|
|||||||||||||||||||||||||||||||||||||||||||||
|
We talked today with Olaf Barthel, the programmer, among others, of the terminal emulator Term and the TCP/IP stack Roadshow. He participated to the development of AmigaOS since the late 1990s. ![]() ![]() My name is Olaf Barthel, I am currently 40 years old and still single. I am self-employed, and my day job is a curious combination of computer and network administration, in-house software design and development, and the occasional web development task. My hobbies still very much revolve around English language literature (I usually read more than 30 books a year), and through no fault of my own I seem to have become a genuine cinephile with a particular interest in movie making and film scores. ![]() That takes me back. It's a nickname which I acquired back at school, more than 20 years ago. ![]() This must have been in 1985/1986 when I read the first set of articles on that emerging Amiga computer, which initially was described as "like the Apple Macintosh, only with colour" (not that I had ever used an Apple Macintosh before, but I knew what it was). The more details emerged, the clearer the picture became: it was a dream machine, the kind of computer which was so far ahead of its time that one could only dream to ever be able to use or buy it. Truth be told, I never felt so incensed about a computer until I caught a first glimpse of the NeXT workstations in 1990. ![]() I started out with a humble Amiga 500, back in 1987/1988, and gradually built up the machine with a second disk drive, a memory expansion, a huge 20 MByte hard disk drive originally intended for an Amiga 1000, and eventually an A590 hard disk drive unit. Following that, earnings from a computer game I co-designed and developed allowed me to buy my own Amiga 3000 in 1991. This I replaced with an Amiga 3000T three years later (which had more memory and an 68040 CPU). Finally, the Amiga 3000T was replaced by an Amiga 3000UX in around 1996 (which had even more memory and an 68060 CPU). The Amiga 3000UX is still my development machine, although I admit that because I have no Amiga in my office, any Amiga software development done today takes place within WinUAE. I still have a truckload of Amiga computers in storage which somehow accumulated over the years, though... ![]() I learned how to program in 'C' first, and then how to make Amiga programs. Throughout 1988-1990 I wrote a commercial computer role playing game ("Legend of Faerghail"), and subsequently did a lot of consulting work in the Amiga market. Some of the work was done for Village Tronic GmbH, for example. I wrote the user manuals and contributed utility software, and sometimes driver software for products such as the "Ariadne II", "Picasso II+", "Picasso IV", "Concierto", "Paloma" and others. Through the early 1990'ies I also kept busy writing software such as my "term" terminal emulation program. ![]() This must have been around 1996/1997. I was doing consulting work for Amiga Technologies GmbH, and part of the job was to sift through the Commodore backup tapes provided to them. These tapes would contain, among other things, the Amiga operating system source code. My involvement with it was something along the lines of getting it into shape in order to resume development work. The updated FFS, SetPatch, etc. that came out of Amiga Technologies GmbH were a product of this work, though Heinz Wrobel was the principal creator of this software. Still, I kept working on the operating system for the next few years to follow... The subsequent AmigaOS 3.5 and 3.9 updates, which I was involved in, came about because somebody convinced Gateway 2000 (as its name was back then) that the Amiga bits they had bought could produce some commercially viable product. Since I was friendly with a number of software developers in Germany, among them Haage & Partner GmbH, I had the good luck to be able to put the plans made for the 3.5 update into practice, and became part of the development team. The AmigaOS 4 development work came about when Fleecy Moss of Amiga, Inc. introduced me to Ben Hermans of Hyperion Entertainment, in around 2001, and suggested that it might be a good idea to work with them. ![]() As far as I know, the period between 1995/1995 was less about developing Amiga products, but much more about acquiring Commodore's assets. When I came to work for Amiga Technologies GmbH, I heard bits and pieces of the story how ESCOM AG managed to snap up the Amiga. It sounded like a really tall tale to me, but here they were and setting up shop in Bensheim, a small town near Frankfurt, and were hiring former Commodore employees. What Amiga Technologies GmbH was supposed to be doing with the Amiga was still not completely certain at the time. Part of the company management wanted to push ahead and get new product developed. This is what eventually produced the Walker prototype gear, and the Amiga surfer