[svlug] open source mimicking existing software?

Sanatan Rai sanat at stanford.edu
Wed Mar 2 15:13:15 PST 2005

: Here's a question that came up in one of our MBA courses: "Is open
: source software always something that mimics an existing software?"
: I haven't given it any exhaustive thought...anyone have any examples
: of the contrary? -- Sameer

	TeX is open source, and when it appeared there was no
equivalent. Not clear what you mean by mimicking. TeX is a typesetting
system, so in some sense it would have mimicked typesetting systems that
were extant, had there been any in the early 1980s. (Well there was

	On the other hand, GNU software was explicitely supposed to
provide free versions of tools that were available on most unix systems.
So it mimicked extant software inasmuch as hp-ux's implementation of
more mimicked BSD's. Since GNU's version tended to be better in the
sense that they provided additional functionality and standard options,
I am disinclined to regard them as clones. Of course, there are quite a
few GNU tools that had no real non-free equivalents, Emacs for instance.
(People who thing Emacs = TECO++ are dangerous to say the least.) By the
same token games such as GNU Go are definitely _not_ clones of extant
systems. Look around the GNU archives, you'll find many such examples. I
should guess---for instance---that many of the network
troubleshooting/analysis tools were new when they appeared.

	How about the Linux kernel? It evolved from the source code in
Tannenbaum's book on OSs. So again, it isn't really a clone of anything.

	In fact, I should argue that most of open source programmes are
not clones of extant software, as they have distinct native modes of
execution. I think the perception that OSS aima to mimic extant software
is due to the fact that many users started life on MicroSoft products,
and on migration to Linux, sought the same kind of tools. Thus, they
ended up using OpenOffice and similar programmes. Naturally, they tend
to conclude that all that open source people wish to do is to provide
free versions of Windows software.

	You might find it interesting to talk to native Unix users, ie
people who started life on Linux/Unix platofrms. You'll be surprised as
to what they regard as useful software. For instance, the first thing I
do on getting a Windows box (say, at work so I can't format the HD and
install Linux), is to install MikTeX and Cygwin on it, so that it
actually becomes useful to me. I do not use Microsoft Office though my
expertise extends to being able to programme VB applications therefor.
However, life without bash, less, cp, wc, file, g++, python, latex, sane
piping, sane I/O direction,...would be very, very difficult. Ah, yes,
language tools and compilers are another example of tools that strictly
speaking, are not clones. As in the case of the more example, gcc cannot
be called a clone of cc, simply because each Unix vendor provides his
own version of cc, most of which are dismal compared to gcc.

	Hope this gives you some food for thought!

Sanatan Rai,                      | EMail: sanat at stanford.edu
Dept. of Management Sc. & Engg.   | Home: 143 Ayrshire Farm Lane #104
Stanford University,              |       Stanford, Ca 94305.
Stanford, Ca 94305.               | 'phone: (650) 498 1655 (R)

More information about the svlug mailing list