The following is an interview with AROS Developer Adam Chodorowski. I want to thank Adam for taking the time to answer all of the questions.
Hello Adam, 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 Amigas in 1992 through my friends; those that had a computer had an Amiga. My first computer became an A600 (I'm too young to be part of the C64 generation ;-)), which I had successfully badgered my parents into getting for me. There was never any doubt about which computer to get; PCs or Macs weren't even considered. Obviously it was the games that interested me in the beginning, and the Amiga blew everything else out of the water at that time.
After a while I gradually shifted into more serious computing interests like programming (initially because I wanted to make my own games). I've always felt that the Amiga somehow encouraged this; I doubt it would have happened quite the same way if I had a PC or Mac (and obviously it wouldn't had happened at all if I only had some game console). Eventually I upgraded to an A1200 which I used for a very long time (it's IMHO one of the best Amigas ever produced), and later I also had an high-end A4000D.
I don't remember exactly how I stumbled upon AROS, but when I did I quickly became very interested. By this time I was already quite cynical about the future of Amiga (having seen both the Commodore and Escom bankruptcies, and not much of anything coming out of Gateway), and AROS seemed like a great way to ensure the survival of the system. There was also the lure of creating an OS that was just like I wanted it. After having managed to discover that AROS was in fact *not* dead (the website was horrible at the time, and it looked like it hadn't been updated for several years) I joined the mailing list, applied for CVS access and started poking around in the source.
How long have you been involved with AROS?
I joined the project the first time in 1999. Sometime during 2001 I completely dropped out of the project for roughly a year; my interest had shifted towards more academic OSes. During this time I also sold my Amigas and started studying at the university. After having used only Linux and Windows for quite some time, I realized it all sucked horribly and what I really wanted was something like AmigaOS and that AROS could give me this in a reasonable amount of time. And thus I rejoined the project.
What were some of your favorite things to do on a classic Amiga?
In the beginning it was exclusively to play all the great games (after all, one of the main reasons I got myself an Amiga was the games). When games got progressively less interesting I started dialing BBS:es, programming (in AMOS and ARexx at first, then Amiga E and later C) and tinkering with the computer in general. I had my own BBS for a while. During a time I also enjoyed playing with raytracing software like Imagine. However, I've always enjoyed a good game now and then, so I've never really abandoned my gaming roots.
Do you currently own any Amiga's?
No, unfortunately not. I sold them when I was short on cash (one I actually exchanged for a laptop), which is something I now regret having done. However, I regularly come in contact with and use Amigas since I'm a member of Swedish Usergroup of Amiga (SUA, http://swedish.usergroup.amiga.tm/) and attend the weekly meetings. This makes it possible for me to satisfy my craving for Amigas without having to own one. :-)
Do you use Amiga OS4 or MorphOS at all?
Not very much. I neither own an AmigaONE nor a Pegasos, which obviously makes it hard to use those operating systems regularly. ;-) SUA owns a Pegasos though, so I've used MorphOS quite a bit. I've also tested AmigaOS 4 at several occasions, but not really used it for any longer amount of time.
What work have you completed for AROS up to this point?
A complete recount would probably take far to much space; I've been active in many areas of the AROS source tree. ;-) Some of the larger and/or more (in my opinion) important things I've done include making AROS bootable from CDROM, implementing the first simple harddrive installer, backporting Intuition from MorphOS, working on Wanderer (especially the workbench and icon libraries, where I implemented support for launching workbench programs and the deficons-like system that uses datatype descriptors), and reworked the website and updated the documentation. I've also ported Python, Lua, VIM and some GNU tools (like diffutils, grep, patch and m4) to AROS. A more complete list of what I've done can be found at http://chodorowski.com/adam/aros.html, although it's probably slightly out of date right now.
Since AROS cannot compile within it's self, what Linux environment do you prefer when developing for AROS?
I run Gentoo Linux, using the Ion window manager (no desktop environment like Gnome, KDE or XFCE). I'm very much a CLI guy, and do most things from the shell (including file management). My text editor of choice is Vim.
What projects are you currently working on for AROS?
