Power Programmierung/Power-Programmierung CD 2 (Tewi)(1994).iso/doc/mir/04how

Text File | 1992-06-29 | 14KB | 278 lines

        ═══════════════════════════════════════════════

            4.    HOW THE MIR PROJECT WORKS FOR YOU

        ═══════════════════════════════════════════════


        ═══════════════════════════
4.1           "Free" software
        ═══════════════════════════

            In the MIR project we are using the "copyleft" strategy
of the Free Software Foundation.  The Foundation's GNU General
Public License is included as Topic Five; it applies to all
software created as part of the MIR project.  This software has
been created specifically for this purpose by Marpex Inc. since
March 1991.

            The Free Software Foundation

                "is dedicated to eliminating restrictions on
                copying, redistribution, understanding, and
                modification of computer programs.  [They] do this
                by promoting the development and use of free
                software in all areas of computer use... 'Free'
                pertains to freedom, not to price...  You have two
                specific freedoms once you have the software:
                first, the freedom to copy the program and give it
                away to your friends and co-workers; and second,
                the freedom to change the program as you wish, by
                having full access to source code.  Furthermore,
                you can study the source and learn how such
                programs are written.  You may then be able to
                port it, improve it, and share your changes with
                others."

            What is copyleft?

                "The simplest way to make a program free is to put
                it in the public domain, uncopyrighted.  But this
                allows anyone to copyright and restrict its use
                against the author's wishes, thus denying others
                the right to access and freely redistribute it.
                This completely perverts the original intent.

                "To prevent this, we copyright our software in a
                novel manner.  Typical software companies use
                copyrights to take away your freedoms.  We use the
                copyleft to preserve them.  It is a legal
                instrument that requires those who pass on the
                program to include the rights to further
                redistribute it, and to see and change the code;
                the code and rights become legally inseparable."

            Quotes in the above three paragraphs are from page 3 of
the January 1992 "GNU's Bulletin" semi-annual newsletter of the
Free Software Foundation, 675 Mass Avenue, Cambridge, MA  02139
USA.

            The argument for this strategy is set out nicely in an
article "Programs to the People" in the February/March 1991 issue
of the M.I.T. Technology Review.  With permission of the author,
Simson L. Garfinkel, the text of the article is included in a
separate file on the CD-ROM release.  The file is named "TOPEOPLE".


        ════════════════════════════════════════════
4.2           Interactive shareware publishing
        ════════════════════════════════════════════

            The Mass Indexing and Retrieval (MIR) project is
releasing five sets of shareware tutorials.  Shareware has three
advantages for the user:

        »   easier access through broad exposure on electronic
            bulletin boards and copying for friends;

        »   opportunity to review tutorials prior to making a
            commitment... minumum risk and no surprises;

        »   much lower prices since normal marketing costs are
            bypassed.

            Seed funding for the MIR project was provided by the
Canadian government with the understanding that the underlying
indexing and retrieval techniques developed in the project shall be
made broadly available under copyleft rules.  Personnel from two
companies are carrying out the project.

            Innotech Inc. of North York, Ontario (416 492-3838)
aims toward excellence in CD-ROM publishing services.  It is
developing interfaces and applications based on MIR technology.
Innotech offers consulting services as well as service bureau
processing in CD-ROM publishing.

            Marpex Inc. is a firm founded in 1976 by the author of
the tutorials and the related software.  Marpex developed the
techniques and pilot programs for the pioneering FindIT CD-ROM
system, and more recently collaborated in the design of the Discis
Knowledge Research CD-ROM books.  Marpex provides consulting in
records management, and seminars related to the techniques in the
MIR tutorials.

            MIR tutorials are designed to be an exercise in
co-operative development.   We hope to engage you, the readers and
users, in the project.  We know that co-operative development will
lead to improved end results; many minds are better than one.  Text
and software is modified according to your input... clarifications,
improved methods, more powerful source code, etc.  Each tutorial
will evolve to reflect significant improvements, with your name
attached to the improvements you provide.

            After the interactive phase is over, Marpex hopes to
compile a reference text based on the tutorials.  This will be
accompanied by a CD-ROM containing all software and support files.
Since ISO 9660 CD-ROMs are operating system independent, your
ported versions of programs can be included.

            Why not release everything at once?  Reasons for
progressive releases are:

        »   Scope of the project:  Look at the tables of contents.
            There is simply too much for one tutor to complete in
            a single step.  Extensive new research is continuing to
            be carried out, particularly in concept recognition.
            Apart from standardized functions, we are not carrying
            forward source code used in any proprietary system.
            Much of this work in the past has been on UNIX
            workstations; now we are achieving levels of efficiency
            that can make preparation of large databases feasible
            on a personal computer.

        »   Market readiness:  Until the introduction and Tutorial
            ONE have been on the market for a few months, we do not
            know if our target groups are sufficiently interested.
            We want to know that our work is meeting a genuine need
            and that co-operative development under shareware and
            "copyleft"  rules is viable.

        »   Financing:  The Canadian government provided seed
            funding, that is, enough to get the project off to a
            good start.  We are using the same approach as the Free
            Software Foundation to provide the money required to
            carry the project forward.  Their major financing is
            through distribution of tapes containing their work -
            at roughly $200 for each of several tapes.  We aim to
            carry forward the MIR project through distribution and
            shareware registrations.  People are free to make
            copies of all materials.  We trust that buyers will
            honor the shareware provisions for the tutorials.


