Beta! | About us | Contact | Submit news | RSS | Twitter	Webspace | Tech docs | Downloads | BBC Micro | Gallery | Wallpaper

Inside an unofficial Shared C Library By Chris Williams. Published: 20th Mar 2006, 23:51:42. Open sourcing RISC OS by stealth Of the 344 functions in the Castle Shared C Library, Graham Shaw has so far managed to implement 200 in his alternative module. A hundred of those have been fully tested, and this is after six weeks of development. When applications use his project's module, it prints a little banner to remind Graham that the program is using his unofficial SCL, although this will be disabled in the eventual public release. By Graham's reckoning, his target is to recreate the official SCL's 48 base functions, 185 C library functions, and 111 functions related to 64bit support, the C99 standard and other bits and pieces. The module is an open source affair, but will allow applications to use it without repercussions, such as requiring proprietary program authors to reveal their own source code. Developers can, it's expected, choose to use their own choice of 'stubs' library, as provided by the Castle C/C++ compiler kit, RISCOS Ltd's StubsG or with the GCCSDK. He explained: "The library will be licensed under the GNU GPL, version 2 or later, with the libgcc exception, plus a formal statement explaining how I believe it should be applied to a RISC OS module." It was anticipated that Graham, the man behind RiscPkg, would re-use source code from other projects, such as Unixlib, to craft his own SCL. However he said: "I'm keeping my options open. I have nothing against using other code, it just hasn't been worthwhile so far. "So far all of the code has been written from scratch. This is because the SCL needs to meet quite a few requirements above and beyond ISO C compliance if it is to be fully compatible. For example, streams are partly implemented by macros, so they must be implemented in a very specific way. "Also, although I have patched GCC so that it is usable within the SCL, there are parts of the language which aren't supported and would require substantial effort to fix. You can't, for example, use string literals because they are not relocated correctly. This is a significant barrier to the re-use of existing code." On how long it will take to finish, he added: "To use a Debianism, it will be released when it's ready. "This isn't a module you want to install on your computer in a half-baked state." The development of an alternative Shared C Library by Graham Shaw opens many possible avenues of future development. It reminds us how modular RISC OS is, and how it's the sum of some 150 individual components - most of which can be swapped out and replaced, within reason. While some modules do complex work, such as the Font Manager, others can be produced from scratch or through code re-use. For example, the MessageTrans module, of which a replacement is rumoured to be floating around, could be produced from scratch to replace the one supplied by the operating system. This patchwork approach to introducing, bit by bit, an open source RISC OS by stealth can be criticised for perhaps introducing a degree of instability into the system by relying on software produced by different groups of programmers most likely working in their spare time - the headache for users juggling different variants and version numbers will not be insignificant. On the other hand, it could lend a hand in updating components of the operating system that have been left untouched for too long. The debate isn't about to end any time soon. Links Alternative Shared C Library in development Comments? Discussion Viewing threaded comments | View comments unthreaded, listed by date | Skip to the end One of the things that has long niggled me with the SCL is that floating point errors are not handled conveniently. On other platforms, with pre-emptive multitasking, the error signals can be picked up and C source code does not have to wrap its arithmetic inside error trapping. Can Graham do anything about this? I do not want to pester him with vague queries when he is busy doing valuable work, but maybe he will recognize the problem that I am fumbling to express. GavinWraith on 21/3/06 8:44AM [ Reply | Permalink | Report ] It's a shame it's not released under something less restrictive than GPL. That way the pressure could have been put on RISC OS Ltd and Castle to adpot the "third way" and use this open Shared C Library, and we'd be back to having one SCL for all. Whether or not they would have done so is, of course, a different matter. cmj on 21/3/06 9:38AM [ Reply | Permalink | Report ] In reply to GavinWraith: Sounds feasible, but it might cause compatibility problems with existing software so I would want to tread cautiously. (There is the possibility of making this sort of behaviour optional - either controlled globally or on a per-program basic - in much the same way as some of the features provided by UnixLib.) A related question is whether to use the FPE or soft floating point. (Initially I'm going for the FPE, but a floating point library could be incorporated later.) gdshaw on 21/3/06 9:50AM [ Reply | Permalink | Report ] In reply to cmj: The licence only specifies what you can do without asking the author beforehand. Having said that I doubt my module would integrate very well with the way RISC OS is built, not to mention any political or support issues. gdshaw on 21/3/06 10:06AM [ Reply | Permalink | Report ] cmj: Under the GPL, RISC OS Ltd, Castle (and everyone else on the planet) would be free to use, distribute, bundle with machines, include in ROMs and modify the code. Their only obligation is that they must provide the source to modified versions that they distribute. Given Graham's version would pretty much solve their compatability issues that's not a lot to ask in return. Note: this is because of the extra exemptions of the libgcc clause, that effectively negate the 'viral' aspects of the GPL that buisnesses generally have a problem with. In otherwords, feel free to pressure ROL and Castle to use this version. flibble on 21/3/06 11:26AM [ Reply | Permalink | Report ] In any case, as Graham points out, it only imposes conditions on what they can do without asking him (as would the plain GPL). Because he developed it from scratch without using any code from elsewhere he's got full control over it, so there's nothing to prevent him releasing it separately to any individual or company under a completely different license if he wishes to. SimonC on 21/3/06 11:41AM [ Reply | Permalink | Report ] Will such open source modules, intended as 'drop-in' replacements for parts of RISC OS, be able to avoid further complicating the problems with conflicting version numbers that we've seen when (e.g.) ROL and Castle each make separate releases of the 'same' component with their own numbering schemes? chrisrayson on 21/3/06 4:12PM [ Reply | Permalink | Report ] In reply to chrisrayson: I can avoid making the problem any worse. My intention is to match the version number of the most recent ROL or CTL module which which I believe mine to be fully compatible, and then provide further details of exactly what is loaded in the comment field. I'm not so sure what to do in response to the function _clib_version. Does anyone use this, and if so do they interpret it programmatically? gdshaw on 21/3/06 5:41PM [ Reply | Permalink | Report ] Please login before posting a comment. Use the form on the right to do so or create a free account.