software package, to be bundled and sold with the Amiga 1200 computers that were assembled from components ESCOM had acquired along with the Commodore assets. Bigger plans were in the making, too, and might have given us an Amiga built upon the PowerPC platform. But, as you will know, the effort did not come to fruition because ESCOM AG collapsed, and with it the Amiga Technologies GmbH subsidiary. ![]() If there was any involvement on my side, it may have been in the planning of the AmigaOS 3.5 update, for which Allan Odgaard, Fleecy Moss and me collaborated. It was part of an effort to do something useful with the assets Gateway 2000 had acquired. From what I learned at the time, the plan for the AmigaOS 3.5 was ultimately rejected, though. To my great surprise the project was eventually resurrected by Amiga International, Inc. and Haage & Partner GmbH. Having a plan must have helped. ![]() I am the wrong guy to ask. My contributions to the AmigaOS 4 project involved providing a working operating system build, and I did not participate in designing or programming the PowerPC-specific hardware abstraction layer and technology which allowed the operating system to run on it. While the early development work was much about cleaning up the source code, figuring out what to improve, adding enhancements, etc. the heavy lifting of replacing the very foundations of the operating system was being done by Hyperion's own Frieden Brothers. ![]() There was not just the ICOA, there was CORE, COSA, Save the Amiga, KOSH, the Jay Miner Society, OpenAmiga, Phoenix, and Team Amiga, too. While the state and the fate of the Amiga was very much in flux near the end of the 1990'ies there was no shortage of goodwill and societies busily discussing and trying to shape the Amiga's future. It has been a long time since I read any of the messages exchanged within these groups. I recall that the results produced by them were somewhat smaller than the goodwill and ambition present within them. By the year 2001 not a lot of energy was left. The one thing which these organizations eventually led up to become reality was the AmigaOS 4 development project, and the associated AmigaOne hardware development work. Both very ambitious undertakings, but at the time of their inception they were, as we know now, some 5-6 years away from producing results. Too little too late? Maybe... ![]() I do not know if the ICOA, specifically, was pivotal in making the work happen which gave us MorphOS, AmigaOS 4 and the hardware platforms for both. These are, I believe, the lasting results the Amiga community organizations of the late 1990'ies came to affect, shape or help become a reality. Give how much ambition was present in these organizations, these results may not be on the scale one could have hoped for. I would say that they were useful in that they helped to articulate what a possible future for the platform could or should look like. I am doubtful that any single organization would be helpful today. As in the example of the late 1990'ies shows, you did not have one single organization filling the void, you had at least ten. And they were not necessarily pulling into the same general direction. Policy and opinion were just as fragmented as the voice of the Amiga community is today. One big difference being that nowadays there may be a greater division between the parts of the community. ![]() That's difficult. I'm a member of the development group, but I am not involved in management, never was. The best I can do for you is give you a list of the people who have, over the years, worked on and contributed to the AmigaOS 4 project: Tobias Abt, Andrija Antonijevic, Simon Archer, Olaf Barthel, Don Cox, Richard Drummond, Stefan Falke, Stephen Fellner, Philippe Ferrucci, Hans-Jörg Frieden, Thomas Frieden, Peter Gordon, Timothy de Groote, Stéphane Guillard, Paul Heams, Joshua B. Helm, Steffen Häuser, Matthew Kille, Alexander Kneer, Christopher Kossa, Adam Kowalczyk, Jens Langner, Henning Nielsen Lund, Martin McKenzie, Martin Merz, Costel Mincea, Rene Olsen, Sven Ottemann, Andrea Palmate, Almos Rajnai, David Rey, Thomas Richter, Rudolph Riedel, Oliver Roberts, Stefan Robl, Hans de Ruiter, Stephan Rupprecht, Csaba Simon, Vit Sindlar, Steven Solie, Ignatios Souvatzis, Martin Steigerwald, Dirk Stoecker, Jörg Strohmayer, Massimo Tantignone, Rene Thol, Thomas Graff Thøger, Laszlo Torok, Massimiliano Tretene, Andrea Vallinotto, Ross Vumbaca, Charles Warwick, Scott Wegner, Davy Wentzler, Colin Wenzel, Frank Wille, Heinz Wrobel, Tony Wyatt and Detlef Würkner. I hope I did not forget anybody. Not every developer named above is still involved in the work. I myself dropped out for a while due to burnout and other commitments. ![]() As far as I can recall the reaction was sort of a collective groan. The motivation on Amiga, Inc.'s side to go to court over a property which they had divested themselved of in order to focus on