                  XTRACT Version 4.1 Overview
                  ===========================

           1.  What's NEW in XTRACT v4.1
              a.  An inline viewer option
              b.  NEW to search the 1% US-Sample
              c.  Extra search options implemented.
              d.  On-screen help is now available.
           2.  XTRACT v3.3 - Bug Fixes
           3.  What's NEW in XTRACT v3.2
           4.  Changes from XTRACT v2.2
           5.  What two of our customers say about XTRACT!



What's NEW in XTRACT v4.1
=========================

a.    The most significant addition for UK users may be the inline
      viewer option.  This means that you can configure XTRACT to
      give you the option, at the end of your search, to invoke any
      text BROWSE, or LIST utility to examine the results of your
      search.  You can choose any utility you wish, provided you can
      call it up with a single file name as operand from the DOS
      prompt - as, for example, you might have typed:-

          LIST WHATSNEW.TXT

at the DOS prompt to read this file.

      To activate this option, follow the instructions in the
      XTRACT.INI file, using the Viewer keyword to specify your
      choice of utility, for example:-

      Viewer=LIST.COM


b.    A 1% sample of the US 1880 census has become available to us as
      a result of a study undertaken within the history department at
      the University of Minnesota.  Version 4 of XTRACT has been
      released specifically to search and format this data, as an
      alternative to, or as well as the UK 1851 2% sample.  For a
      fuller description of the US data, please see the file
      US8_DESC.ZIP, which complements the equivalent for the UK 2%
      sample, UKC_DESC.ARJ.  This file should be available wherever
      you found your copy of XTRACT4x.ARJ [where 4x is the most
      recent version number].

      Note that in order to cater for additional information on US
      census, this version supports the following new keywords, with
      values which may be adjusted in XTRACT's configuration file:-

ShortBPL  -   by default, XTRACT displays birthplaces in short form,
              using ICPSR, or ISO abbreviations.  Specifying
              ShortBPL=No allows display in full.

ShortOCC  -   by default, the unregistered version of XTRACT displays
              occupations in general form at all times.  The
              registered version displays detailed occupational
              information, but it can be forced into general form by
              using ShortOCC=Yes.

ShortREL  -   by default, the unregistered version of XTRACT displays
              relationships to heads of households in a very general
              form at all times.  The registered version displays
              more detailed relationship information, but it can be
              forced into the more compact general form by using
              ShortREL=Yes.

ShortSick -   by default, XTRACT displays information (when it is
              present) as to whether an individual is disabled, in
              terms of being blind, deaf, crippled, &c.  The US
              census form also contains a category for recording
              sickness on the day of the ennumeration, which may be
              independent of the former.  Specifying ShortSick=No
              forces a display of sickness on day of ennumeration,
              rather than in more general terms.


c.    The option to search on ALL fields in a census record, apart
      from ages has now been fully implemented.  So in particular,
      you can now search for matches of Relationship, and Marital
      Status; and Disabilities, or Country of Origin, as well as
      previously, Forenames, Occupation, &c.

      Note also support has now been added for the Maxdrive keyword,
      which was documented for version 3.x, but disabled.


d.    On-screen help is now available - the brief "Usage" screen
      which was displayed if the wrong parameters were supplied has
      been expanded, and help text is immediately available on
      typing:-

          XTRACT -h
      or
          XTRACT -?

      as well as if the command line parameter syntax is incorrect.



XTRACT v3.3 - Bug Fixes
=======================
XTRACT v3.3 is essentially a bug fix release.  A list of new features
arriving with XTRACT v3.2 follows this section.

The following is a list of bugs which have been fixed:-

1)    Fixed a bug present since version 2.2.  If searching for a
      valid three letter surname in a single county, say:-

          XTRACT DBY LEE

      it would fail to recognise DBY, and proceed merrily to search
      the surname index, and pull out all LEE.

2)    Fixed another bug which was also in version 2.2.  It couldn't
      find the archiver if it was in the startup directory (reported
      ARJ failed with an errorlevel of -1).

3)    The maximum number of working drives has been adjusted to
      reflect the current working drive.  This will hopefully cater
      for systems with more than 4 fixed logical drives.

4)    If requesting .CSV output, the parish name was omitted when
      switching parishes.

5)    The IBMChars keyword was being ignored if set to "NO".



What's NEW in XTRACT v3.2
=========================

1)    The ORDER, and displayed WIDTH for fields (surname, forename,
      &c) in the output files are now configurable.  See the
      FieldSize keyword in XTRACT.INI for details.

