Dear Select Subscriber, Following comments made to Drobe and the subsequent discussions that have followed on that web portal and on the Select discussion group I want to make the following points absolutely clear about the current state of RISC OS 4 / Select development. Firstly I want to apologise for the delays in the Select 4 release. Anyone involved in commercial software development will appreciate the problems, even allowing for the uncertainty caused last year by Castles' actions. We have however moved on since then. I understand the frustration of some users, who think that because nothing has been released in the last year that nothing at all has been done that will ever be of use to them. That is not the case as I will try to explain. As recently mentioned on Drobe we had previously only taken 18 requests for Select32 subscriptions from Iyonix users up until the start of September 2005. Of those we have only had payment from 6 people who paid the extra £15 to have both a Select and Select32 subscription - these were taken at the Guildford Show last year. We have not taken payment from anyone who has solely requested Select 32 for Iyonix. Since the Drobe article this number has effectively risen to 41 expressions of interest. This is still nowhere near the 100 minimum new subscriptions we have publicised before we can afford to make any commitment to Select 32 for the Iyonix. We took a big risk by originally starting the Select scheme without really knowing how large the market was going to be as the total sales of Risc PC and A7000 etc has never really been accurately known. However the 500 subscribers that we set as a minimum to justify the scheme was comfortably reached within the first 6 weeks. Many people seem to be overlooking the point I have repeatedly made about Select development in speeches at exhibitions etc. i.e that the route from previous RISC OS 4.02 / Select to Select 4 has been via Adjust32/Select32 (which is the basis of the OS used on the A9). We have not deliberately detoured to produce the version of RISC OS used on the A9 it is built from the same sources and is not "yet another fork in the versions of RISC OS available". This route should not be taken as deliberately ignoring Iyonix users, it is simply the logical way that the software has developed. The subscriptions of all current Select subscribers will ultimately benefit all users as we are only working on one version of the source. If Iyonix users drop out of subscribing to the scheme then it simply places the option of Select for Iyonix that much further away, and ultimately increases the price we would have to charge new purchasers. The way RISC OS 5 has been developed will require a lot of other changes to the RISC OS 4 kernel in order to integrate with RISC OS 5, but once the Select 4 work has been completed the Iyonix can then became a practical goal. In comparison once the necessary Video Hardware independence had been built into the RISC OS 4 ROM it became far easier to support other processor / video combinations. The A9 doesn't use a HAL as some people have suggested, it has its own Bootstrap loader which initialises the hardware and then passes control over to RISC OS. The RISC OS 4 ROM image used in the A9 has built in support for the appropriate hardware, as has always been the case in all previous versions of RISC OS. The hardware specific features of the A9 have been developed by Advantage6 to sit alongside the core RISC OS 4 and have NOT been paid for by the Select subscribers. I would like to point out that to continue to develop both a 26 bit source code version and a 32 bit source code version of Select was not a practical option. It would have been great if we could have developed a full 32 bit version of RISC OS 4 in a few weeks, but the fact is that the work needed is substantial. I would suggest that if Castle had really felt that they could develop the equivalent of Select for the Iyonix via the Merlin scheme in a couple of months last year, then I am sure they would have done so. The fact is that there are very many man years of development (and even more testing) that has gone into Select and that fact that we have been unable to produce a release version of Select 4 in the past year is unfortunate, but is just a fact of life for software development. Simply throwing money or people at a project is very rarely a cure. Anyone involved in commercial software development will realise the enormity of looking after an entire Operating System. Sometimes changes have to be made, as Microsoft realised last year when they scrapped the entire Longhorn development project and went back to scratch and restarted Microsoft Vista using a different Windows core. As previously announced we are giving a three months grace on Select subscriptions. i.e memberships that aren't renewed will have a three months grace period before they lapse. Memberships that are renewed will have an extra three months added to their expiry date. I hope everyone will realise that if no-one renews their Select