General Comments

The tape database is implemented by storing all fields as control sequences, thus it is suggested that the database be constructed using reasonable-sized files; subsequently to make tape labels, only the necessary files be 38, otherwise TEX may run out of memory.

The tape database has been implemented with the aim that the database files be sharable amongst various users so long as every album has a unique <#63#>album-id<#63#>. In order to load a tape database, all one needs to do is 39 the appropriate files and subsequently type the 40 commands for each desired pair of albums. Following suggestions are made for conventions to be adopted so as to enable a large community to share files without much effort:

• Each album database file have a ``<#65#>.bum<#65#>'' file-name extension¹. For such files, the extension will have to be explicitly specified in the 41 statement, otherwise TEX will look for a file with a <#67#>.tex<#67#> extension.

• Separate files be maintained for different composers or groups.

• Any character can be used in the <#68#>album-id<#68#>, thus following formats are suggested for <#69#>album-id<#69#>'s
  42, or
  43, where <#72#>#<#72#> is an integer.

  For example, one may have a file <#73#>mozart1.bum<#73#> with various album descriptions of Mozart's compositions, each having an album-id of the form <#74#>mozart1.1<#74#>, <#75#>mozart1.2<#75#>, and so on. If a single piece of music spans both sides then we might label the two parts using another level of subscripts, for example <#76#>mozart.12.1<#76#> and <#77#>mozart.12.2<#77#>.

The only problem with above conventions is that if two users have a file with identical album-id's then the file will have to be copied and the album-id's changed.

Following page contains an example. Commands used to generate the label would be: 25pt

verbatim9#

<#101#>2 #center304# #center305#

verbatim10#