Yonas Fisseha was an undergrad student at Michigan State University in East Lansing, Michigan when I met him thru email. We had started working together on a project to send text books to Addis Ababa University. Before long we were buried under the enormity of the effort and had to step back from it. We were humbled by our first experience at working on an Internet based project. It had started with good intentions, a positive if naive spirit, and an energy that very much remained, waiting for a new direction in which to be channeled. I didn't know where Yonas was going next, what this young man 5 years my juniour had in mind, but I was ready to follow.
From out of the midst of the online political frays of late '92 - early '93 Yonas had appeared. There was so much energy being expelled into email volleys then, exploding in my mail box like ASCII hand grenades. Turmoil that beget turmoil. Endless, mindless flame wars over nothing, pointless and never heading anywhere. Were those mails but lines of code... Was this all an "online community" could offer? Then he came. No rage, no vile words for anyone, just simple messages proposing to work on something positive and with no prerequisite political affiliation required. This is what I had been waiting for, what I had subscribed to mail lists for, at last I found myself no longer deleting my email upon arrival.
Before long Yonas was into projects more in tune with his skills. Until the next project came up we continued communicating as friends and I was discovering in him uncommonly deep knowledge and experiences on Unix and network systems. The son of Ph.D. parents Yonas may have been gifted from the start. He could absorb new things wholly, from top to bottom.
As I communicated more and more with Yonas he became for me the über hacker. It seemed there was nothing that he couldn't do. He was not a hacker in the negative sense of the word (e.g. "cracker") he never brute forced access into anything. He was the architype that would experiment and experiment with high level and low level interfaces and take advantage of what was already there but not common knowledge (he's also the type of guy who coded assembly because he "liked it").
I could ask him what Perl was in '92 and he would send me scripts he had written on his Linux box. I once asked him what a "worm" was and moments later he sent me code that had taken down government servers. This was in the era before search engines, of Windows 3.1 and when "the web" was a TEX tool written by Knuth. He lived in the ethereal world of the cyber underground, of outlaw FTP sites. He turned me on to Phrack and the day after I had downloaded a few issues the site was gone. The rest of the week I lived in fear that the FBI would come knocking at my lab door to take me away, having traced me back through the ftp session logs.
He had a creative, if not geekish, sense of humour that were always new lessons for me. The few people who know what /dev/zero is few also know what purpose it serves (I still don't). Yonas, knowing what it was or not, had the twisted genius to link it to his .plan :-)
He would compose email and USENET articles by telnetting to server ports and issuing raw protocol statements (old timers might recall Amha Lisan's posting to Cleo "If this get's through the Tesh has a lot of work to do!" -that wasn't Amha! I still sting from the flood of email I received when he posted a USENET article in my name advertising low cost RAM chips after the primary production factory in Japan had blown up). I had been employed previously to code on Unix but this "student" was in a league far beyond me. He was turning my machine against me and I learned a lot just trying to defend myself from his mischievous side.
The turning point that blew the doors open for work on Unix systems in Ethiopic occurred towards the end of 1993. Unix was a natural environment for communicating in non-Roman and large character sets such as Hebrew and Japanese. There had been talk of exploiting these possibilities and creating Amharic email and the like. But such talk never grew beyond the hypothetical exercise in large part from the lack of personal time people had as well as for the total dearth of documentation (or at least easily consumable documentation) on the defacto windowing system for Unix: "X".
While this was a barrier to the kid hacker Yonas, he did not stop there at the sight of it and turn back like all the others. Like the rest of us, Yonas did not know how to use real Fidel under the X11 window system or in software and over Internet. He did not know how programs or computers used fonts, he did not even know how to create fonts. Like the rest of us he didn't know the system.
In less than a month and a half Yonas, during the end of the school semester and in the face of final exams, had devised his own mathematical model for representing letters. He also devised his own character coding system to communicate between the software and the letters. He then created a graphical font editor, based on this model, to start hand painting his characters.
Now that he had his own font system, a way to read and display it plus a hand built editor, he was ready to make an app. The natural thing to do next would be to make a simple file viewer and later an editor. This would be the careful, formal and systematic approach, the safest road thru a software development cycle.
But following the standard path is not what had brought Yonas this far. "The Expected" and "The Predictable" were not his companions. Like quantum mechanics if you stick with your intuition you've already lost. He didn't veer left of the path nor towards the right, nor up nor down. He came screaming down the path from the other end.
His first application would be a talk client. It would require reading and writing Ethiopic text streams to and from ports, safe transmission over the Internet, a typing system, the font system, and display system. It was called "EthioTalk".
If you have never used standard "talk" before the principle is not difficult to understand. Internet "chat" rooms today are similar, the difference with "talk" is that you see text as it is typed. Typing is done in a window that is divided into top and bottom halves. In the bottom half of the window will be the text that you are typing and in the top half you will see the text that your friend is typing from any other location world wide. Yonas' had created such a software as the first Unix software specifically designed for any form of Ethiopic communication.
EthioTalk and Yonas' fonts opened the doors to a world of new possibilities. Every software that was in English seemed possible to have in Ethiopic now. EthioTalk evolved thru a few releases and spawned the file viewer offspring "EView" and at the end of 1994 the same work was incorporated into Multilingual Emacs and is now a part of the standard Emacs. Other components of his original work were redeployed in the Ethiopic document services at the Ethiopia Online domain.