The Icon Bar: Codename: Merlin

Castle have just emailed with plans for the future of RISC OS. This includes a "6 week period of open consultation with the RISC OS community to assist Castle's management in allocating development resources through its internal Green Paper".

Assuring us that the company is "committed to the development of core RISC OS and desktop extensions", here's a quick rundown of the three main areas of development:

1. Tool Chain
Mainly info on the C/C++ suite (and keeping it up to date), with new features such as "inline assembler", "packed structure support" and "instruction scheduling" mooted.

2. Core
Highest priority, as it brings in the most cash, but slightly devoid of punter-grabbing info. The mention of "further ARM cores" support is interesting though.

3. Dekstop Extensions
The bit most readers will be interested in. As well as bug fixes, the following are suggested:
Filer Enhancements
- Extended Configuration
- Extended Graphics/Display Capabilities
- Extended Network Services
- USB Performance and Connectivity
- Audio Input
- WIMP Enhancements
- ROM Applications Enhancements

More info can be found on the Merlin webpage, and it looks like a SmartGroup has been set up to handle the discussions. And Hey, there's the press release to check out too.

Links:
Merlin webpage
press release

Codename: Merlin

flibble (15:25 18/5/2004)
  andrew (16:00 18/5/2004)
    flibble (16:45 18/5/2004)
      andrew (18:03 18/5/2004)
        ajps (18:25 18/5/2004)
          arenaman (10:26 19/5/2004)
            Col (11:10 19/5/2004)
              Horse (12:24 19/5/2004)
                Gulli (16:04 19/5/2004)
                  arenaman (01:34 20/5/2004)
                    rdenk (07:41 20/5/2004)
                      guestx (08:27 21/5/2004)
                        Ian (08:57 21/5/2004)
                          guestx (13:34 21/5/2004)
                            Ian (15:29 21/5/2004)
                              guestx (15:55 21/5/2004)
                                ad (16:52 21/5/2004)
                                  arenaman (23:30 21/5/2004)
                                    guestx (09:48 24/5/2004)

Peter Howkins

Message #92979, posted by flibble at 15:25, 18/5/2004

Posts: 892

1) Make it open source, and get some real 'open consultation'.

[ Log in to reply ]

Andrew

Message #92980, posted by andrew at 16:00, 18/5/2004, in reply to message #92979

Handbag Boi
Posts: 3439

That depends on Castle getting significant revenue from elsewhere does it not?
Castle need to be make profits if RISC OS has a future.

[ Log in to reply ]

Peter Howkins

Message #92981, posted by flibble at 16:45, 18/5/2004, in reply to message #92980

Posts: 892

Such as hardware perhaps ?

[ Log in to reply ]

Andrew

Message #92982, posted by andrew at 18:03, 18/5/2004, in reply to message #92981

Handbag Boi
Posts: 3439

If it's enough of course.

[ Log in to reply ]

Antony Sidwell

Message #92983, posted by ajps at 18:25, 18/5/2004, in reply to message #92982

Member
Posts: 48

If RISC OS was open source, Castle wouldn't need to make a profit for RISC OS to have a future. For Castle to have a future, they would indeed need to make money from elsewhere. :)

[ Log in to reply ]

Michael Stubbs

Message #92984, posted by arenaman at 10:26, 19/5/2004, in reply to message #92983

Member
Posts: 114

Stop trolling. It is childish to suggest that a company releases it's intellectual property into the public domain. Castle is a business. Their aim is to survive and flourish, which often means making money. They'll hardly make a lot of money if their central product is open source, will they.

[ Log in to reply ]

Colin Cartmell-Browne

Message #92985, posted by Col at 11:10, 19/5/2004, in reply to message #92984

Member
Posts: 89

Besides if Castle made RO open source wouldn't you end up with 100's of different OS's rather than just two?

[ Log in to reply ]

Horse

Message #92986, posted at 12:24, 19/5/2004, in reply to message #92985

Unregistered user

No, if RISC OS was open source, you'd quite possibly have a single community-developed OS with substantial momentum, entering markets where the commercial entities don't want (or can't afford) to go. You'd also see some real enhancements being made, too, rather than seeing the emperor's dinner service, looted from his palace when the horsemen rode into town, being polished for the nth time.

[ Log in to reply ]

Gunnlaugur Jonsson

Message #92987, posted by Gulli at 16:04, 19/5/2004, in reply to message #92986

Member
Posts: 138

dream on!

[ Log in to reply ]

Michael Stubbs

Message #92988, posted by arenaman at 01:34, 20/5/2004, in reply to message #92987

Member
Posts: 114

Horse: I guess it would have stopped the two different versions of RISC OS appearing, too, just like Linux being open source has ensured there aren't several different versions floating about and that the interface remains constant. Oh, no, wait, that didn't work...

