RISC OS : Unique Situation
All information herein is the view of the author. Use at your own risk, no warenty of any kind is provided.

Unique place, among long time OSes:

There are many Operating Systems that people have been using on Personal/Home computers since the mid to late 1980s. Of these, most have been cloned, or forked, and do not have any central authority, though do very well at providing a good system on which to work. To support this view think about the others:

  • MS/PC-DOS : Has RxDOS, DR-DOS, and FreeDOS as clones. No central authority, even when it was only by one company.
  • Amiga OS : Has some partial clones, then AROS, Morph OS, etc as full clones. Never had a central authority as to what to do or name or use.
  • Atari TOS : Again has many many clones (too many to list), and never had a central authority for what names, APIs or other to use/create or how.
  • Sinclair QDOS : Never caught on anywhere. Though still the same story as the others listed, heavily cloned and no central authority.
  • DR-GEM (x86) : Somewhat cloned, in late life. Has never had a central authority for third party applications, libraries, etc.
  • Win16 : Only minimally cloned (oddly enough), never had a central authority for third party creations.
  • OS/2 : Not yet cloned, though some are working on it. Again no central authority for third party software at all.

Then we have RISC OS. RISC OS has always come with the requirement of registering software with a central authority, name, SWIs handled, etc. The authority was originally Acorn Computers, then RISC OS ltd, then RISC OS Open (where it is to this day). The concept being that if everyone registers there software it will help to avoid conflicts. The problem is that it relies on a central authority to register with.

While the concept has some merrit, it is not future proof, or even present proof. There have been a number of programs that do exist that did not get registered, because the programmers did not know to register there work, including a lot of the stuff made here in the USA that I was more familiar with, we definitely did not know. The only other company that I am aware of trying to put forth this requirment was Apple with Extensions (through the init 31 hack) for the Macintosh System Software, Apple quickly gave up on that as programmers did not bother to register there extensions.

I feel that a large part of the reason that many of the other OSes have done explosively well, without interuption, and continue to, is that they did not have to worry about who the central authority is. Just write your software, distribute it however you want, and if it conflicts with something and you find out, add a incompatibility note to the documentation. This is the easy way of doing things, a way that encourages people to write software.

With the large number of new to RISC OS users we have seen come up since the wide availability of the Raspberry Pi, I do wonder how many of them we have lost when they found out they are supposed to register there software with a central authority. This is alien to most programmers that target other Home Computer Operating Systems (Home Computer OS, not the multi-user OSes that some now use).

I write this document, knowing that sooner or later RISC OS will become a purely hobby / community maintained OS, like most Home Computer OSes have become. When this happens and there is no longer a central authority in existance, then what do we do? Why should we have to wait to change our philosiphy on this issue?

Then there is the question of a different distribution of RISC OS in the future. Again does it make sense to stick with the central authority model? Or is it more reasonable to go with a model like most others, where programmers generally do there best to avoid conflicts, and when they are noted users can choose the combination of software that suits them without conflict?

The central authority model is flawed, as it is by its nature not future proof. RISC OS Open is doing well, so is RISC OS Developments. There could be a time in the future when these two are no more, and there is no central authority, then where are we sitting. It seems that it would be reasonable for ROOL and ROD to begin to change the model away from the central registery, to the trusted user model of almost every other Personal/Home Computer OS. Though this is my personal view.