Right now I'm working on implementing FAT filesystem support (Team AROS' bounty #12), a new and improved harddisk installer (Team AROS' bounty #6) and UAE integration (Team AROS' bounty #7). I also have several other projects in the pipeline, but I haven't worked much on them for a long time now (so it's probably best not to mention any details ;-)). There are simply too few hours in a day...
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?
What I think is needed right now is some more focus on fixing and polishing existing components, rather than more applications, new features or additional drivers. Those things are important too, but I think AROS really needs some consolidation of existing features before adding more.
AROS could be quite usefull for many people already, given improvement in this area. However, these kinds of tasks are quite boring and lack glory, so unfortunately they are neglected.
This includes making AFS less lethargic than it is (it's close to unusably slow), converting the remaining applications to use Zune (for a consistent look and feel), including more applications by default on the ISO, getting the installer to work properly, making Wanderer more usable for day-to-day tasks, making existing libraries more feature-complete and a lot of small annoyances all over the system. But the most important thing is probably a decent development environment (it should "just work" out of the box, with good documentation).
Having said that, AROS does really need more networking applications. At least a simple web browser, an IRC client and a SSH client. You can't really live without those things in the modern computing world. A SSH client would allow you to do a lot of things remotely that you might not yet be able to do with AROS (like using an IRC client, reading and writing mail or even using a textual web browser), so it would definately be very useful.
If you could have another skilled developer implement some features or applications for you in AROS, what would you like to see implemented?
I think I would need several skilled developers. ;-) There is a huge amount of improvements and applications I would like to see in AROS, and it would take far to much space to discuss (or even list) them all here.
I'd better focus on some of the more important ones...
A specific goal of mine is to be able to use AROS as my main desktop OS. For this to be feasible I'd need at least: a good web browser (a fairly simple one would suffice, but it needs to support SSL and work with the websites I visit), an email client, an IRC client, an SSH client, NFS (or other network filesystem) client support, a usable (speed- and stability-wise) native filesystem, a media player (a port of mplayer would be fine) and a complete development environment (including CVS and Subversion clients). AROS could also do with some (large and small) improvements in many places, but they are not completely critical to me.
How do you think the development of AROS is going now?
How does it compare to earlier days of development?
It's hard to judge. On one hand, it seems to me that commits to the repository have become less frequent in recent years and that there is less activity on the mailing list. On the other hand, several very large and/or important chunks have been commited in a single dump, like eg. AROSTCP and many drivers. Maybe there's simply been a shift in focus from small-scale tinkering to larger projects, but somehow the vitality of the project seems to have decreased. On the gripping hand, there are a lot more third-party developers now that develop software for AROS but aren't directly involved in the core development. A few years ago there was no such thing: all software written for AROS came from developers in the AROS development team. Not to mention that the user community has practically exploded in size in a quite short time.
How does a person become a developer for AROS?
You apply for access to the source repository (it's well documented on the website) and start coding (not necessarily in this order; you can download the nightly source archive and work with that to start with). You should also join at least the development mailing list; otherwise it gets hard to coordinate with the other developers and avoid stepping on their toes. Joining us in IRC at #aros and/or the AROS-Exec forums might also prove useful.
Could you briefly explain to us how a single developer's source code is implemented into the main source code of the OS?
It's quite well documented on the website, at http://www.aros.org/documentation/developers/svn.php.
How do developers currently communicate with each other? Is there anything you would like to change in the way AROS developers communicate with each other?
The main communication channel for the developers is the mailing list, although in recent years the IRC channel and the forums at AROS-Exec have become increasingly important. The IRC channel has proved very useful for quick discussions that would otherwise have taken several days on the mailing list. However, only by a subset of all developers are present on the IRC channel. When a discussion on IRC comes to the point where input is deemed needed from all developers, it normally migrates to the mailing list. Unfortunately, many discussions never reach the mailing list. A similar situation exists with the forums, although to a lesser degree (or so it seems to me). In a way, the communication has become somewhat fragmented, which is a pity. If you want to be sure that you won't miss anything, you have to follow the mailing list, the IRC channel and the forums.
