Dimitris Panokostas: Hello there! I’m a software developer (mainly, but not only) and Amiga fan since the 80s. I first came across an Amiga 500 in the late 80s and was blown away by its capabilities. Like others having a similar experience, it was love at first sight. Soon after I got myself an A500 and later on an A4000, which I still have and use today.
Generation Amiga: When did you started working on Amiberry and what motivated you?
Dimitris Panokostas: I liked the idea of the Raspberry Pi from the beginning, although I didn’t own one until the Pi 2 came out. There was this idea of having Amiga emulation on it in the back of my mind, but I never actually did anything about it. I found the idea of a minimal distro booting directly into Amiga emulation very interesting, so when I saw Amibian I was very excited. However, I found certain problems with the approach taken in Amibian, so I offered to help improve it. Unfortunately the creator of Amibian was not very responsive in my attempts to collaborate*, so I ended up starting something on my own, implementing the vision I had on how this “should” be.
The first iteration of Amiberry (before it even got that name) was a simple script to transform a stock Minibian installation to a minimal distro booting into UAE directly. I spent a lot of time in optimizing the boot speed, so that it resembled a real Amiga (you didn’t see any Linux information popping up when you powered up the Pi, no boot logo, just directly jumping into UAE). The back-end was designed to be fully functional, therefore eliminating some of the issues I had with Amibian (for example, you could not upgrade Amibian’s Linux packages – it would break). To cover popular demand, although it wasn’t originally planned, I also released pre-configured SDcard images with everything in them, ready to go.
To improve Amiga emulation on the Pi even further, it was clear that the UAE port needed some love and attention. I also had some ideas on how to improve the back-end as well. To minimize the amount of work needed, I tried to look for a minimal distro that might already offer what I wanted, and that’s when I came across DietPi. DietPi offered everything I wanted and more as a back-end, being smaller and faster than Minibian at the same time. I thought if I could use that, then I can focus on the UAE port only and make everyone’s life better. 😉
I contacted Daniel Knight, the author of DietPi, who happened to be an old Amiga fan as well. He was very interested in getting an Amiga emulation system in DietPi, so we started working on that, bringing all my previous work from Amiberry v1 into his distro. The result was Amiberry v2, a much improved package with many more features that the original, which also included a version of UAE with my improvements.
Since then, I have continued working on improving UAE4Arm for the Raspberry Pi, fixing several bugs, optimizing it for each Pi version, and so on. Here’s a list of the things implemented after forking the project so far:
- New target platform: Pi 3
Optimizations for Pi 3 added
Pi 3 is now the default target if no Platform is specified
Added support for custom functions assignable to keyboard LEDs (e.g. HD activity)
Code formatting and cleanup
FullHD (1080p) resolution supported in Picasso96 mode.
Pi Zero / Pi 1 version now has full Picasso96 support.
Removed Pandora specific keyboard shortcuts which caused crashes
Loading the Configuration file now respects the input settings
Fixed bugs and crashes in GUI keyboard navigation
At the moment I’m working on porting the emulator to SDL2, so we can take advantage of the enhanced support it provides for the GPU, game controllers etc.
Generation Amiga: What is your favorite programming language and why?
Dimitris Panokostas: Nowadays it’s more high-level languages, like C#. I guess as we get older, we tire of dealing with the low-level stuff and just want to get things done. The enormous resources we have at our disposal from today’s computing devices make this easy, but you also lose out on a very important aspect: optimization. It’s good to remember how to do that when needed. Sitting on an older computer (like the Amiga) and trying to create something will teach you that lesson, or remind you of it. 😉
Generation Amiga: Can you tell us more about future Amiga projects?
Dimitris Panokostas: I had a few projects started while experimenting with Hollywood, but run into some bugs in the language and I was waiting for them to get fixed before I resumed them:
– One was a multi-platform Lightwave render client, like ScreamerNet, which also included a real-time chat system.
– Another was an automatic downloader for WHDLoad slaves, which will check your collection, get any new versions of slaves as they become available and install them for you. Basically a tool to help you manage and keep your WHDLoad collection up-to-date.
– I was thinking about an improved version of iGame, since MrZammler has not updated it in a very long time. I had a few ideas on how it could be improved and discussed them with MrZammler, but he didn’t have the time/motivation to implement anything new. Sadly he hasn’t released the sources either, so it would have to be done from scratch.
Generation Amiga: What are your Amiga configurations?
Dimitris Panokostas: Nowadays I have two A4000 which I use occasionally.
– One is equipped with a Cyberstorm 060 MK2, with 128MB RAM and the SCSI module. It boots of a CF-card which is connected on the SCSI bus (using an adapter), and also has a PicassoIV, X-Surf, Deneb (with a few custom ROM modules on it), SCSI multi-card reader and a hand-optimized version of Workbench 3.9.
– The other has a QuadDoubler 040@50MHz, 16MB of RAM, PowerFlyer IDE with a CF-card adapter, Indivision scandoubler and a Kickflash OS4. It has a minimal OS 3.x installation to “keep it simple”.
Generation Amiga: What is your opinion about upcoming Amiga systems ? (X5000, A1222)
Dimitris Panokostas: I’m not a big fan of expensive hardware which has a limited usage scope, and I’m afraid the PowerPC architecture is just that. I also dislike the bitching and moaning of many Amiga users, whenever something new comes up and it doesn’t fit their expectations. I wish A-EON the best of luck, but in my opinion the whole platform should move towards another direction if it is to survive. Personally I believe that cheap hardware that is accessible to all, with a very efficient OS (such as AmigaOS) is the way to go. Currently that seems to be the Arm architecture, so something like the Pi, but without the Linux part on top. The AROS project has the right idea, but it needs more people behind it to move faster. My ideal solution would be AROS (or even better, OS4) running natively on the Pi, with full support for its hardware, and an internal UAE covering all the “classic” applications. It’s not too far away from that already, but it’s not quite there yet either.
Generation Amiga: What are your prefered modern AmigaOS4 games and programs?
Dimitris Panokostas: I don’t own a physical machine that can run OS4, so I can only use it under emulation. It would be great to see it running on common and affordable hardware, like the Pi, but I sadly doubt we’ll see that day. Which is a pity, since it seems like it has some potential which might be lost due to the restricted hardware platform. If a small platform is to survive and expand, it needs new developers to start writing code for it. Code will eventually become applications, some of which might be attractive enough to get more users in the platform. More users means more potential developers as well as customers, so the whole thing forms into a feedback loop shortly after. If the platform is too expensive, restrictive or not widely available, there is very little chance it would take off.
Generation Amiga: Thanks for the interview Dimitris and good luck with future Amiga projects
Dimitris Panokostas: Thank you for having me and best of luck with your work!
*note: to clarify, this was more of a misunderstanding and I hold no grudges against the author of Amibian, with whom we maintain a good relationship and exchange ideas often.