![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to myths.tex CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [CSRG BSD Unix] / 43BSD / contrib / mh / papers / myths|
Return to myths.tex CVS log | Up to [CSRG BSD Unix] / 43BSD / contrib / mh / papers / myths |
1.1 root 1: % run this through LaTeX
2:
3: \input lcustom
4: \draftfalse
5: \input version
6:
7: \documentstyle[12pt,DScustom,sfwmac]{article}
8: \setcounter{page}{0}
9: \pagestyle{empty}
10:
11: \begin{document}
12:
13: \title{The Rand MH Message Handling System:\\
14: Myths about MH}
15: \author{Marshall T.~Rose\\
16: Northrop Research and Technology Center\\
17: One~Research Park\\
18: Palos Verdes Peninsula, CA 90274}
19: \date{\ifdraft \versiondate/\\ Version \versiontag/\else \today\fi}
20: \maketitle
21: \footnotetext[0]{\hskip -\parindent
22: This document (version \versiontag/)
23: was \LaTeX set \today\ with \fmtname\ v\fmtversion.}%
24:
25: \begin{abstract}
26: \noindent Mail handlers, like text editors, text formatters,
27: programming languages, and computer-communication network technologies,
28: have become the basis of religious wars.
29: Although the UCI version of the Rand Message Handling System is well beloved
30: (and deservedly so) by its disciples,
31: there remain those who spread non-truths about \MH/.
32: This document seeks to set the record straight.
33:
34: Of course,
35: the gentle reader should understand that the author uses this forum primarily
36: as a non-violent release of hostility.
37: Note however,
38: that this document is not a case of ``me against them''.
39: Rather it is a case of ``us against them''.
40: There are a lot of \MH/ supporters,
41: though
42: I don't pretend that this paper represents anyone's views other than my own.
43: Furthermore,
44: I only write like this when I'm upset.
45: As such,
46: this paper shouldn't be taken {\em too\/} seriously.
47: \end{abstract}
48:
49: \bop\pagestyle{plain}\pagenumbering{arabic}
50:
51: \section* {The Plain Facts}
52: I really hate writing, especially documentation.
53: I don't mind having written so much, but I don't like writing.
54:
55: To put the reader in the right mood,
56: let us recall the words of Lord John Whorfin:
57:
58: \begin{verse}
59: Sealed with a curse,\\
60: \qquad as sharp as a knife;\\
61: doomed is your soul,\\
62: \qquad and damned is your life.
63: \end{verse}
64:
65: This is my standard response to people who want more documentation on \MH/.
66: The \MH/ documentation set is way too large as it is.
67:
68: Now with that out of the way,
69: onto the plain facts.
70:
71: \begin{enumerate}
72: \item Speed versus Performance\hbreak
73: For some reason,
74: people think that \MH/ is {\em slow}.
75: This is not true.
76:
77: It is true that \MH/ is {\em slower\/} than some monolithic user agents.
78: The reason for this is that since each \MH/ command is a \unix/ program,
79: there is a larger initial cost for running each \MH/ program.
80: Of course,
81: once an \MH/ program has loaded its state information,
82: it can execute quite quickly.
83: In particular,
84: a lot of time has gone into tuning \MH/ towards fast execution.
85: (Everyone owes a hearty thanks to Van Jacobson who did most of this work.)
86:
87: It is important to distinguish between how fast a system runs,
88: what work that system does,
89: and the ratio between the two.
90: Although \MH/ may not be fast,
91: it does perform well.
92:
93: \item Hardware/Software Dependencies\hbreak
94: For some reason,
95: people think that \MH/ is riddled with \vax/, \bsd/~\unix/, and/or \SendMail/
96: dependencies.
97: Nothing could be further from the truth.
98:
99: \MH/ does run on all \bsd/ releases of \unix/ since 4.1\bsd/.
100: It also runs on V7~\unix/ and various \xenix/ variants of \unix/.
101: Recently, the \MH/ distribution has taken to supporting the AT\&T variant of
102: \unix/, System~5.
103: \MH/ does not have support for System~3.
104: If someone would like to port \MH/ to System~3,
105: please let me know.
106:
107: \MH/ does run on Digital Equipment Corporation's \vax/-family of computers
108: (providing that host is running a \unix/ mentioned above).
109: It also runs on a large number of other hosts,
110: such as SUNs, Integrated Solutions, Pyramid-90x's, Gould FireBreather's,
111: ALTOS's, 3B2's, \pdp/-11's, and so on.
112:
113: \MH/ does run with \SendMail/ as its message transport agent.
114: It also runs with \MMDFI/ and \MMDFII/,
115: and provides it's own stand-alone delivery system with \UUCP/ support
116: (e.g., \pgm{rmail\/}).
117: Furthermore,
118: if your host can make an SMTP connection to another host,
119: you don't even need to use any of these programs.
120: You can simply instruct \MH/ to open an SMTP connection,
121: using a flexible search-list of service hosts (and network),
122: to any service host accepting mail.
123: The bottom line is that \MH/ can really run with any 822--based message
124: transport agent.
125:
126: \item The BBoards channel and distribution lists\hbreak
127: For some reason,
128: people think that the UCI BBoards facility is good only for local receipt of
129: BBoards.
130: It is true that the UCI BBoards facility provides excellent support for local
131: BBoard subscribers.
132: However,
133: the UCI BBoards facility also supports continued distribution.
134:
135: The BBoards channel does both local delivery and remote distribution for lists.
136: For local delivery,
137: it delivers messages into maildrops in a spool area.
138: In addition,
139: it allows the \MH/ user to shorten the typein of distribution list addresses.
140: For remote distribution,
141: it does the usual list exploding and error trapping.
142: \end{enumerate}
143:
144: \section* {The End}
145: That's it for now.
146:
147: \bibliography{myths}
148:
149: \showsummary
150:
151: \end{document}
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.