Symbian Queue Peter Naulls and others that have strongly objected to this for various (mostly technical) reasons
No, not really. You didn't provide any references to such arguments, and I think you've rather misstated what has been argued for. And just because someone repeatedly argues for such things doesn't mean (a) the argument is cohesive (b) the person in question is offering to do any work themselves (c) the solution makes any kind of financial or practical sense (quite apart from technical considerations).
In particular, and apart approaches pointed out by SW, no one has been able to put together a complete solution as to how this would actually work.
Anyway, in the particular case of wanting RISC OS on x86 hardware, this problem has long since been solved, in the form of RiscPC emulation - you get hardware abstraction without any real effort, nearly 100% compatibly, and best of it, it although exists. The only other practical approach for RISC OS on new hardware is direct ports, such as the BB port, or other systems suggested on riscos.info
Elsewhere:
If there was ever a chance of RISC OS being attractive to the wider market, for now we should be less worried about 3rd party apps (though nethertheless having a good choice of apps is always a good thing) and more concerned in addressing the shortcomings in the operating system, like its lack of memory protection etc.
I find a certain about of comedy in such posts, with the use of "we", suggesting poster is going to get technically involved somehow. Apart from that, RISC OS does have memory protection (but could be substantially improved with some minor changes), and lack of apps remains the number one failure point for RISC OS, despite the hype over hardware. |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112119, posted by solsburian at 18:35, 30/11/2009, in reply to message #112118 |
Member
Posts: 23
|
I find a certain about of comedy in such posts, with the use of "we", suggesting poster is going to get technically involved somehow. Apart from that, RISC OS does have memory protection (but could be substantially improved with some minor changes), and lack of apps remains the number one failure point for RISC OS, despite the hype over hardware. I'm pleased you found some comedy in my post :S. As a matter of fact I program for my Job (mainly Java but I do have some experience with other languages) and I hope to get involved in some capacity.
My view on why the core OS should be improved first is that you need a good foundation to support Applicatios. My reasoning is based on this page: http://wiki.netsurf-browser.org/Caveat_RISC_OS |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112120, posted by filecore at 19:02, 30/11/2009, in reply to message #112119 |
Posts: 3868
|
The sad truth is that if RISC OS development had continued at the same pace it was going throughout the late '80s and the '90s (A410x -> A3000 -> A5000 ->  ->  ->  ), it would all look very different now and would in fact be entirely different to what today's developers are forced to work with (and drop down to the level of). Ports aside, the principal fact here is that native OS and hardware architecture is at best more than a decade old. Perhaps holding onto the old stuff is a step backwards - or at the very least, only something that will develop (maintain?) a certain enthusiast cliqueishness. |
| |
[ Log in to reply ] |
| |
|
Daniel Nesbitt |
Message #112122, posted by solsburian at 20:57, 30/11/2009, in reply to message #112120 |
Member
Posts: 23
|
Its probably safe to say that people holding on to old hardware is holding everything back. I can imagine that if Acorn were still around and producing desktop computers, support for RPC's and even Phoebe would have been dropped years ago.
What has happened is that we have a catch 22 situation where the most common hardware that people use is Acorn era stuff so only targeting post Acorn hardware would exclude a large chunk of users and possible revenue for any developer.
[Edited by solsburian at 22:49, 30/11/2009] |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112123, posted by filecore at 21:02, 30/11/2009, in reply to message #112122 |
Posts: 3868
|
"Damned if you do, and damned if you don't." |
| |
[ Log in to reply ] |
| |
|
Jason Togneri |
Message #112125, posted by filecore at 10:00, 2/12/2009, in reply to message #112107 |
Posts: 3868
|
One of the reasons why Inkscape hasn't hit 1.0 yet is because its missing some SVG support but given its stability and features its already got, it if was a commercial product it would have probably have passed the 1.0 version mark a long time ago! Here's my take on versioning systems.
The engineer's method
An engineer's approach (I'm applying these terms arbitrarily) would be such that version 1 = 100% completion. This makes a certain amount of logical sense, and explains why Linux apps in particular rarely seem to get to a version number of 1 or higher:
 0.1 - 10% complete
0.2 - 20% complete
0.3 - 30% complete
0.311 - 31.1% complete
0.35 - 35% complete
0.37 - 37% complete
0.4 - 40% complete
etc
While this makes logical sense, it's not very practical because it seems to presume that either a) a 100% completion rate is unattainable, and that intermediate version numbers will simply go on forever (like taking half a step, then half that step, then half again...) or that b) there is no need for work after 100% (version 1).
The manager's method
A manager's approach would be more along the lines of subdivision - also logical, but to my mind it makes more sense. Each release is assumed to be a subset of some release group:
0.1-0.999 - beta or pre-release
1 - first discrete release ('ready for market')
2 - second discrete release
3 - third discrete release
 3.1 - incremental change to third release (small feature changes, bugfixes, etc); still recognisable as version 3
3.11 - a very small additional change
3.5 - a more sizeable change; still recognisable as version 3
3.7 - a substantial improvement over the original version 3, but still recognisable as such
4 - fourth discrete release; clearly distinct from version 3
etc
The major difference between these methods of visualisation is that with the engineer's method, version 0.6 is clearly a vast improvement (10%) on version 0.5, while with the manager's method, version 6 is distinct from version 5 but not explicitly assumed to be an improvement - just different. This is where some people have problems with the whole different approaches to version numbering.
Thoughts?
EDIT: yes, I did intentionally choose examples that match major RISC OS releases. Yes, I was also taking a dig at the whole RO5 vs RO6 palaver. You see my point.
[Edited by filecore at 16:22, 2/12/2009] |
| |
[ Log in to reply ] |
| |
|
Chris Williams |
Message #112126, posted by diodesign at 13:26, 2/12/2009, in reply to message #112125 |
 The Opposition
Posts: 269
|
4 - 40% complete I assume you mean 0.4?
Thoughts? Now work in the marketing managers, who want new products to start off at version 2 or 3. Version 1 sounds unknown, unreliable and untrustworthy. Version 3, now there's a fine figure you can shell out cash for. This is how they think. |
| |
[ Log in to reply ] |
| |
| Pages (7): |< <
4
> >|
|
The Icon Bar: News and features: What is the point of RISCOS Ltd? |
|