Another interesting form of communication that seems to have evolved in the AROS project is through commit messages. Every time someone commits something to the repository, a email (which includes the commit message) is sent to the aros-cvs mailing list. Since pretty much everyone follows this list in some way, the commit messages are often worded with this in mind, and discussions can often start from such a commit email.
AROS is based after Amiga OS 3.1, but it is also being improved and coming into it's own. What features set it apart from 3.1 right now? What features do you think will help it become even better in the future?
The biggest feature (which you might easily overlook) is of course that AROS has a driver subsystem, where as AmigaOS 3.1 was pretty much hardcoded to run on the classic Amiga hardware. There a few other user-visible improvements, like a more advanced layers subsystem which allows us to have things like offscreen windows (moving windows outside the screen), child windows (windows inside windows) and irregularly shaped windows. Most other extensions are only visible to developers.
Do you ever communicate with any of the developers of Amiga OS4 or MorphOS?
Very little (both personally and as a team). I don't think we've had any sort of official communication with the AmigaOS 4 developers at all. A few AROS developers have personally discussed with various AmigaOS 4 developers on forums. It's better when it comes to MorphOS, since many MorphOS components are based on AROS code. There have been some discussions on the mailing list and on the IRC channel with MorphOS developers.
What do you think about the new project "adtools" where the goals are to unite future development of the Amiga development tools like gcc, binutils and gdb to a common developer platform?
In my opiniion it's a bit too late, unfortunately. All the different systems have already ported the majority of those applications, and all of them have chosen different solutions. Since many choices are integral to the systems (like the ABI) and can't be changed without massive backwards compatibility problems, it's not really possible to unify the development tools. I have a hard time seeing any benefit for AROS from this project.
To much effort, too little gain.
Where do you see AROS two years from now?
Hmmmm, tough question. I would like to say that I see myself using AROS as my main OS by then, but as I am a quite cynical person (being an Amiga fan(atic) for so long can do that to you) I have a hard time being so optimistic. Maybe it will be possible; I don't know. I really hope so.
Hello Adam, 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 Amigas in 1992 through my friends; those that had a computer had an Amiga. My first computer became an A600 (I'm too young to be part of the C64 generation ;-)), which I had successfully badgered my parents into getting for me. There was never any doubt about which computer to get; PCs or Macs weren't even considered. Obviously it was the games that interested me in the beginning, and the Amiga blew everything else out of the water at that time.
After a while I gradually shifted into more serious computing interests like programming (initially because I wanted to make my own games). I've always felt that the Amiga somehow encouraged this; I doubt it would have happened quite the same way if I had a PC or Mac (and obviously it wouldn't had happened at all if I only had some game console). Eventually I upgraded to an A1200 which I used for a very long time (it's IMHO one of the best Amigas ever produced), and later I also had an high-end A4000D.
I don't remember exactly how I stumbled upon AROS, but when I did I quickly became very interested. By this time I was already quite cynical about the future of Amiga (having seen both the Commodore and Escom bankruptcies, and not much of anything coming out of Gateway), and AROS seemed like a great way to ensure the survival of the system. There was also the lure of creating an OS that was just like I wanted it. After having managed to discover that AROS was in fact *not* dead (the website was horrible at the time, and it looked like it hadn't been updated for several years) I joined the mailing list, applied for CVS access and started poking around in the source.
How long have you been involved with AROS?
I joined the project the first time in 1999. Sometime during 2001 I completely dropped out of the project for roughly a year; my interest had shifted towards more academic OSes. During this time I also sold my Amigas and started studying at the university. After having used only Linux and Windows for quite some time, I realized it all sucked horribly and what I really wanted was something like AmigaOS and that AROS could give me this in a reasonable amount of time. And thus I rejoined the project.
What were some of your favorite things to do on a classic Amiga?
In the beginning it was exclusively to play all the great games (after all, one of the main reasons I got myself an Amiga was the games). When games got progressively less interesting I started dialing BBS:es, programming (in AMOS and ARexx at first, then Amiga E and later C) and tinkering with the computer in general. I had my own BBS for a while. During a time I also enjoyed playing with raytracing software like Imagine. However, I've always enjoyed a good game now and then, so I've never really abandoned my gaming roots.
