README
1===============================================================================
2 FSP client version 0.0-h (pre-alpha)
3-------------------------------------------------------------------------------
4
5 This software is copyright 1992, 1993 Philip G. Richards, All
6 Rights Reserved. Any copies that are made of this code MUST keep an
7 unmodified copy of this notice within a file called README in the
8 same directory as the source code. This is a PRE-ALPHA release --
9 I would prefer that the code does not get posted to Usenet, thank-you
10 very much :-)
11
12 THIS SOFTWARE COMES WITH ABSOLUTELY NO WARRANTY.
13
14 The author will accept no responsibility for any use or misuse of the
15 software, nor for any consequences produced by using it. In other
16 words, if it deletes all your files, don't blame me -- by using
17 it, you are accepting responsibility and liability. (It will not,
18 however, intentionally delete them -- it will have been an error.)
19
20 The author can be contacted (currently) as <pgr@prg.oxford.ac.uk>.
21
22===============================================================================
23
24 **************************************************************************
25 *** Note: I have written the *program* called `fsp' -- I have ***
26 *** *not* written the entire FSP package. Questions about ***
27 *** the server and the like should be sent to Wen-King Su ***
28 *** <wen-king@vlsi.cs.caltech.edu> -- or the new maintainer of the ***
29 *** code Joseph_Traub@seismo.soar.cs.cmu.edu. ***
30 **************************************************************************
31
32
33---------------- What the client can do ---------------------------------------
34There is a help command; not a very good one, but it does tell you the
35commands. Type `help' to get a list of all commands; to get a brief
36description of a command, type `help name' (`name' being the command);
37to get brief help on all commands type `help all' (warning: there are
3846 commands).
39
40See the manual page fsp(1) for a somewhat longer description...
41
42Note for versions since 0.0-g: the command `cat', `get', `du', and
43`tar' can all take a `-r' flag -- this causes the commands to process
44subdirectories as well as files. e.g., `du' will give the disk usage of
45the current directory; `du -r' will give the disk usage of the current
46directory and _all_ subdirectories (resursively). The `timeout' command
47causes communications to return an error code if the timeout occurs;
48this means that:
49
50 timeout 30
51 pro
52 iferror exit
53
54is a good way of determining whether a remote site is alive or not when
55writing scripts. The `ver' should only return an error if the timeout
56occurs -- however, pre-2.6.3 servers return version strings as an error
57which messes up that usage...
58
59Macro's are, for want of a better word, stupid. Until parameter
60variables are allowed, they can not do very much. Common ones (and
61user set up options) should be defined in the file ~/.fsprc (i.e., .fsprc
62in your home directory). This filename can be changed by setting the
63environment variable FSPRC. An example .fsprc is supplied in the file
64`fsprc' in this directory.
65
66It is also now possible to have commands like:
67 ls -l | less
68Anything after the first `|' symbol is fed the output of the builtin
69command...
70
71rehash may need a little explanation -- when you do an ls, the directory
72information is fetched to the local end and stays there (thanks to
73the original authors stuff); if you want to check if anything has been
74updated since you did an ls of a directory, then you must rehash first.
75The client is now a bit more intelligent than it used to be (i.e., pre
76`d' release); it will automatically mark directories as out-of-date if
77it knows that it has modified them (e.g., by put, or mkdir or whatever).
78It's the wrong command name, but it brings back fond memories of csh...
79It may sound like a disadvantage -- in fact it is an enormous benefit
80having it work this way... *much* faster.
81
82All the commands which have a f<name> equivalent in the original
83distribution behave as before; well, pretty much.
84
85Oh, one last thing -- command lines can't have continuations...\
86yet :-)
87
88-------------------------------------------------------------------------------
89
90If you use it and like it, mail me. I need the encouragement :-) If you
91use it and don't like it, mail me. I need bringing back down to earth...
92
93Basically, even though it is pre-alpha, I would still like to know
94whether people like it or not... just don't complain too loud when it
95doesn't work right :-)
96
97enjoy,
98pihl <pgr@prg.oxford.ac.uk>
99