2)    You can now search on fields other than surname, and combine
      your targets logically, eg:-

          XTRACT S(SMITH) & F(JOHN)

      [Search for Surname (SMITH) and Forename(John)]

3)    At the other extreme, it's now possible to disable all
      searching, so that, for instance, you can select a whole county
      to view.

4)    Or you can specify "wild-card" selection.  This is not a true
      wild-card to "match anything", but rather to select any entries
      which otherwise would be discarded by the normal matching
      process - for instance blank surnames, or surname "N K".

5)    It is now possible for XTRACT to read the census data in a
      variety of archived formats, for instance .ZIP files.  You can
      also specify the input file mask to use, to override the
      default UKC_ccc.ARJ.

      NB: XTRACT will NOT read file formats other than the original
      2% standard ASCII text files.

6)    For those with a lorra lorra disk space, the new XTRACT can
      read the *.T* files directly.

7)    You can limit the year you search by specifying a "mask"
      parameter for selecting from the *.T* files.  See the
      TxtFileMask keyword, in XTRACT.INI for details.

8)    Input and output directories are configurable.  See the file
      XTRACT.INI, and the XTRACT Reference Manual for how.

9)    You can add archives containing your own data to the
      UKC_ccc.ARJ series to be searched.  Typically these would be
      "one name" collections.  Naturally, you do need to ensure you
      create the same format and file structure as the 2%, original,
      but thereafter, all you need to do is choose a unique trigraph
      to replace the 'ccc', and to enter this, and a description of
      contents, in your XTRACT.INI file.

10)   It's now possible to define a "scratch" directory to unpack the
      census data to.  This may be particularly useful for users who
      need to keep their data files on floppy disk, as their scratch
      directory can be on their hard disk, or - for instance - a RAM
      drive.  You will still need sufficient free space to unpack the
      data, however.

11)   You may now select comma delimited (.CSV) output (suitable for
      importing into a database or spreadsheet). XTRACT will report
      maximum field widths identified at the end of your search to
      assist you in setting up a fixed length record database.

12)   You get "cookies" in the registered version, and two
      additional utilities - XTPRINT for paginating output, and
      XCHART for producing a soundex chart of surnames matching
      the one you supply.

13)   XTRACT now displays a progress bar for each file searched, and
      both lists and logs the name of the parish being searched.

14)   There is limited support for changing screen colors.

15)   Other options are configurable - check out the XTRACT Reference
      Manual.



CHANGES from XTRACT v2.2
========================

1)    XTRACT v3.2 will by default OVERWRITE existing files of the
      same name on a fresh search.  Version 2.2 APPENDED to existing
      files, which you can still do in Version 3.2, using the
      OutputMode configuration option in XTRACT.INI, XTRACT's
      configuration file, and in fact, this is how it is set up in
      the distribution package.

2)    Results by default will be saved in the startup directory,
      rather than the directory containing the UKC_ccc.ARJ files. You
      can re-configure this using the RESULTS keyword in XTRACT.INI.

3)    If you specify a search for more than one Surname, results are
      written to one file only, rather than separate files.  However,
      you *can* force separate files, by  separating the surnames
      with a Caret symbol (^).  See the XTRACT Reference Manual for
      more details.

4)    Output in the same format as the original (using the '-' switch
      on the command line) will be written to a file with extension
      .TXT, rather than using the same extension as a standard
      search.  This is to enable you to retain the results of a
      previous search more easily.

5)    XTRACT v2.2 allowed a width of only two columns for printing
      Ages of Males.  This has been amended for v3.2 to default to 3
      characters, as it does for Ages of Females.  Note that if you
      are using the default FieldSize settings, and comparing the
      results from a search using an earlier version of XTRACT, the
      effect of this "bug fix" is that your search results file will
      be larger, almost certainly, than the earlier one.

6)    The original v2.2 SOUNDX utility is no longer distributed with
      XTRACT.  In fact this has been revamped, to become two separate
      utilities XCHART (to produce a printed chart of soundex
      matches) and XTSNDX (which is still under development, and will
      allow selection, and therefore filtering of the soundex matches
      identified.




What two of our customers have said about XTRACT
================================================

"I find it to be very effective and fast and overall a nice
improvement over v.2.2.  I particularly like the new progress
bars..."  TW, Hampshire.

"I had a number of the UKCxxx.ARJ files but the bother of having
these expanded files cluttering up my disks, the problem of reading
them and sorting out the wheat from the chaff was a pain.  XTRACT
removed that pain."  RD, Norfolk.


-+-
Rosemary Lockie, 2:253/188
September 1995
