A&B Computing
1st December 1984
Categories: Review: Software
Author: Mark Webb
Publisher: Silversoft
Machine: BBC Model B
Published in A&B Computing 2.01
What useful features can Silversoft's database offer the disc system user?
Sorting On Index
Silversoft has set out to provide an easy-to-use disc based filing system, a no frills but equally no complications, database. It's a standard sequential file with a separate index held in RAM, and stored as a separate file on your data disc.
Index comes on disc and !BOOTs up to a title page and an initialisation sequence. This sequence sets up the program to work with the combination of hardware available and in use for this particular session.
The hardware concerned is disc drives and printer. If we answer "yes" to printer, we are invited to set the BBC to send or not to send carriage returns, to enter a short sequence of control codes to trigger special print facilities (Epson compatible) and to specify the page length we wish to work with. The program will work with a single or double disc drive configuration. With a single drive, prompts are issued throughout the program to remind the user to swap between data and program discs.
Index does not however utilise the second side of a double-sided disc, and most 80 track drives are double sided. This means that only two files can be stored on an 80 track disc. A 40 track 100K disc drive can cope with only one per disc. This fact could be a considerable constriction if a number of separate small files need to be stored. From a management point of view however, one file to each clearly labelled disc is a good idea.
Atlhough at this stage we can tell Index that we do not wish to use a printer this session, unfortunately this does not stop it asking us whether we wish to print records later on, when we may be just examining records.
Planning and setting up your file, the records and fields within it, is the most important stage of the initialisation process. Index allows us to change the definitions of field and record that we specify at this point but it is wholly desirable not to have to enter into the time consuming process of re-initialisation.
Space Restrictions
Index allows 15 fields within each record, a maximum of 30 characters per field, but only a maximum of 250 characters overall. It is surprising how quickly characters can be eaten up by the simplest records and if 250 looks like a small number, do some mental arithmetic with your records before using the software.
There are also two internal restrictions on the field and record lengths of Index. If you wish to make use of the label printing facility then account must be taken of the label width when assigning numbers of characters to each name and address field. Secondly the link to View (and since Index was called Viewbase before Acorn stepped in, it may well be purchased with the intention of combining the two) is restricted by the number of characters allowed within a single View macro - 132. In fact with punctuation and control code overheads this becomes 120 in practice. Once again, if this figure looks a little on the small side then do some forward planning to see if it can encompass the sort of records you are going to keep. Index is ideal for the name, address, phone number type of record keeping but a personnel file or resource database would not fit.
Index initialises its file according to the original specification and any under utilisation will be wasted storage on your disc. The manual is very good in this respect and describes the best way to handle initialisation including the suggestion that a test run be made with only a few records entered. When you have decided upon the format of your file, the program creates a file of your chosen name on disc and a similarly named file in directory D. This second is the index file, pointing the program to the records stored on the disc.
Re-initialisation, if the need occurs, does not actually change your file in any way but rather creates a new framework into which the data is read, and for which a separate disc is needed. You tell Index which old field links to which new field. Watch out if you are reducing the size of fields since Index will cut the data to fit the new format. Fields can be inserted or deleted.
Field Trials
Index is menu-driven and operations from now on involve making choices 1 to 9 for entering, altering, deleting, examining records, printing the file, file definition, labels and linking to View. Entering records is not as easy as it might be. For instance, you have to hit Space and Return to leave a field blank. At the end of each record, we are asked if the record is OK to store. "Yes" transfers the data to the disc buffer and "no" offers the chance to edit the record or start again.
Each field of the record has to be typed in if it is to be changed, although Return will take you on to the next field. Each record is a simple field name and data combination but some editing with the cursor keys would make altering records less of a chore.
There is no way to economically reproduce regularly used data. The price of a number of items in stock might be the same, but you will have to type each one individually. The field names are there to remind you what information is required and these are not printed out with the records. When a session of entries is over, the TAB key initialises the saving of the file definition. Records are continually being moved into the buffer space and out to disc (automatically achieved by the DFS).
When considering the security of your file, the definition file must also be taken into account. If it has not been updated then even records on the disc will not be retrievable. I had the unfortunate experience of losing the definition file to a flawed disc. All the records were intact but apparently beyond reach. Fortunately Index is written in BASIC and not protected. It is possible to construct your own definition file but this sort of repair work can be avoided by regular backing up of both files and, I suggest, hitting the TAB key now and again to save your file definition - just in case.
Sorting And Selecting
All the options, alter, delete, examine and print file are straightforward in operation. Each time the name of your data file has to be specified and the file definition displayed before you. All these functions have access to the fairly powerful sorting and searching facilities of Index. Records can be chosen according to any field in the record. Any selected set (searched out by string or number with > and <) can be displayed in sequence, sorted according to any field and displayed, or further reduced by selecting a further subset of the file according to any field.
These operations enable you to pinpoint records, to skip through them, editing, deleting or printing. Unfortunately when you merely examine, you still get the choice of printing or not. There is no going back if you hit the wrong key. With altering and deleting, Index asks for conformation that it has found the correct record before starting anything drastic.
The print option lets you print the whole field, or just selected subsets. You specify the number of lines per page and whether a pause should take place at the end of each page. The records are printed out as seen on the screen display, without field titles but with an underline separating each record. The 30 character width does not make full use of your paper.
Into View
This is potentially the most powerful aspect of Index - its ability to produce a file which can be read into the Acornsoft View wordprocessor. The action is designed to allow the merging of names and addresses with word-processed text, utilising View's macro facility. A macro can be set up so that it will incorporate, on meeting with a control code, a specific string of text within the main body of a letter or form. In this way the same standard letter can be printed but with individual information on each.
The link between Index and View is made by creating a separate file, with your file name, in directory V. The restrictions detailed above apply and data can be lost if they are not adhered to. Trailing spaces are stripped off each field if they exist. The new file needs yet another disc and the process can be a bit monotonous if you have a single drive. Even with a double, it is fairly time consuming with around the 50 record mark. If you intend to get up to hundreds of records per file, it will undoubtedly take a number of minutes.
The file contains all your specified records ready to be incorporated into a View macro using the symbols @0, @1 and so on.
Summary
Index surprised me in how useful it could be, especially with my records which follow the simple name, address, data format, small chunks of information which need to be bashed in via the keyboard and searched or sorted. There is no sophisticated screen formatting, or form writing, such as you might find in a DataGem, File Plus or Cardiff disc database. The method of searching convinced me of its usefulness and the printed output is neat. The link to View is there for those who wish to use their wordprocessor for business mailing although this is not a particularly friendly mail merge when compared to a properly integrated package, such as the Merlin one promises to be and which the Acornsoft disc database will provide.
Now and again Index proved infuriating. There is no recovery from many situations you might get yourself into. For instance if your disc is not entirely empty, the program will crash with disc full error and you will not be able to reuse your carefully thought out file definition. Still, if you follow the instructions, things like that will not happen and Index will perform adequately.
There are also a few errors in the screen displayed text which I am sure the final version will iron out and one strange use of the word "printing" for writing to the disc.
The facilities of the more advanced databases are not available with Index but the price is fair at £24.99. You pays your money and you takes your choice!
This article was converted to a web page from the following pages of A&B Computing 2.01.