BSD 3 development
[unix-history] / usr / man / man4 / ht.4
CommitLineData
e6817382
BJ
1.TH HT 4
2.SH NAME
3ht \- RH-11/TE-16 magtape interface
4.SH DESCRIPTION
5The files
6.I "mt0, ..., mt15"
7refer to the DEC RH/TM/TE-16 magtape.
8The files
9.I "mt0, ..., mt7"
10are 800 bpi, and
11.I "mt8, ..., mt15"
12are 1600bpi.
13The files
14.I "mt0, ..., mt3"
15and
16.I "mt8, ..., mt11"
17are rewound when closed; the others are not.
18When a file open for writing is closed, a double end-of-file is written.
19.PP
20A standard tape consists of a
21series of 512 byte records terminated by an
22end-of-file.
23To the extent possible, the system makes
24it possible, if inefficient, to treat
25the tape like any other file.
26Seeks have their usual meaning and it is possible
27to read or write a byte at a time.
28Writing in very small units is inadvisable,
29however, because it tends to create monstrous record
30gaps.
31.PP
32The
33.I mt
34files discussed above are useful
35when it is desired to access the tape in a way
36compatible with ordinary files.
37When foreign tapes are to be dealt with, and especially
38when long records are to be read or written, the
39`raw' interface is appropriate.
40The associated files may be named
41.I "rmt0, ..., rmt15,"
42but the same minor-device considerations
43as for the regular files still apply.
44.PP
45Each
46.I read
47or
48.I write
49call reads or writes the next record on the tape.
50In the write case the record has the same length as the
51buffer given.
52During a read, the record size is passed
53back as the number of bytes read, provided it is no greater
54than the buffer size;
55if the record is long, an error is indicated.
56In raw tape I/O, the buffer must begin on a word boundary
57and the count must be even.
58Seeks are ignored.
59A zero count is returned when a tape mark is read;
60another read will fetch the first record of the
61next tape file.
62.SH FILES
63/dev/mt?,
64/dev/rmt?
65.SH "SEE ALSO"
66tp(1)
67.SH BUGS
68The magtape system is supposed to be able
69to take 64 drives.
70Such addressing has never been tried.
71These bugs will be fixed when
72we get more experience with this device.
73.PP
74The driver is limited to four transports.
75.PP
76If any non-data error is encountered, it refuses to do anything
77more until closed.
78In raw I/O, there should be a way
79to perform forward and backward record and file spacing and
80to write an EOF mark explicitly.