[Simh] SIMH tape images to real tapes

[Johnny Billquist]

On 2014-05-15 23:10, Clem Cole wrote:
> I wrote one of these a long time ago for Tru64, but I do not think I
> still have it.  The concept is easy:
> http://simh.trailing-edge.com/docs/simh_magtape.pdf

Right. Conceptually it is very simple to write a tape image to a 
physical tape, as long as the tape image contains the necessary information.

> But the problem is it becomes extremely driver specific.   You need to
> know what marks the driver is going to write for you - which is highly
> OS dependent - as well as what the devices does behind the scenes which
> is highly device dependent.  I was going to 9-track and I know the
> details for all of that, but I have not idea how TKxx wrotes tape marks
> (I always avoided them as a "bad idea").

Well, there are only one kind of tape marks, and it's not usually any 
different for different drivers, but exactly how you write a tape mark 
is absolutely OS-dependant.
So any program for Tru64 is essentially useless for someone running VMS 
(for example).

> In 9-track world, the end of each file is mark as a single meta record
> (tape mark),  Two in the row marks end-of-tape.   So the when you write
> a take the driver check to see if its at start of tape and if not, has
> to backs up over the last tape mark and then start writing.  On close it
> writes 2 tape marks.

The two tape marks for logical EOT is a convention, and not all software 
adhere to it. But in general, right. I think someone else mentioned it, 
but it's worth pointing out that for ANSI labelled tapes, the tape marks 
are not used this way, for example.

> QIC tapes do not work that way.  Exabyte and DAT are closer to 9-track
> in rules, but I've forgotten most of the details and I do remember there
> was something funky about the original DECtape but those bits in my
> brain have long rotted away.  You really should open the driver for the
> TK50 and see if you can grok it.  Look at the open and close code very
> carefully.

I don't remember exactly how QIC tapes work, but it rings a bell that 
they might have been different.
DECtapes are not tapes in this sense, and so they should be totally left 
out of this discussion. DECtapes work the same way a floppy do, but 
slower. Fixed length blocks, block addressable, and you can rewrite 
individual blocks. There are no tape marks in the way a 9-track have them.

A TK50 works the same way as a 9-track. So do Exabyte, DAT, DDS, DLT, 
and any "modern" tape technology of today that I can recall.

	Johnny

>
> Clem
>
>
>
>
> On Thu, May 15, 2014 at 4:38 PM, Cory Smelosky <b4 at gewt.net
> <mailto:b4 at gewt.net>> wrote:
>
>     Hello all,
>
>     Googling around only seems to want to show me how to copy real tapes
>     to images.  I need to copy a SIMH tape image to a real tape!
>
>     I seem to recall SIMH including a utility for this...but I could be
>     mistaken.
>
>     I will need a utility that will run on VMS (VAX) as I need to use a
>     TK50 to make a TK50. (Unless someone wants to doante a TK70. ;) )
>
>     --
>     Cory Smelosky
>     http://gewt.net Personal stuff
>     http://gimme-sympathy.org Projects
>     _________________________________________________
>     Simh mailing list
>     Simh at trailing-edge.com <mailto:Simh at trailing-edge.com>
>     http://mailman.trailing-edge.__com/mailman/listinfo/simh
>     <http://mailman.trailing-edge.com/mailman/listinfo/simh>
>
>
>
>
> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
>


-- 
Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email: bqt at softjar.se             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol