Navigation
« 

Anonymous




Register
Login
« 
« 

Amiga Future

« 

Community

« 

Knowledge

« 

Last Magazine

The Amiga Future 168 will be released on the 5th May.

The Amiga Future 168 will be released on the 5th May.
The Amiga Future 168 will be released on the 5th May.

The Amiga Future 168 will be released on the 5th May.
More informations

« 

Service

« 

Search




Advanced search

Unanswered topics
Active topics
« 

Social Media

Twitter Amigafuture Facebook Amigafuture RSS-Feed [german] Amigafuture RSS-Feed [english] Instagram YouTube Patreon WhatsApp
« 

Advertisement

Amazon

Patreon

« 

Partnerlinks

Interview with Michal Schulz

Description: from The AROS Show

Categories: [EN] Eng_Interviews

Link to this article: Select all

[url=https://www.amigafuture.de/app.php/kb/viewarticle?a=1298&sid=8099882e76230e3adf3378e4cc195b8c]Artikeldatenbank - Interview with Michal Schulz[/url]

This is an interview with AROS Developer Michal Schulz. I want to thank Michal for doing this interview.

Hello Michal, could you give us a history of how you became interested in Amiga's and then eventually came to be involved with AROS?

I became interested in Amiga's somewhere around 1992. I was already a young computer freak (owning an old and still working Atari 65XE) trying to do something useful with a computer I owned. I had heard and read about Amiga everywhere around me. Then I saw it :)) -- my classmate got an A500 as a gift and needed help from a computer "specialist". Since I was only 13 years old, I was attracted to the games. The A500 outperformed every single computer I knew of and was just amazing.

The following years I worked on my neighbor's A500 occasionally and was learning more about the Amos language. And then, in 1994 I got my own A1200. Finally! It came to me with only a 4MB memory extension. I was amazed by its speed and beauty. I will never forget the impressive look and feel and the great performance of the operating system working on such a slow CPU. After a relatively short gaming stage I came back to my beloved proffesion -- programming.

Since I was always working on some kind of lower level (on Atari, I used to write tiny utilities either in 6502 assembler or in Action! language) I abandoned learning AMOS (I found it too system-unfriendly) and I had chosen m68k assembler.

In 1997 I switched to PC. The situation of Amiga was not clear and it seemed that Amiga would dissappear forever. On the PC I didn't waste time on games. I just continued with programming, this time in x86 assembly.

In the same year I came up with the idea of my own operating system. The Mephisto project was born. It consisted of a bootloader with a bootsector and a small kernel - exec.library, utility.library and small bits of other libraries. I used old AROS source and rewrote these libraries into pure assembler. It worked and multitasked pretty well. Since completing this project would be a huge effort for one man, and since AROS didn't have native in those days, I decided to join the team.

How long have you been involved with AROS?

I joined the AROS team over seven years ago, in the middle of 1998. In the beginning it was a really hard time for me, getting used to the make build system, trying to figure out all the mysteries of AROS code and fighting with my child -- the i386-pc target.

What classic Amiga's did you own in the past?

It was an Amiga 1200 with an additional 4MB of FAST memory. Later, I bought a harddrive, CD-ROM, external floppy, and attached my machine to a monochrome SVGA monitor. At that very moment I started to hate all system-unfriendly software.

What were some of your favorite things to do on a classic Amiga?

At the very beginning I was just playing games. Although they used to look way worse than the games we have now on our PC's, they have always had this specific spirit. It was really a pleasure to play on an Amiga. I really enjoyed programming of course, with AsmOne and Devpac later. Thanks to my friend, Tomasz Cechowski (http://www.cechowski.com/), I have been playing with 3D software a bit.

Do you currently own any Amiga's?

Unfortunately not. I've sold my Amiga in order to be able to buy a CRT for my brand new (K6-2, 266MHz) PC. It's a pity, really, but on the other hand, my Amiga has found a new user, which is way better than staying in dust forever in my cupboard.

Could you briefly list the specifications of your development computer or computers?

It's a regular PC with Athlon XP 2000+ CPU and 512MB of RAM. The mainboard is based on the nForce2 chipset and an ATI graphics card allows me to see what I do. AROS boots on this machine (still without ATI support, but I'm working on it). It boots native (from it's own partition on my main harddrive) and it's hosted.

Soon I will work on another machine, donated by terminillis -- a x86-64 system.

Do you use Amiga OS4 or MorphOS at all?

Not really. Since I was involved in the AROS port for Pegasos machines, I do have our AROS-Team Pegasos at home. Since my target was the native port of AROS, I was never in need of using MorphOS. I have tried it of course for a short time and then removed it from the harddrive. Since I do not own either a classic Amiga nor any new Amiga, I cannot use Amiga OS4 at all.

Were you ever involved in projects for Amiga OS4 or MorphOS?

No. I have committed my Amiga-related work for AROS completely.

What work have you completed for AROS up to this point?

