This is the TU58 emulation program (zipped C-Source) written at Rockefeller Univ., Dept. of Neurobiology. We copyright (C) it and permit its use provided it is not sold to others.
It has been tested and extensively used as a mass storage replacement for DEC LSI 11/23's and 11/03's running RT-11 at 9600 baud. On a VAX 11/750 running 4.2BSD, the character handling is not quite fast enough to permit reliable writes at 9600 baud. I believe writes work fine at 2400 and 4800 baud. Given a faster host machine or a strip-down tty driver, 9600 baud and possibly above should be feasible.
As it is, it is much faster in booting up RT-11 than a real TU58 - it can boot RT-11 in less than 30 seconds, a real TU58 takes more than five minutes. This speed difference is obviously because the real TU58 has a average access time of 30 seconds whereas the access time for the emulator is just the protocol overhead plus the UNIX filesystem lacency.
It should be possible, though we have yet to try it, to replace the diagnostic TU58 on the VAX 11/750 with a serial line to this program running on another machine. This arrangement should speed up 11/750 diagnostics. If someone is really interested in this but doesn't have a real TU58 to dump the 11/750 diagnostics to a UNIX file, we might do that.
The TU58 emulator uses a non-blocking read() call which is possible under 4.2BSD and implementable using a call like the old RAND empty(), which checks if a read() on a tty would block (is the read queue empty ?). If you don't have 4.2bsd, most versions of UNIX support some form of this functionality.
The command to use to mount filesystems is:
tu58 [-flags] serial_device filesys1 [filesys2] 2> /dev/null &To create a bootable 'tu58' tape it is easiest to copy an RT-11 bootable filesystem and add the your own files with rtpip.
If you need a sample RT-11 TU58 bootable filesystem and you have an RT-11 license, I could send you one (they are usually 256kbytes of binary).
We please report bug fixes and enhancements to us, we would like to track this program.
Sorry if the documentation is not up to snuff.
Cheers, Dan Ts'o Dept. Neurobiology Rockefeller Univ. 1230 York Ave. NY, NY 10021 212-570-7671 ...cmcl2!rna!dan rna!dan@cmcl2.arpa