• Select on ROM. Well, not much to say here; RISCOS Ltd has had an apparently brisk trade in the ROMs it offered for sale, and is fitted in the machines of three drobe writers.
  
  
• Iyonix NetBSD. Of course, we'd known Gavan was working on this since the middle of 2003; unfortunately it's not as complete the the Linux port yet, but hopefully that will be soon.
  
  
• No Iyonix Laptop. The reasons we stated remained true throughout 2004, and Castle's focus remaining on supporting their desktop machine; but see what we say further down for 2005.
  
  
• Shared Libraries. Well, they didn't appear in a form I was hoping for - Steven Simpson did do more work with his AOF system - but more work remains to be done, some of which was covered in our recent article on AOF and ELF.
  
  
• Package Management. Despite numerous instances of cases where much grief could have been avoided by a proper package management system in 2004, a comprehensive system has yet to arrive, although Graham continues to press on. The Unix Porting Project hopes to make uses of this system in 2005, and will hopefully encourage a larger take up.
  
  
• ARM Twister. What can we say? MicroDigital have had a very quiet year, and there's no evidence of any of the promised upgrades since the ethernet cards at Wakefield 2003.
  
  
• Web Browsers. NetSurf continues to be developed apace; the Unix Porting Project demonstrated a port of a simple but functional browser, Dillo, and continues to make important progress towards bringing Firefox to RISC OS.
  
  
• Cino. Oh dear! It's been demonstrated at various RISC OS shows, and information about its development has been quite open, but we've not yet seen a release. Maybe next year, guys?
  
  
• PCI Cards. Well, there wasn't a proliferation of PCI cards supported in RISC OS in 2004, but we did see the release of Geminus. We hadn't known anything about the project, but familiarity with the Iyonix video system in RISC OS and Linux suggested that it was certainly possible, which is why it made the predictions list. And of course, PCITV was released too.
  
  
• GCC 3.3.3. Not only was GCC 3.3.3 released (although upstream versioning meant it was labelled a pre-release), but also was 3.4.1 and, on Christmas day, 3.4.4. Unixlib too, has had an unprecedented number of updates, improvements and bug fixes.
  
  
• GemPrint. This has been very popular, and probably doesn't require any further explanation.

• RISC OS Portable. No, we're not going to predict a portable RISC OS machine in 2005. Why? Because we believe there will be not one, but two. We have no idea what forms these will take in terms of OS versions, processor speed, physical design, etc. But what we do know is that things have changed hugely since the debacle of the RiscStation laptop. There's now a plethora of ARM-based devices around, with a huge range peripherals and made from a vast range of components - there's bound to be something suitable for RISC OS (and plenty more terribly unsuitable). Of course, putting RISC OS on any new machine is a vast amount of effort, but at least it is now rather more practical.
  
  
• Development tools and Unix compatibility. 2005 will be another busy year for Unixlib and GCC. Work has already begun on porting GCC 4.0, which has substantial changes and promises even better code output. There's also the ELF changes we previously discussed, and some major work on Unixlib to close some of the gaps with Unix compatibility. I won't talk too much about those here, as I'll be discussing those soon in a separate article. I think we will also finally see the ability to produce modules using GCC, leaving you with no reason not to use it should you choose.
  
  
• Non-updated RISC OS news sites. No, really? Yes, sadly. The Iconbar, despite more erratic and sometimes odd updates, has managed still to produce quite a volume of articles in 2004, and we don't want to see them disappear. On the other hand, the number of news items on riscos.org, Acorn Cybervillage and MyRISCOS has continued to drop. It's our hope that in 2005, they'll decide it's not in the best interests of RISC OS to display old news.
  
  
• Select 4 and Select 32. Originally intended for release this year, but held up by Castle-RISCOS Ltd controversy, 2005 should see a Select 4, some of the details which were discussed during RISCOS Ltd's Wakefield presentation. We also expect to see further development of Select 32 and machines using it, including the availability of the A9Home, which promises to be a nice proposition for new entry level machines into the RISC OS market (compared purchasing souped-up, but old, RiscPCs).
  
  
• Iyonix 2? We're not sure we want to put our money on this one for 2005. Although it's clear that one based around the IOP332 is certainly possible, or possibly some of the newer and faster ARM9 devices, we think that Castle's focus will not be on a producing an Iyonix replacement until 2006, but you can be certain that they will be keeping their minds open about the possibilities.
  
  
• New Users and more publicity. There's been a trickle of new and returning users in 2004, although perhaps it's not entirely obvious why. With new software and new hardware in 2005, we might reasonably expect this to increase. RISC OS has also done quite well with appearances in wider computing circles, with for example, drobe.co.uk being linked numerous times from OSNews. This may be part of the reason for the new users.
  
  
• New Software. Apart from the previously mentioned items, we might reasonably expect to see USB2 support from Castle (with the USB card in the Iyonix already providing hardware support), a 32-bit Impression Publisher that XAT have been working hard at and possibly Java/Flash plugin solutions via various free software developments.

Related articles
Best of 2005 awards results
Best of 2005 awards voting open
South East 2005 show report

This article has been linked to, or is available in the following formats:  

[Printable]

[Digg this]

[Blog search]