Held in sessions during Saturday & Sunday in parallel with the exhibition
Real Programming on 1950s hardware: the Story of Mel
The 1956 LGP-30 is a computer that demands a rethink - which is why it is interesting. Just 1 bit of data flows through its veins. Instead of RAM or ROM, a rotating drum stores data. Built from 113 tubes, the CPU only has 16 instructions, making machine code programming simple. Except it isn't, because if you want to get performance out of the LGP-30, you have to time the drum. Store a variable at the right distance from the instruction using it, and the serial stream flows uninterrupted. Store it somewhere else, and you have to wait for the drum to rotate up to where your data sits. Hence the nicely unintelligible code from the Story of Mel. We'll go through the programming of the ultimate Real Programmer.
Oscar Vermeulen, ZG
1977, The Year of the Trinity
The first three home computers are introduced in a brief few months of 1977. Machines you can actually use! There's this Apple II from a few kids in a garage, and small calculator company Commodore has the PET. But everyone waits for the TRS-80 from Tandy, a serious company. Except owner Charles Tandy, who only allows 3000 computers to be made. 2,999 unsold units can at least decorate his stores. So the revolution begins.
Rob Carke, SZ
From transistor to microprocessor
No PC or smartphone without a microprocessor. But how did it came about ? What were the steps from tranisitor to integrated circuit to CPU ? Who were the people behind it ? This fascinating talk will give answers !
Robert Weiss, ZH
Graphics on Homecomputers and Game consoles
Numererous were the discussions, back in the highdays, on which computer was the best. Graphics were a main factor. Here we will discuss the graphics infrastructure, on 22 different machines, from the programmers point of view.
Neil Franklin, ZH
Surf the Web on Vintage PCs Using Slim Alternative Web Browsers on Linux
The triumvirate of Firefox, Chromium und WebKit becomes more and more resource hungry. If you try to use them on 20 years old computers (486, Pentium 1, m68k, PowerPC, Sparc, etc.) it gets horribly slow — if working at all. Sure, you can use a browser from the heyday of these computers, but nowadays you won't find many fitting web pages anymore. But there are alternatives: Besides multiple, frugal and nevertheless usable graphical web browsers there are the text-mode web browsers. Frugal and still uptodate. (Talk will be held in German.)
Axel Beckert , ZH (Slides)
Teil 1: Memory - but how ? Main memory, separated from CPU and controlling circuitry, was not always as clear cut an idea as it is now. Here we will discuss memory strategies, starting from Zuse and Eniac, right up until the current paradigms.
Teil 2: Memory - 64KiB are not enough. Very quickly microprocessors very quickly gained the possibility to address 64 KiB of main meory. What seemed giant at first quickly became a bottleneck. We discuss memory strategies on early micro's, starting from the KIM-1 and ending with the IBM PC.
Teil 3: Memory - Everything and now! 16 Bit Adressing space, as is common on the 8 bit processors, is simply not enough. Already in the early 80's datastructures and higher graphics resolutions demanded more memory. We discuss possible strategies for implementing a MMU on the 6502 processor.
Hans Franke, DE
FUZIX - because small is beautiful
Coming out of a fusion of UZI variations - small UNIX-alike systems originally for Z-80 computers, FUZIX aims higher and broader and might show to be one of the most significant projects in the retrocomputing world of this decade. It is a true open-source project and has already been ported to multiple microprocessors and platforms, taking lessons from many UNIX projects and striving for a good balance between features and bloat. The presentation will give an overview and current status of the project, as well as look under the cover to see how it is implemented on our beloved 8-bitters. Got root yet?
Tormod Volden, LU
Implementing a TCP/IP stack on a PDP11 running RSX
A talk about some challenges of implementing a general TCP/IP stack on a machine with a small address space and an operating system with a very different I/O model than Unix. Also presenting some ideas and approaches to how an API for TCP/IP can be done in a different way. I'll also talk a little bit about RSX itself in the context of all of this.
Johnny Billquist, ZH