In sci.electronics.repair Dave wrote:
The manual in question (HP 5370B time interval counter) is dated 1995,
part number 05370-90031. The equipment is no longer supported.

A few observations:

- If the manual is separate from the instrument, it will get separated
from the equipment and lost. Doesn't matter if it's on paper, a CD,
a website, or whatever.

- In these latter days, memory is cheaper than dirt. Especially if
it's ROM.

- Small flash-memory drives with a USB connector are rather ubitiquous.

So... why not store the manuals INSIDE the instruments, and not have
this problem again?

Have a USB port somewhere on the instrument. Plug in a flash drive,
push a button, and get a .txt or .pdf dumped to the flash drive. A
fancy instrument could use a menu selection to dump a nice PDF from
a big ROM; an inexpensive one could use a little recessed switch on
the back panel to bit-bang a text file out the USB port with an 8051
or something. Perhaps even a serial port doing an ASCII (or Kermit
or similar) transfer for a really low-dollar solution. Instruments
fancy enough to have their own Ethernet / Web server could simply
serve documents through that interface. If they just have Ethernet
and TCP/IP, maybe a "magic packet" to a well-known port (17?) on a
non-routable IP address could trigger a manual dump via FTP.

The storage inside the instrument would need to be in ROM, or else it
will eventually get erased. If the instrument takes firmware updates,
there should be a mechanism for the updates to include addenda pages
in the manual dump, but the updates shouldn't be able to overwrite the
original manual.

This won't do a thing for all those instruments floating around out
there now. (Or maybe this has already been thought of and implemented;
I don't get to buy much brand multi-kilodollar test equipment at work.)
But if the market could agree on some kind of standard, and get the
vendors to accept it, the "missing manual" problem could be reduced
a great deal.

Matt Roberds