        ═════════════════════════════════════════
4.3           Engine-independent techniques
        ═════════════════════════════════════════

            The ISO 9660 CD-ROM standard and Microsoft's MS-DOS
extensions opened the way to accessing the files on any conforming
CD-ROM.  But having access to files is not the same as being able
to search conveniently.  Because indexing systems and interfaces
are proprietary, the user has been faced with the nightmare of
having to learn a new retrieval method every time a CD-ROM title is
purchased from a new vendor.  The plea goes up:  "Why can't I use
the same program I've already learned?"

            Why not, indeed?

            Two ideas have emerged in the literature.  One is full
"interoperability"...  the ability for a person to select her/his
own preferred retrieval interface software and use it to search
within any CD-ROM title on any CD-ROM drive under any operating
system.   That's far off yet.  The second idea, a subset of the
first, is now before a Standards Committee (SCAD) of the
International Standards Organization (ISO) and may show up in
commercial products in 1993.  That is the possibility of separating
the software into a client interface and an underlying server which
fetches data from the CD-ROM.  The server module resides in RAM and
communicates with the client interface through standardized ASCII
strings.  The intention is that the server is specific to the data
and the indexes in place; the client interface is the user's
preference of any retrieval software conforming to the standard.

            These engine-independent techniques do away with the
high cost and inconvenience of re-education.  There are perhaps
five contending proposed standards.  The Information Handling
Committee of the Intelligence Community Staff in Washington, D.C.
has commissioned the CD-ROM Read-Only Data Exchange Standard (CD-
RDx).  The aircraft industry appears seriously committed to
Structured Full-Text Query Language (SFQL), an extension of the ISO
approved SQL.  Other contenders are V39.50 (a library system
networking protocol), Silver Platter's DXS, and DFL, an earlier
outgrowth of Standardized Query Language.  Unknowns at this point
include the data structures supported (whether columnar relational
databases and subsets thereof, or whether more generalized forms),
and the actual syntax of messages that pass between the interface
and server modules.

            We believe that cooperative development through the MIR
project can contribute to this process.  If software is freely
available under copyleft rules, it can be adapted very readily as
standards evolve.  No-one has to hold back until the Standards
Committee makes its one year or three year or five year report.

            We also believe that it is unnecessary to limit the
discussion to CD-ROM.  The basic problem (frustration at being
forced to learn new interfaces) is independent of the medium on
which the data are stored.  MIR technology may be applied to data
held on hard disk, floppy diskettes, Write Once Read Many (WORM),
Bernoulli, rewritable laser optical disks, laser cards or whatever
other media can retain data as byte streams.


        ═════════════════════════════════
4.4           The software provided
        ═════════════════════════════════

            Scope:  The source code for data analysis and
preparation, search term selection, and to some extent automated
indexing require little interaction with a user.  The programs in
TUTORIALS ONE through THREE are therefore considered complete.

            TUTORIAL FOUR presents an engine (a "data server
module") which may be used with interfaces compatible with engine-
independent techniques.  The number of different interfaces that
might be written is infinite.  Interface source code can be (and is
likely to be) handled in traditional proprietary ways, simply
because of the great variability in features that end users desire.
You or your firm may write a "client module" interface and keep it
proprietary, provided the data server module is kept separate and
under copyleft rules.  If you care to write a client module under
copyleft rules, and if it works well, we will be glad to pass it
along.

            The software provided with TUTORIAL FIVE might be
classed as "discussion starters".  We carry the discussion a fair
distance, but look to readers to pursue their specific interests.
In an ideal world, that pursuit would take the form of a public
exchange of ideas under copyleft rules.  As Captain Jean-Luc Picard
would say, "Make it so!"

            Naming conventions are applied to many of the programs.
DOS constrains source code names to eight characters plus a ".C"
extension.  Where a six letter name is workable, a single letter
followed by an underscore precedes the name and has one of the
following meanings:

                    A_*.C    analyze, report
                    B_*.C    build indexes
                    C_*.C    compress / integerize data
                    E_*.C    expand content of a file
                    F_*.C    filter out parts of a file
                    I_*.C    invert token matrix
                    J_*.C    join words into useful phrases
                    M_*.C    merge files
                    P_*.C    preprocess particular layouts
                    Q_*.C    quality assurance
                    R_*.C    rotate content within a line
                    S_*.C    server module for retrieval
                    T_*.C    transliterate language to ASCII

            Support files include 05LICENS, OVERVIEW, COPYRIGH,
FRONTISn, NEWSREL and ORDER.  On the diskette version, the install
program gives you a choice of whether to extract the files in
WordPerfect 5.1, ASCII, or a generic form suitable for other word
processors which can handle ASCII files.  05LICENS is the Free
Software Foundation's GNU General Public License which governs
permissions for software supplied with the tutorials.  CD-ROM
release(s) contain extra worked examples, and articles such as
TOPEOPLE.

            We recommend you place executable copies of all
programs in one area on your hard disk.  That way, you can create
easy access to the programs with only one small addition to your
DOS path (something of the form "\C:\BIN;" added to the PATH line
in your AUTOEXEC.BAT file).