posted by Fulvio Peruggi on Mon 17th Jul 2006 11:14 UTC

"Page 4"
MorphOS development

In the first phase of its history, MorphOS was a dream that slowly came true by means of a number of very talented young programmers. Then, when the hardware development by bPlan (that is now the hardware branch of Genesi) became closer and closer to finalization, and MorphOS was the unique OS able to use this hardware, the development was well supported and accelerated. In the last two years, MorphOS development slowed down again because Genesi concentrated most of its activities on the design of new hardware, on further development of very basic software like the HAL/OF, and on other operating systems like the various Linux distributions, which of course have a larger base of potential users.

Today, the current development of MorphOS goes on slowly but constantly with a number of simultaneous activities.

(1) The core internals of the boot.img are handled exclusively by the MorphOS Development Team. When improvements in this area have been tested enough by the team members and ready for user trials, a CD ISO image is released which also includes a complete MorphOS installation and establishes a baseline for a release.
Little is known about current improvements, but it is publicly known that the members of the MorphOS Development Team already use a new boot.img file where many components were strongly enhanced. For instance, Altivec (the floating point and integer SIMD - Single Instruction, Multiple Data - instruction set implemented in high-end PowerPC processors) is fully supported in all the components of the system software where it can be used for major speed gains.

(2) Other parts of MorphOS which reside outside the boot.img are updated and released in binary form that registered users can download and manually install. These releases are handled on an "as needed" basis to accommodate new features, or correct problems, or even offer a glimpse of what is to come (alpha and beta software).
For instance, this concerns version 4 of MUI, version 6 of AHI, version 3.3 of Poseidon, improved CGX 3D drivers, debugged versions of some high-level libraries, and so on.

(3) Development for MorphOS in the open source/third party arena is not only active but well appreciated and heavily discussed/debated. One special case is the Ambient desktop, which has become an open source effort where some MorphOS team members are actively involved.
Ambient evolves quickly and very visibly to the users, who can download and install nightly builds of this component of MorphOS.

(4) In an interesting move to focus third party developers' attention, users have gotten together to provide a "bounty" system where users (and coders) can submit ideas for development and contribute money for their realization.
Several noticeable projects have emerged from this "bounty" system, including SFSDoctor and MOSNet (both mentioned in the previous sections), and MorphUp (a sophisticated package manager for automatic installation and upgrade of applications). The bounty system is being used even to speed up the development of parts of the OS that have a particular value for users. This requires the collaboration of members of the MorphOS Development Team, like in the case of the last missing native ARexx library (rexxsyslib.library).

What is the final goal of this somewhat anarchic development process? We know its probable name: MorphOS 1.5, and know that its scope is ambitious. This version of the OS should ultimately remove most needs for external program support, and will qualify MorphOS for its debut outside the Amiga community. Unfortunately nothing is known about its release date, though it does not seem to be very close. In a recent private communication, Frank Mariak, one of the leaders of the MorphOS Development Team, wrote me that MorphOS 1.5 is still a thing to come because "its feature set is not finally defined".

Platform expansion

Cubic IDE One of the most important positive effects of the compatibility of MorphOS with Amiga programs is the fact that MorphOS users can still run almost all the commercial software they purchased for their Amiga, with great advantage in power and speed. The MorphOS/Pegasos computer platform does not start from scratch! Although the official death of Commodore is dated April 29, 1994, though many applications for the Amiga were developed for years and years after that date. And a number of important programs are still actively developed today, like, for instance, the extremely sophisticated Desktop Publishing program PageStream (which currently is available at the same time for Amiga, Linux, MacOS, and Windows platforms, as well as in native PPC code for MorphOS), the advanced editor GoldEd (that now is the core of Cubic IDE, an Integrated Development Environment that covers all the major programming languages and SDKs available for AmigaOS/MorphOS), the state-of-the-art presentation program Hollywood (that inherits the illustrious legacy of Scala, preserving full compatibility with that program, and adding all the features allowed by modern graphics systems), and so on.

Anyway, when a computer platform has a small user base, like MorphOS, the development of new software becomes difficult. The production of commercial software declines because of decreased likelihood of finding a sufficiently large number of purchasers. The production of open source and shareware software also declines because it does not find a sufficiently large base of coders: everyone is already concentrated on a number of projects and has no time for others. In such a case there is a solution that sometimes can drastically reduce the development time of an application: porting software from other platforms.

Blender When AmigaOS was designed, a number of structures and features were inspired by Unix, and of course this reflects in the ABox API of MorphOS. So the porting of small commands, utilities, programs, and games from Unix, and now from Linux, to MorphOS is sometimes not difficult. Two specific system libraries (ixemul.library and ixnet.library) make a number of porting efforts easier that require special Linux-like routines. Even large and complex applications like MPlayer, MEncoder, MLDonkey, E-UAE, MAME, and Blender have been ported to MorphOS.

The main obstacles for code porting are the absence of the fork() function in AmigaOS and the ABox of MorphOS, the fact that AmigaOS/MorphOS are not fully POSIX-compliant, and the extreme difficulty of GUI porting. Linux GUIs are based on windowing systems which are usually parts of larger desktop environments, and are not integrated in the OS. Besides a very early port of X-Windows, no Linux windowing system has ever been ported to AmigaOS/MorphOS. There is no real advantage in doing such a port, because the effort would be very hard, and the smallest windowing system for Linux is at least five times larger than MorphOS as a whole. Easy GUI porting would require the complete loss of the small footprint character of MorphOS, and would transform it into a useless new Linux-like OS.

So, unlike other platforms, AmigaOS/MorphOS never had a port of very large and important applications like Mozilla and OpenOffice. The absence of programs like these, that are fundamental for normal users who want to interface their computers with the whole cyberworld without compatibility problems, is the biggest obstacle for a larger adoption of MorphOS in the desktop computer market.

Once people recognized the uselessness of porting large pieces of Linux distributions to MorphOS, a better idea emerged: the creation of wrappers that relate all the calls to certain basic structures of one system to their equivalents in the other system. The first attempt in this direction involves KHTML, the HTML layout engine created by the KDE project. A giant wrapper that relates KHTML to MUI is in the works (most likely it will be ready within the end of the year). The success of this effort will have a strong influence on the future of the MorphOS/Pegasos platform. MorphOS users will have a state-of-the-art browser (current native browsers are not up-to-date), possible new users will be less reluctant, and the success in this field will encourage similar efforts in other directions (Open Office porting).

Table of contents
  1. "Page 1"
  2. "Page 2"
  3. "Page 3"
  4. "Page 4"
  5. "Page 5"
  6. "Page 6"
e p (8)    56 Comment(s)

Technology White Papers

See More