I've been using Damn Small Linux on low-end x86 machines for a while now and have been very happy with its performance, general utility, and creative approach. While the effort to regroup for a possible Etch+1 release for the m68k architecture is underway, perhaps it would be worth considering a release of something very similar to DSL for Atari, Amiga, and Mac.
Since DSL is Debian-derived, at least some of the effort put into such a port might be preserved and carried forward into a future Debian release. (Pick your source package from the upcoming testing or unstable release, make it run on m68k, then hack it down to size for the DSL-68k version, for instance.) Meanwhile, it would provide a vastly more achievable short-term goal than Debian in its entirety. The number of packages that would need to be maintained would be far more manageable. And the kinds of programs selected for such a release could be chosen on the basis of what is actually useful on m68k hardware. In fact, the package selection for DSL is already based on the exact criteria that would make a good OS release for 68k machines -- packing as much functionality as you can in the lowest amount of disk space, with memory use and good performance on slower processors a high priority. The availability of a lightweight distribution for m68k that could be installed like DSL's frugal install (by copying a ready-made ~50M filesystem image to a disk partition and installing a bootloader -- with extra apps available as MyDSL packages that could be located on a different partition and integrated at boot-time) would really demonstrate the potential for fun and utility that still remains in m68k machines. So we can't boot off of a USB drive. I've got loads of Zip drives lying around. Rolling such a thing out could serve to boost the morale of disappointed Debian-m68k developers and potential users, and could provide a stable platform from which to build toward a more expansive future release. For the time being, forget about a complex and system-taxing installer, monstrous desktop environments, and java VMs. Concentrate on a relatively modern kernel with good support for the hardware; a stable toolchain; and a core selection of sensible, usable apps. Contribute to Emile so all those orphaned Macs out there can stand on their own. I remember running Debian's Mozilla browser on one of my Q840AVs w/128M RAM. It took an eternity to start up, and the UI was painfully slow. But it actually loaded and rendered web pages in a finite amount of time. The browser functionality itself was almost barely usable. If only it could keep up with my keystrokes when I typed in a URL... At that time I was running Galeon on RedHat 8 (or something like that) on a Pentium 200, and noticing that it ran quite perceptibly faster than the Mozilla browser on the same machine because, of course, it used GTK (or whatever) to render the UI instead of the XML rendering engine. Since then I've been really curious to see what Firefox would be like on the Q840. I suspect it would be slow, but far more usable than the full Mozilla suite ever was. DSL allows me to use an improved version of Dillo (older versions of which ran _very_ nicely on my Q840AV) to do a lot of general purpose browsing on older machines. When I come to a page that would really benefit from Firefox, I can copy the URL, start up the big browser, paste it in, and have the benefits of a modern, full-featured, albiet slower browser when I need it. Sylpheed, which is a standard DSL app, ran great on my Quadra. Ted and Beaver look like they should prove useful on m68k as well. I really think that a complete port, or near-port of DSL would be an advantageous way to redeem all the effort that has already gone into getting Debian to run on the m68k machines. That shouldn't go to waste, and the next Debian release date is too far away to be the only hope for anything more recent or usable than the current version. The m68k platform was the first non x86 Debian port. Why not make it the first non x86 DSL port? Keep it small, fast, functional, and free. JCE -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