Let's start from the beginning. A few months after joining the team I had made AROS boot natively on a regular PC. It didn't do much at that time, but the functionality has been added later on by our Team. I wrote the first (supervisor mode only) and the second (supervisor+user mode) core of native AROS. I wrote the initial version of trackdisk.device, which was fixed (thank God!) by Johan Grip pretty quickly. Based on the assembler sources provided by Frédèric Requinn, I wrote our first version of ide.device (fixed later by Johan Grip of course).

Apart from that I wrote our legacy vga.hidd video driver, mouse.hidd, keyboard.hidd and irq.hidd. Later I wrote the new set of PCI classes - our pci.hidd (Bounty #10, http://www.thenostromo.com/teamaros/bou ... ls_10.html), the accelerated 2D nvidia video driver (Bounty #9, http://www.thenostromo.com/teamaros/bou ... ils_9.html), new ata.device which support BusMaster DMA transfers (Bounty #11, http://www.thenostromo.com/teamaros/bou ... ls_11.html), the AHI audio driver for ICHx Intel chipsets, with misleading ac97 name (Bounty #13, http://www.thenostromo.com/teamaros/bou ... ls_13.html) and the network driver for nForce based NIC's (Bounty #27, http://www.thenostromo.com/teamaros/bou ... ls_27.html). In the meantime I wrote a simple screengrabber utility, PCITool utility, and added UNIT_MICROHZ support in native version of timer.device.

What projects are you currently working on for AROS?

First of all I am working on the accelerated 2D video driver for ATI Gfx cards (coding stage) and on the x86-64 port of AROS/AROS core (planning stage -- I'm waiting for the hardware). The former project should be completed within two months, where the latter will start officially early in the next year, on my birthday :)

Since AROS cannot compile within it's self, what Linux environment do you prefer when developing for AROS?

I am using SuSE Pro 9.3 right now, with KDE 3.5 desktop environment. Until now I have been using vim (and/or kvim) as my text editor of choice, but right now I'm slowly switching to the Eclipse SDK.

I'm sure people are always telling you what they would like to have available in AROS. What features and applications do you think AROS really needs right now?

AROS needs a bit more drivers to be really usable. Currently, AROS does support 2D accelerated NVIDIA cards, soon it will hopefully support ATI cards. In order to satisfy Laptop owners, AROS should support chipsets like the VIA ones too. Then, we definitively need USB support, at least for the keyboards and mice.

AROS needs a fast and efficient filesystem as well. Our AFFS should be either improved or completely replaced with any other filesystem. I'm really depressed when I see my harddrive working on native at 50MB/s speed, whereas the filesystem manages 3MB/s in the very best case.

A typical user would probably benefit from the network applications. We do have a TCP/IP stack that has been available for some time now, but we still lack of applications using it. Email/news and IM clients are just a must.

How do you think the development of AROS is progressing right now?

Better then ever! AROS has been developed slowly for ages, but not really disturbed by the actual situation of Amiga and Amiga-related companies. Since we are finally attracting new developers, it has sped up recently.

I do understand that many people are very anxious about our system and do not expect AROS reaching beta stage (let's say full OS3.1 or OS3.9 compatibility) in the next (x amount) of years. They should always remember, that most of us develop AROS in our free time, shared between family and other non-IT activities.

What features do you think will help AROS become a better OS?

Documentation. Really. AROS lacks good documentation right now. Since we are an AmigaOS-API-compatible OS, one could assume that the AmigaOS documentation can be used. But this assumption is plain wrong. A badly documented system is not a serious system. We cannot expect all developers interested in AROS to look for AmigaOS documentation all over the Internet.

Then, we need better hardware support. Right now many people want to try AROS out, but they get scared when they find out that it won't work with their USB mice and keyboards.

And last but not least, we have to provide a complete compiler toolchain for native AROS. I do know that some people will not contribute with their programs to AROS because they do not want to install linux first and compile there.

Do you think the AROS project could use better organization?

Well, I find the organization of our project really good. We have few communication channels and we always try to keep a decent level of conversation. If we develop slowly, then it's not due to bad organization, but rather due to very few active developers and their lack of time in general.

What I sometimes miss are the long- and mid-term goals of AROS. People observing AROS, suspect that we are just spinning around without any real target.

Where do you see AROS two years from now?

First of all, I would like to see AROS working on embedded machines, where the POSIX-compliant operating system with full memory protection is not a must. I think AROS could fit there pretty well as a fast and small operating system with very low latencies caused by the system itself. The modular approach of the system may be a huge feature in embedded environments. API compliant with the leading Amiga-like OS, whichever it would be.

Apart from that I would like to see AROS running flawlessly on common x86 hardware - with full support of the most typical hardware and better filesystem performance.

Is there anything at all you would like to add?

At this place I would like to thank my wife for understanding and accepting my AROS-coding, which is very important for me since I do it in my free time only, which I could spend with my family instead.