Open source isn't as good as it sounds. If it was open source, for example, the RISC OS interface could be simply added to a Linux base (a bit like MacOS X) and then goodbye RISC OS on ARM hardware.

[ Log in to reply ]

rick Denkers

Message #92989, posted by rdenk at 07:41, 20/5/2004, in reply to message #92988

Member
Posts: 4

Did Castle realised that the codename ' merlin' was also used by IBM?
IBM used it for their os/2 warp V4, codename merlin.....

Regards,

Rick

[ Log in to reply ]

GuestX

Message #92990, posted by guestx at 08:27, 21/5/2004, in reply to message #92989

Member
Posts: 102

I guess what Horse is trying to say is that, with an open source RISC OS, at least the users get a decent choice about which variant of RISC OS they get to use, and that (like Linux distributions) the choices are laid bare. With RISC OS, on the other hand, it's all about commercial politics. Want Select on your Iyonix? Too bad that it isn't on the agenda of the company which developed it; too bad that company seemingly have it on their agenda to actively *not* support that platform. At least a community-driven RISC OS would attempt to tackle such issues.

[ Log in to reply ]

Ian Jeffray

Message #92991, posted by Ian at 08:57, 21/5/2004, in reply to message #92990

Member
Posts: 4

guestx> You're talking utter bollocks.

[ Log in to reply ]

GuestX

Message #92992, posted by guestx at 13:34, 21/5/2004, in reply to message #92991

Member
Posts: 102

OK. Which part is bollocks, then? The bit about the differences between Linux distributions being completely in the open and widely known, or the bit about the fish in the small pond splashing around trying to nibble each other's fins?

First off, let's brush aside the doubt about Linux distros. Yes, some distros make different decisions about the various pieces of software that gets shipped with the kernel, but you usually have (1) a ton of choices about which distro floats your boat (look at distrowatch for a start) and (2) a fair degree of control over whether you get to run the latest KDE (for example) or not.

As for the fighting minnows situation, I can well imagine a number of useful and substantial projects that could get done (or have been done) on RISC OS if the community were (or had been) involved. I might as well use the past tense, however, because by the time RISC OS does become open source (ie. if some commerical entity doesn't insist on being buried with it), there's likely be more community interest in nostalgia-driven, leather-clad Peter Bondar dolls than in reworking the RISC OS source code for the post-microcomputer age.

[ Log in to reply ]

Ian Jeffray

Message #92993, posted by Ian at 15:29, 21/5/2004, in reply to message #92992

Member
Posts: 4

You're just a Troll. Why do you even bother using this website?

[ Log in to reply ]

GuestX

Message #92994, posted by guestx at 15:55, 21/5/2004, in reply to message #92993

Member
Posts: 102

If stating the obvious (and correcting comfortable misconceptions) is trolling, then I suppose I'm doing a pretty good job of it. ;-)

[ Log in to reply ]

Andrew Duffell

Message #92995, posted by ad at 16:52, 21/5/2004, in reply to message #92994

Prints of Yarm

Posts: 3262

Okay folks... time to count to ten and chill a bit.

[ Log in to reply ]

Michael Stubbs

Message #92996, posted by arenaman at 23:30, 21/5/2004, in reply to message #92995

Member
Posts: 114

Troll/guestx:

I agree ROL could well die, which is due to not meeting the demands of the customer base (32bit RISC OS).

However, even if the desktop market was to die, there is nothing saying that RISC OS and Castle would go with it. As you ought to know, RISC OS is not just used on the desktop.

But it looks very far from being dead, what with the IyonixPC and RISC OS 5 and all the software that is now 32bit.

The market has looked as far from being dead as it has been since Acorn stopped trading.

Now stop trolling.

[ Log in to reply ]

GuestX

Message #92997, posted by guestx at 09:48, 24/5/2004, in reply to message #92996

Member
Posts: 102

Trolling indeed! Anyway, I think your fears of RISC OS on ARM disappearing after a hypothetical open source release of the software are misguided. It's quite possible that RISC OS on ARM would have less significance over time, especially since one of the first things likely to happen would be an attempt to port as much of it as possible to other platforms and architectures. But the emergence of Virtual Acorn suggests that the market entertains such developments anyway, even when the means of executing them is substantially more convoluted.

Meanwhile, consider the benefits for RISC OS on ARM from an open source release. Creative thinkers (eg. Niall Douglas with his Wimp2 stuff) would actually have a realistic chance of delivering production quality goods with full access to the source code, rather than having to second guess the workings of the software. Moreover, such developments would stand a chance of becoming mainstream. Meanwhile, in the closed source world of RISC OS, are either of the commercial guardians willing to take the risks in such an overhaul of the software?

[ Log in to reply ]