To have things come somewhat full circle, there’s also a MicroPython port for the BBC micro:bit, based on the C version.
I remember a game called Le Manoir de Mortevielle, a detective/mystery game where all the dialogs were synthesized (not a single line of text for any dialog!) and it fitted on a floppy disk. We spent many hours with a friend on the Amiga 500 version…
That’s a very interesting way to do it. I wonder how well it sold.
The C version of SAM works splendidly on modern systems. Remarkable what it could accomplish on 6502 home computers. The Apple II version was audibly inferior to the Commodore and Atari versions, though (because of its primitive audio hardware).
Cool. I’m not surprised that the Atari version worked better.
It appears that this software was released at about the same time as the Echo II add-on card for the Apple II. Here’s a demo of a real Echo II and an emulator. (Note: though this clip is on my server, I didn’t create it; unfortunately, I’ve lost track of who did.) On the one hand, the Echo II’s voice, or more accurately the voice of the Textalker software, sounds less harsh than SAM. On the other hand, SAM’s intonation and pronunciation seem better to me. I wonder if Textalker was more restricted by the LPC encoding used by the Echo or by the tight memory requirement; if I’m not mistaken, Textalker had to fit in the upper 16K of RAM on an Apple II+ with the so-called language card that expanded it to 64K. It would be interesting if someone reverse-engineered Textalker, as someone reverse-engineered SAM. I don’t know the copyright status of Textalker; it was originally created by the Echo’s manufacturer, Street Electronics, but it seems that it was later maintained by the American Printing House for the Blind.