software for mobile platforms (phones and whatnot) was hard to fathom if you were "just" a software developer. It certainly did not help motivation in the AmigaOS 4 development group, and it certainly took its toll on Hyperion. ![]() Hm... I can think of about three: 1. AmigaOS has been ported to the PowerPC platform. Considering the scope of the work, and that most of the code that makes up AmigaOS is based upon the original work of the Commodore engineers, this is a major achievement. As I know by now, this is also the kind of achievement which is extremely rare. Outside the domain of military contractors, hardly anybody is willing to make an effort to port an existing, large operating system to a completely different hardware platform. This is because it is risky, expensive and bound to take a long time to come to completion. 2. The result is a commercial product which can be bought today, with hardware it will run on. We actually got this far, in spite of all the risks and the prolonged development work. 3. This is not just a port of code which existed in this form largely unmodified since 1993. The operating system has been significantly enhanced as a result of the development work. The major changes are, however, under the hood and may not be directly noticeable by the end user. ![]() I know what I would like to see happening in the future, but it may not entirely mesh with what the official development path is intended to deliver. In no particular order: 1. Our printing system needs to be replaced. It's not just the lack of drivers for current printers, it is also the architecture which has been on the brink of collapse under its own limitations for ages. 2. The user interface toolkit is not powerful enough for today's needs. It has to be either replaced or augmented. What we have today, just like the printing system, is fundamentally constrained by the same limits imposed upon it in 1991. 3. The file system part of AmigaDOS is badly in need of an overhaul. Not only is it excrutiatingly difficult to implement or port a file system, the current architecture severely limits file system performance. 4. The built-in networking functionality is insufficient for today's needs. The TCP/IP stack is very old, and the tools supplied with it leave much to be desired. Simple things such as file sharing are not even part of AmigaOS. 5. The architecture of the device driver system needs to be updated and made into a framework. We have no uniform device driver model to start with, and it is not even safely possible to query the capabilities of a device driver. ![]() I do not know the plans for updating the Workbench, but I have an opinion on what should be done. The first thing to be done would be to start over from scratch and leave the existing Workbench code behind. This is what should have been done decades (!) ago. We are still using very much the same kind of Workbench that existed in 1985. ![]() Well... I already said a bit what I would like to see done with the operating system in the future. This small list should suffice to keep developers, such as myself, busy for a while. ![]() I "ported" the original UAE to AmigaOS 3.x back in the 1990'ies, out of curiousity. This was not what could be considered a "serious" port with a decent degree of platform integration, e.g. allow you to move data in and out of the emulated system. From my experience with E-UAE, I fear that the emulation may never be as stable and mature as WinUAE is today. It will require a major effort to make it work well within AmigaOS 4, if your goal is to let 68k application software run within the emulator as sort of a sandbox system. Personally, I am not fond of running old games within the emulator, which may work well enough with the current state of E-UAE. The operating system and its application software ought to work well, too. ![]() No. ![]() It is the most powerful piece of hardware any Amiga operating system ever ran on, but certainly not the most expensive. I'd be tempted to shell out for such a machine myself, but I found that the daily work I do at the office, and the Amiga development work I manage to do, require a machine which I can conveniently carry around with me. As powerful as the X1000 is, it is not as portable as a laptop. ![]() In this small market price is always an issue. The kind of power the X1000 provides, for such a small customer base, naturally results in a high price. As things stand today, you cannot make this kind of gear in sufficiently large enough volume to bring down the cost and consequently the price. I consider the x86 path a pipe dream. AmigaOS has no platform/porting layer: it is hardwired to a big-endian host platform, not just the fundamentals but also its data structures. I would say that the chances to see AmigaOS run on an ARM are much higher than to see it run on an x86 family processor. If you wanted to make it work on an x86 host, you would have to