Open Source HW/SW bridge for pre-Ethernet LaserWriters?

A cheap PC, a LocalTalk card, and some software like COPS might work. I've never done it. The biggest problem will be using generic serial hardware in the Apple LocalTalk mode. I'm not up on the hardware details anymore, but it might need custom hardware to get the LocalTalk signals into a microcontroller.

A LocalTalk card? Never heard of such, do you mean a PCI (or ISA) card that has a LocalTalk connector? Any idea who makes (or made) such a thing?

There were a couple companies that made them...Apple, Dayna and Farallon being the main ones, but I've never come across one myself, nor have I even seen one on ebay.

IMHO, the localtalk->ethernet bridges like your Shiva or Asante's, are probably the cheapest route...though not very cheap.

That's the problem...it's hard to justify spending up to $40 for a bridge, when doubling that can probably get you a new laserprinter.

Jon: Since your post, I talked to My Friend Google and found some good summaries. ("LocalTalk PCI")

Eeun, you are speaking right to my point. Enough time has passed that none of the vendors of that technology are around anymore, or if they are, they don't care a rat's patoot about such ancient products. My thought/goal is to take a PicAXE (or similar $10 microcontroller) and make it do the needed signal and protocol transformations. I -have- to imagine that the protocol specs are public somewhere, right? Maybe going for the lowest goal first is the thing: DIN-8 serial to USB serial, with whatever it takes (maybe nothing) in terms of protocol translation. The goal would be to end up with NEW (and reproducible) technology in place to substitute for old (EOL) bridging technology. Pre-Ethernet Laserwriters are ONLY useless because they don't have the electrical and protocol bridges to talk to contemporary systems. My G4 (more to the point, the latest Apple printer drivers) work happily with Percy the LaserWriter 4/600, but if that Shiva bridge goes poof, Percy is abandonware.

I'd dive into this myself, but I my uptake of software development environments stopped in the previous Millenium. (PowerBuilder was enough to put ANYBODY off their feed... What a POS.)

Well, the obvious solution for laserwriters that support serial is to just use a standard serial dongle. OS X can be hacked to support printers hanging off a serial port:

Serial Printing in OS X

This Sourceforge project provides a working OS X driver for those *really cheap* generic USB to serial dongles that sell for under $10, so you won't have to pay $30 for a Keyspan to make it work.

It's also worth noting that a number of the more expensive Applewriters had a PC-style Centronics parallel port. I see no reason why those won't work with the easily available parallel port->ethernet "mini print server" devices. Most of those emulate a UNIX-style LPD server, which OS X has no problem working with.

Just as a "reality check", those mini print servers generally retail for about $50. That's probably the lowest target price you're likely to hit with an Appletalk->Ethernet version, should you engineer one. Doing it with an embedded ARM CPU system like a Gumstix board would probably be pretty trivial (If you can find documentation for the Appletalk protocol these boards should be fast enough to just "bit-bang" it in software through an I/O port), but the 1x unit price of ethernet equipped models is usually in the $150-$200 range, which means you'll need a fair bit of help from "economy of scale" to get it affordable. A cheaper solution would be an ethernet-equipped 8 or 16 bit CPU (say an eZ80), which is what those print servers generally use, but you'll have to do more of the Appletalk port in hardware.

--Peace