Login Username Password Create a new account Forgot your password? Search this website This week's poll Recent developments have left me feeling... Assured ROS will appear on new hardware 55% Pleased OS desktop features are being developed 10% ROL and ROOL will eventually kiss and make up 5% App developments are critical 10% Dave Holden sleeps easy at night 19% Discuss this | Archives Featured articles Wakefield 2009 wrap-up, photos and video The weekend's RISC OS event has been and gone and we've got the rest of our lives to look forward to. Here's a round-up of extra news and Drobe's show-related coverage and some photos taken from Wakefield 2009 - plus a video from the show floor.  16 comments, latest by AW on 29/4/09 7:41PM. Published: 27 Apr 2009 RISC OS 5 pictured running on ARM Cortex-A8 kit Picture exclusive - This grainy photograph shows a port of RISC OS 5, sourced from the RISC OS Open project, running on a Beagleboard - a device powered by a 600MHz ARM Cortex-A8 processor with a built-in graphics chip. The port, developed by Jeffrey Lee with help from Uwe Kall and ROOL staff, is seen as a major breakthrough for the shared-source project as it proves the OS can be ported to new hardware without the need for a large team of engineers.  75 comments, latest by rjek on 30/4/09 3:15PM. Published: 25 Apr 2009 Open documents from Windows-using pals with handy online tool It can be a pain when someone sends you a file that can only be opened on Windows, Mac OS X or Linux - but with the help of a free-to-use website and NetSurf, Paul Stewart reveals how these documents can be viewed on RISC OS.  6 comments, latest by AW on 8/5/09 12:12AM. Published: 19 Apr 2009 Useful links News and media: Iconbar • MyRISCOS • ArcSite • RISCOScode • ANS • C.S.A.Announce • Archive • Qercus • RiscWorld • GAG-News Top developers: RISCOS Ltd • RISC OS Open • MW Software • R-Comp • Advantage Six • VirtualAcorn Dealers: CJE Micros • APDL • Castle • a4 • X-Ample • Liquid Silicon • Webmonster Usergroups: WROCC • RONE • NKACC • IRUG • SASAUG • ROUGOL • RONWUG • MUG • GAG • RISCOS.be Useful: RISCOS.org • RISCOS.info • Filebase • NetSurf Non-RISC OS: The Register • The Inquirer • Apple Insider • BBC News • Sky News • Google News • xkcd • diodesign

Recently logged in:  jmb •  blahsnr •  jess •  killermike •  librarian16 •  diomus •  dkb •  Revin Kevin •  littlebird •  miskin •  Stats
© 1999-2009 The Drobe Team. Some rights reserved, click here for more information | Powered by MiniDrobeCMS, based on J4U
"On Drobe's view of confidentiality and 'public interest'... I find the words 'invasion of privacy' more appropriate"
Page generated in 0.112 seconds.