throw out all existing Amiga software designed to run on the 68k platform, much of which is still useful today (I would go so far as saying that it is not just useful, it is necessary). You would have to throw away much of the operating system and replace it. Even if you were to make all of that happen as part of an x86 port, you would have to make significant sacrifices. I doubt that any of these would result in a viable product. ![]() When the AmigaOS 3.9 update was released it became clear that the future Amiga operating system would need a TCP/IP stack, and none of the product existing at that time could be expected to ship with it. Out of curiousity I began to investigate how difficult it would be to port a new TCP/IP stack to the Amiga. It turned out not to be that difficult after all, but it took its time to figure out how to make it work well on the Amiga. I wrote new configuration commands, added new Amiga specific APIs and also took the plunge to write PPP and PPPoE network drivers from scratch. ![]() I somewhat doubt that AROS really needs another TCP/IP stack, there is also the issue of not having a big-endian host, which Roadshow assumes to be present. I sort of "cut corners" when I ported the code, and this would come back to haunt me. MorphOS may be a different matter, but contractual obligations prevent me from releasing Roadshow for this platform. ![]() While CygnusEd 4 was still in development, considerable help from the MorphOS developers allowed the CygnusEd port to be made and tested. This is no longer the case with CygnusEd 5. While I wish I had the resources myself to make a good port to MorphOS, I just cannot shoulder this alone. Help would be appreciated. As for AROS, CygnusEd is too much of an AmigaOS application, developed for the kind of machine the Amiga was in its 68000 hardware form, to allow it to be ported. I am afraid that a more reasonable approach would be to start over from scratch instead. That kind of effort is not on my todo-list right now. ![]() I appreciate and admire the work. Had Amiga history taken a different turn, it could have been just one operating system carrying the torch, but now we have three operating systems whose designs draw upon the very same great ideas embodied in AmigaOS. Who is to complain, really? :-) ![]() If I remember correctly, we never talked about the matter. When MorphOS emerged, I was already involved in my own small work, and later the project that came to be AmigaOS 4. The latter complicated matters enormously, I can tell you... ![]() I wish it was. But business decisions and business politics have made things so much more difficult. Also, there is less personal contact between developers than there used to be. The reason why my FFS reimplementation is part of both MorphOS and AmigaOS 4 is because I gave my promise to Ralph that it would be available for MorphOS. I have known Ralph since at least 1993, and met with him and other German developers rather regularly during the 1990'ies. As time went by, these developer meetings became less common, and I really miss them. It is a very diffent thing to have met somebody on a personal level than to just have contact through forums and online chats. There is usually less trust involved, and it becomes so much easier to misunderstand each other's intentions, which can then blow up into incessant flame wars and worse. ![]() I am surprised how long it took for this kind of business opportunity to materialize. The Amiga market always was one in which the platform maker (Commodore) was in a weaker position than the developers creating product for the market. At the best of times players such as Great Valley Products lifted the platform far beyond what the original maker could have achieved. But at the worst of times, you had small businesses taking advantage of the flaws of the platform, and not in a constructive manner. It remains to be seen where "Commodore USA" fits into this scheme. Whatever intentions there are beyond making money, I do not buy into the idea of putting an Amiga label on the kind of hardware that was announced. For me an Amiga came to be enabling technology: put great power into the hands of the users and developers. The effort it took to make a great product, in hardware or software, on the Amiga platform of the 1980'ies and 1990'ies was significantly smaller than on, say the IBM PC compatibles or the Apple Macintosh. Amiga products often put these competing platforms to shame in what you could buy for a lot less money than for comparable solutions, and often they delivered much more than even the more expensive solutions on other platforms did. The Amiga, as a platform, provided enormous leverage, so to speak. I may not know enough about the efforts of "Commodore USA", but right now consider me skeptical about their product being in the same league of enabling technology as the Amiga is supposed to be. ![]() It's not over yet :-)
|