Do you currently own any Amiga's?
No, unfortunately not. I sold them when I was short on cash (one I actually exchanged for a laptop), which is something I now regret having done. However, I regularly come in contact with and use Amigas since I'm a member of Swedish Usergroup of Amiga (SUA, http://swedish.usergroup.amiga.tm/) and attend the weekly meetings. This makes it possible for me to satisfy my craving for Amigas without having to own one. :-)
Do you use Amiga OS4 or MorphOS at all?
Not very much. I neither own an AmigaONE nor a Pegasos, which obviously makes it hard to use those operating systems regularly. ;-) SUA owns a Pegasos though, so I've used MorphOS quite a bit. I've also tested AmigaOS 4 at several occasions, but not really used it for any longer amount of time.
What work have you completed for AROS up to this point?
A complete recount would probably take far to much space; I've been active in many areas of the AROS source tree. ;-) Some of the larger and/or more (in my opinion) important things I've done include making AROS bootable from CDROM, implementing the first simple harddrive installer, backporting Intuition from MorphOS, working on Wanderer (especially the workbench and icon libraries, where I implemented support for launching workbench programs and the deficons-like system that uses datatype descriptors), and reworked the website and updated the documentation. I've also ported Python, Lua, VIM and some GNU tools (like diffutils, grep, patch and m4) to AROS. A more complete list of what I've done can be found at http://chodorowski.com/adam/aros.html, although it's probably slightly out of date right now.
Since AROS cannot compile within it's self, what Linux environment do you prefer when developing for AROS?
I run Gentoo Linux, using the Ion window manager (no desktop environment like Gnome, KDE or XFCE). I'm very much a CLI guy, and do most things from the shell (including file management). My text editor of choice is Vim.
What projects are you currently working on for AROS?
Right now I'm working on implementing FAT filesystem support (Team AROS' bounty #12), a new and improved harddisk installer (Team AROS' bounty #6) and UAE integration (Team AROS' bounty #7). I also have several other projects in the pipeline, but I haven't worked much on them for a long time now (so it's probably best not to mention any details ;-)). There are simply too few hours in a day...
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?
What I think is needed right now is some more focus on fixing and polishing existing components, rather than more applications, new features or additional drivers. Those things are important too, but I think AROS really needs some consolidation of existing features before adding more.
AROS could be quite usefull for many people already, given improvement in this area. However, these kinds of tasks are quite boring and lack glory, so unfortunately they are neglected.
This includes making AFS less lethargic than it is (it's close to unusably slow), converting the remaining applications to use Zune (for a consistent look and feel), including more applications by default on the ISO, getting the installer to work properly, making Wanderer more usable for day-to-day tasks, making existing libraries more feature-complete and a lot of small annoyances all over the system. But the most important thing is probably a decent development environment (it should "just work" out of the box, with good documentation).
Having said that, AROS does really need more networking applications. At least a simple web browser, an IRC client and a SSH client. You can't really live without those things in the modern computing world. A SSH client would allow you to do a lot of things remotely that you might not yet be able to do with AROS (like using an IRC client, reading and writing mail or even using a textual web browser), so it would definately be very useful.
If you could have another skilled developer implement some features or applications for you in AROS, what would you like to see implemented?
I think I would need several skilled developers. ;-) There is a huge amount of improvements and applications I would like to see in AROS, and it would take far to much space to discuss (or even list) them all here.
I'd better focus on some of the more important ones...
A specific goal of mine is to be able to use AROS as my main desktop OS. For this to be feasible I'd need at least: a good web browser (a fairly simple one would suffice, but it needs to support SSL and work with the websites I visit), an email client, an IRC client, an SSH client, NFS (or other network filesystem) client support, a usable (speed- and stability-wise) native filesystem, a media player (a port of mplayer would be fine) and a complete development environment (including CVS and Subversion clients). AROS could also do with some (large and small) improvements in many places, but they are not completely critical to me.