subscriptions then development will ultimately cease from lack of cashflow. We don't have the profit from hardware sales to use to subsidise the OS development, though we are likely to be offering new shares in RISCOS Ltd shortly to enable more development work to be undertaken. Our philosophy has always been that new features have been introduced via Select for the benefit of all end-users and developers. I would suggest that application developers could easily assume that new features are available on a computer that may be running Select / Adjust, and gracefully inform the user of the lack of the required functionality rather than not using the features in the first place. That way everyone would be encouraged to get the latest version of the OS instead of still running RISC OS 3.5 on their Risc PC, because they still rely on an 11 year old PD application. If everyone continues to work to the lowest common denominator then the market will totally stagnate. Aside from additional hardware support there has to my knowledge been no new real enhanced OS features added to RISC OS 5 to match those available in Select/Adjust, or to supercede them. There is really no reason why any currrent Risc PC generation computer user should not enjoy the features of Select / Adjust. We have after all priced the Adjust ROMs and Select at a far lower price than the Mac / Windows equivalent updates over the past five years. I would also like to point out to everyone that Castle did not hand over the Merlin project to us. The Merlin wish list included over thirty features that are present in Select and not in RISC OS 5. 1. OS Support for Alpha channel sprites 2. Colour mapping to allow colour manipulation of individual images. 3. Translucent sprite plotting 4. OS Support for CMYK sprites 5. Image File rendering system 6. Cut and Paste / highlight in writable icons, CTRL- X, C, V etc 7. Button drawing filter support - to allow different button styles 8. FileSwitch now able to reference path variables via $. e.g Music:$ for 9. CDFS supporting rockridge extensions 10. CDFS to use mimemap for filetype conversions 11. File sort by number 12. Filer drags indicating filename/number of objects 13. Refresh option on Filer 14. Active highlighting of selected files when dragging. 15. Filer menu extensions 16. Direct share option for all filing systems and individual directories 17. Filer displays shared directory icons 18. Thumbnailing in filer 19. Network time setting 20. Ethernet settings can be changed without reboot 21. Firewall 22. Ability to act as DNS server 23. ZeroConfig Internet Setup 24. Support for Universal Clipboard protocol in ROM apps 25. JPEG export from !Paint 26. Drag n drop sprite re-ordering in !Paint 27. PNG import/export in !Paint 28. JPEG screen snapshot / individual window snapshot 29. Configure options for screensaver and DPMS 30. Context sensitive configuration e.g ability for configure plugins to be called directly from menu entry. 31. DHCP client name sending 32. Pinboard to front 33. Syslog Of the above we have shown items 5, 7, 8, 11, 12, 13, 16, 17, 18, 23, 31 and 33 on the demo Iyonix at Shows this year. There are far more features available in Select that we would love to make available to RISC OS 5 users but I am not going to lie to you and say that we will definately do it by December 2005 or any other timescale because that would not be a promise I could keep at present. You might also like to consider that the availability of Select for Iyonix would be much more likely if we had support for the project from Castle. It would after all be encouraging users to buy their computers! For anyone interested the full features of Select / Adjust and the comparison with RISC OS 5 can be found at:- http://www.riscos.com/adjust/AdjustROMchanges.htm http://select.riscos.com/S3Features.html Further programmers details are available at:- http://select.riscos.com/Developers/Issue1overview.htm http://select.riscos.com/DBChanges/ A comparison of features between Select and RISC OS 5 can be found at http://select.riscos.com/iyonix_select.htm We will also be posting some screen shots of the new features planned for Select 4 on the RISC OS web site before the South East Show. In the meantime I want to thank everyone who has subscribed to Select in the past. We don't want to lose you. I also want to apologise for the delays in the Select 4 release. Anyone involved in commercial software development will appreciate the problems, even allowing for the grave uncertainty caused to both us, and the hardware developers, by Castles' actions last year. We have however moved on since then. I understand the frustration of some users, who think that because nothing has been released in the last year that nothing at all has been done that will ever be of use to them. I hope the fact that the A9 is a real machine and is in the hands of real users will disprove those who think that because nothing has been released to Select subscribers this year that nothing has been done at all. We announced at the Guildford Show last