For those of you who are interested, to install it all you have to do

is follow the instructions below. Remember, the program is free.

 

KWIC-CONCORD is supplied in the form of a compressed ZIP file

containing:

 

Concord9.exe            (the main program file)

Concord9.ini            (contains defaults)

Concord9.hlp            (helpfile)

Concord9 User Summary.doc (manual describing basic operations)

Readme.doc              (installation notes)

html.wp#                (codes required for searching HTML / SGML files).

 

To extract the files, all that should normally be necessary on

Windows systems is to either double click on the Zip file or right-

click on it and select the 'Open' option. Windows should then

automatically open the default 'unzipper' program. It should then

simply be a matter of selecting all the files and extracting them to

a new folder, preferably a sub-directory of 'Program files' with a self-

explanatory name like 'Concord9'. If you then open the folder using

'Explorer', you can create a shortcut for the EXE file and drag this

either to the desktop or to the Windows \ Start up \ Programs

menu (Windows 95+) or  to the WINNT \ Profiles \ Default User \

Start Menu \ Programs (Windows NT).

 

All that you need to do in order to start the program is double click

on the EXE file or the shortcut icon. As it loads, the program will at

the same time load the HTML codes and the defaults contained in

the INI file. New defaults can be created  when using the program

and saved upon exiting.

 

If you have an older Windows 95 or 98 (first version) system, you

may have problems with out of date system files (certain DLLs).

To look for updates, try contacting:

http://www.microsoft.com/windows95/default.asp or

http://www.microsoft.com/windows98/default.asp.

 

If you are building a corpus, keep this under a separate directory,

preferably in the root, so that any searches will not 'stray' into other

directories. If you create new directories for search and word count

results, make sure these do not lie in the path of your corpus. If

you are working on a network, you may not be able to create new

directories on protected drives.

 

If you have any queries or suggestions or suspect any bugs,

please contact me at jeremy.whistle@northampton.ac.uk.

 

The program is supplied free on an 'as is' basis to anyone working

within an educational institution or engaged in non-commercial

research. Its author cannot accept liability for any damage to files

that may be caused. The program has however been extensively

tested and no such damage has ever occurred.

 

You may use this program for your own use, including at home.

You can also install it in public work spaces for use by students

and colleagues.  If anyone else manifests an interest, would you

please ask them to contact me at the e-mail address above.

 

Updates will automatically be sent to any user known to the author.

 

PROGRAM LIMITATIONS

 

The program is theoretically capable of searching an unlimited

number of files and saving an unlimited number of results to file.

There is no limit on the length of files. The only limitations on the

size of concordances or word counts are imposed by the limits of

the hardware (disk and memory space) on which the program is

being used. Sorting of large concordances or wordcounts may not

be possible on systems with small amounts of memory. List boxes

have an approximate maximum (depending on memory available) of

32767 items.  This will only affect lists of keywords read from file if

an attempt is made to edit the list. The limit will not affect the

viewing of concordance searches or word counts as these are read

from disk.

 

The program will allow a total concordance of a corpus to be made

by first performing a wordcount and then reading in the alphabetical

list (contained in the 'lst' file) using the 'Enter keywords / from files'

option in the main window and using this as the basis of the

concordance search. In the case of a corpus containing several

tens of thousands of words (types as opposed to tokens), this can

take a long time, even a whole weekend. Also, wordcounts on large

corpora, eg several millions of words, can be slow because of the

sheer amount of sorting and compression that has to be carried out.

 

Jeremy Whistle

21 March 2002