How do you think the development of AROS is going now?
How does it compare to earlier days of development?
It's hard to judge. On one hand, it seems to me that commits to the repository have become less frequent in recent years and that there is less activity on the mailing list. On the other hand, several very large and/or important chunks have been commited in a single dump, like eg. AROSTCP and many drivers. Maybe there's simply been a shift in focus from small-scale tinkering to larger projects, but somehow the vitality of the project seems to have decreased. On the gripping hand, there are a lot more third-party developers now that develop software for AROS but aren't directly involved in the core development. A few years ago there was no such thing: all software written for AROS came from developers in the AROS development team. Not to mention that the user community has practically exploded in size in a quite short time.
How does a person become a developer for AROS?
You apply for access to the source repository (it's well documented on the website) and start coding (not necessarily in this order; you can download the nightly source archive and work with that to start with). You should also join at least the development mailing list; otherwise it gets hard to coordinate with the other developers and avoid stepping on their toes. Joining us in IRC at #aros and/or the AROS-Exec forums might also prove useful.
Could you briefly explain to us how a single developer's source code is implemented into the main source code of the OS?
It's quite well documented on the website, at http://www.aros.org/documentation/developers/svn.php.
How do developers currently communicate with each other? Is there anything you would like to change in the way AROS developers communicate with each other?
The main communication channel for the developers is the mailing list, although in recent years the IRC channel and the forums at AROS-Exec have become increasingly important. The IRC channel has proved very useful for quick discussions that would otherwise have taken several days on the mailing list. However, only by a subset of all developers are present on the IRC channel. When a discussion on IRC comes to the point where input is deemed needed from all developers, it normally migrates to the mailing list. Unfortunately, many discussions never reach the mailing list. A similar situation exists with the forums, although to a lesser degree (or so it seems to me). In a way, the communication has become somewhat fragmented, which is a pity. If you want to be sure that you won't miss anything, you have to follow the mailing list, the IRC channel and the forums.
Another interesting form of communication that seems to have evolved in the AROS project is through commit messages. Every time someone commits something to the repository, a email (which includes the commit message) is sent to the aros-cvs mailing list. Since pretty much everyone follows this list in some way, the commit messages are often worded with this in mind, and discussions can often start from such a commit email.
AROS is based after Amiga OS 3.1, but it is also being improved and coming into it's own. What features set it apart from 3.1 right now? What features do you think will help it become even better in the future?
The biggest feature (which you might easily overlook) is of course that AROS has a driver subsystem, where as AmigaOS 3.1 was pretty much hardcoded to run on the classic Amiga hardware. There a few other user-visible improvements, like a more advanced layers subsystem which allows us to have things like offscreen windows (moving windows outside the screen), child windows (windows inside windows) and irregularly shaped windows. Most other extensions are only visible to developers.
Do you ever communicate with any of the developers of Amiga OS4 or MorphOS?
Very little (both personally and as a team). I don't think we've had any sort of official communication with the AmigaOS 4 developers at all. A few AROS developers have personally discussed with various AmigaOS 4 developers on forums. It's better when it comes to MorphOS, since many MorphOS components are based on AROS code. There have been some discussions on the mailing list and on the IRC channel with MorphOS developers.
What do you think about the new project "adtools" where the goals are to unite future development of the Amiga development tools like gcc, binutils and gdb to a common developer platform?
In my opiniion it's a bit too late, unfortunately. All the different systems have already ported the majority of those applications, and all of them have chosen different solutions. Since many choices are integral to the systems (like the ABI) and can't be changed without massive backwards compatibility problems, it's not really possible to unify the development tools. I have a hard time seeing any benefit for AROS from this project.
To much effort, too little gain.
Where do you see AROS two years from now?
Hmmmm, tough question. I would like to say that I see myself using AROS as my main OS by then, but as I am a quite cynical person (being an Amiga fan(atic) for so long can do that to you) I have a hard time being so optimistic. Maybe it will be possible; I don't know. I really hope so.
posted by Paul J. Beel at 11:48 PM
|
1 comments