October that there would be the option of a) transferring from Select to Select 32 b) of adding Select 32 to existing subscription c) of joining straight into Select 32 d) of joinging both Select and Select 32 So if you are an Iyonix user and would like Select to be available for it, then please let us know. The sooner we get the required number of supporters the sooner that work can begin. Please note though that at present we are not ready to start taking subscriptions to Select 32, we now have 40+ people who have expressed an interest in Select for Iyonix and once we have 100 expressions of interest, we will then ask for those pledges to be turned into actual subscriptions. I also want to state that setting a date for the release of any particular version of Select is extremely difficult. In the past we have stopped developing new features once a certain number had reached a suitable stage of development. With the current base of RISC OS that is not an option as it is the entire core of the OS that has changed. We have to get to the point where the entire OS is suitable for normal everyday use before we make the next Select 4 release and then move on to support Iyonix. I'm sorry that this has meant no release in the past twelve months and if some people feel they have to give up supporting the scheme then I would like to thank them for their support so far. It just unfortunate that the year timespan between subscriptions has not been enough to make the changeover to a totally 32 bit neutral source code for RISC OS 4. As I have said before we are not a faceless megalithic corporation, we are a small group of RISC OS devotees who quite often wonder why we bother! However we then remember that it is a very small group who regularly comment on the discussion groups and that the vast majority of them do not actively participate in RISC OS development. Thankfully the other 99.9% of RISC OS users follow the usual social courtesies and even regularly write to us to thank us for the work that has been done over the past 6 years. It is very much more gratifying to think that we are working for the 5,000 plus satisfied users of RISC OS 4 as opposed to a very small number who have nothing better to do than criticise everything nearly everyone in the RISC OS market does. If you want to have personal arguments about RISC OS then do it personally on a one to one basis. I strongly suspect that a large part of the decline of support for RISC OS has been down to the vastly negative comments that some people feel that they must make about every aspect of RISC OS development (not just that done by RISCOS Ltd). Finally I thought long and hard about including these last paragraphs, but I feel it is time to make my feelings known. I want to point out there are a small number of users who feel that if they continue to express their personal displeasures with the market that the mob rule culture will prevail. I want to make it quite clear that it is this type of attitude, more than anything, that has led to the decline of the RISC OS market. New users coming to the market want to see helpful discussions, not constant rantings from a small group who want to have their choices put first over the silent and content majority. For example some people have complained that the Select features are not used by any developers because they are not universally available. We have done everything possible to make the Select features available to the widest possible range of users, via the Select scheme itself and via the Adjust ROMs, (which are now the standard distribution). Those same people then simultaneously complain that because we haven't planned to implement in RISC OS 4, some of the very few features that are only available in RISC OS 5, (such as Unicode fonts), that we are missing out on an opportunity to support new features. Could they please make their minds up which way they want us to go? Do they want us to produce new features or not!! Finally I want state that once we have completed the conversion and testing of the new 32 bit neutral RISC OS 4 source, it will mean that all versions of RISC OS 4, whether Select / Select 32 or Adjust will be being built from the same source tree. Consequently any bugs found and fixed in one version will automatically lead to the same bug being fixed in other versions. Thus the subscriptions that have been paid by every subscriber are all contributing to a common development. Just because there has been no specific releases this year doesn't mean that anyone's subscription has been wasted. BTW If anyone has any ideas for anything you would like to work on for Select 4 then please let me know. I guess a popular item might be a RISC OS sidebar in the style of the Google or Longhorn versions, with a notepad, picture browser, diary etc. So please let everyone know your thoughts but please, please be positive! After all who wants to go to work if they think that no-one cares about the work they are doing, nor will ever use it! regards -- Paul Middleton Managing Director RISCOS Ltd