1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
|
Network Working Group B. Thomas
Request for Comments: 535 BBN-TENEX
NIC: 17454 July 1973
Categories: Protocols, FTP
References: RFC 520
Comments on File Access Protocol
A file access protocol (FAP) of the sort proposed by John Day in RFC
520 is a good idea. The following comments suggest improvements
(mostly additions) to the protocol described in RFC 520.
1. (Philosophical comment) The intent of both FTP and FAP is to
make it possible for a user to remotely access files. In effect,
FTP provides means for a user to have (parts of) file activity of
the sort typically initiated at the command language level
"slaved" across the network to the site where the file resides.
In a similar way the intent of FAP is to provide a mechanism
which allows activity of the sort typically initiated by programs
at the operating system or monitor level to be "slaved" across
the network to the site where the file resides. The OPEN, CLOS,
SETP, etc. commands of FAP can be viewed as attempts to define
"generic" file system monitor calls. The suggestions made below
are further attempts to make features typically available to
local users also available to remote users via FAP.
2. The OPEN command should allow for a third OPEN mode called A for
append. In terms of its action with respect to a file and file
pointer, the command
OPEN A FOO
would be equivalent to the sequence:
OPEN W FOO
SETP E
The difference would be with respect to access control. Many
systems allow a user to control separately write and append
access to a file (e.g., on TENEX a user usually sets the
protection on his MESSAGE.TXT file such that anyone can append to
it but only he can write it). For such systems the append OPEN
would succeed in many cases in which the write OPEN would fail.
The principle here is that FAP (to as large as degree as is
practical) should allow remote users to access files in the same
way as local users may.
Thomas [Page 1]
^L
RFC 535 Comments on File Access Protocol July 1973
3. The protocol as proposed allows for the creation of non-
sequential files but provides no convenient way for remotely
accessing them after they are created. For example if sent to a
TENEX server, the sequence:
OPEN W FOO //byte size assumed = 36
SETP B
WRITE 512
SETP 1024
WRITE 512
CLOS
would create a file FOO with two pages (on TENEX a page = 512 36
bit words). The two pages would be page #0 and page #2; because
page #1 does not exist the file is said to have a "hole" in it.
Access to FOO via FAP would be difficult unless the remote user
knew its (page) structure prior to access. To support remote
access to files such as FOO, FAP should have means for a user to
determine a file's structure. Consider a value-returning command
that returns the value the file pointer should be set to in order
to point to the first byte of the next used page (block or
record) beyond the current position of the file pointer. With
such a command, call it FNUB (Find Next Used Block), the
following sequence could be used to retrieve a holey file such as
FOO:
OPEN R FILE
SETP B
a: FNUB //let x=the value returned
if x=null
then CLOS
else ( SETP x
READ 512 //page size=512
goto a )
This presumes that the remote user knows the block (page) size so
that he can properly access the file. One can imagine files
having blocks of variable size; perhaps FNUB should return two
values: the file pointer position of the next block and the size
of that block in bytes.
4. FAP should provide means for a remote user to acquire certain
status and "descriptor" information about a given file. The
following is a (non-exhaustive) list of information which would
be useful to a user remotely accessing TENEX files:
- user's access to file; can he read, write, execute or append
the file?
Thomas [Page 2]
^L
RFC 535 Comments on File Access Protocol July 1973
- size information; byte size used in last write access (OPEN
W) of the file; file size in bytes (of that size).
- file access dates; date of create, last read, last write.
- on TENEX a user can specify different access control for
different pages within the same file; a remote user should be
able to acquire such access control information about files
(and be able to specify such access control when he creates
them).
5. There are many applications in which a remote user would like to
access several files simultaneously in much the same way as a
local user can. FAP as proposed can not support such multiple
file access (of course, the user always has the option of going
through an ICP to establish another connection with the server).
FAP can be extended in a simple way to support multiple file
access by including the notion of a "file handle" which is used
to specify which file a given FAP command refers to. When the
user does:
OPEN R FOO
the server's response would include a handle for FOO which the
user would use in subsequent references to FOO. The handle
returned would be a string of the server's choice; it might be
the file's name (FOO), a small integer, etc. Use of a (server
chosen) file handle rather than the complete file name enables
the server to respond to FAP commands without incurring the
overhead of re-parsing the file name for each command. To
illustrate, consider the following sequence which opens a file
for reading and one for writing, reads 3 bytes from the first
file as data, computes using the data and writes a 2 byte result
to the second file:
OPEN R FOO //server returns FH as handle
OPEN W MOO //server returns MH as handle
READ 3 FH //user reads data
//User does some computation on the 3 bytes
WRIT 2 MH //user writes the result
CLOS MH
CLOS FH
Reasonable defaults could be provided with handles: e.g., a FAP
command without a handle refers to the same file as the previous
command; etc. (The association of a handle with a file is
probably better achieved via a separate FAP command rather than
as a side effect of the OPEN command; e.g.,
Thomas [Page 3]
^L
RFC 535 Comments on File Access Protocol July 1973
HNDL FOO )
6. It is important to take local transformations into account (page
3 of RFC 520). However, it is equally important to allow a
remote user to suppress local transformations, if he wishes, so
that he can access the file as it is stored. This would enable a
program that manipulates a file to work equally well whether the
file is local (and accessed "directly" via system calls) or
remote (and accessed "indirectly" via system calls that are
"trapped" and transformed into FAP commands which are sent to the
remote site).
[ This RFC was put into machine readable form for entry ]
[ into the online RFC archives by Alex McKenzie with ]
[ support from GTE, formerly BBN Corp. 10/99 ]
Thomas [Page 4]
^L
|
