TMG-L Archives
Archiver > TMG > 2000-10 > 0970583133
From:
Subject: Re: [TMG] File extension .fpt
Date: Tue, 03 Oct 2000 09:25:33 -0500
References: <4.1.20001001091221.00a8da60@pop.mis.net><4.1.20001002111717.009322c0@pop.mis.net><4.1.20001002193416.00931340@pop.mis.net><4.1.20001003093201.00b6b800@pop.mis.net>
Lee,
I have always thought that each dataset was a separate unit with a
kind of protective wall around it; that each of the folders containing
different kinds of data (little databases) were peculiar to one
dataSET; and that "the dataBASE" referred to the entire entity of one
or more dataSETS or the various types of data (little databases)
entered in or attached to a single dataSET. Thus, under this thinking,
damage to one dataSET would presumably not affect the other dataSETS.
I removed Booth_i.fpt which contained 849 kb for which I could not
find an attachment to a person or an event. This action caused damage
in four other dataSETS in many different kinds of files, other than
fpt files. As a consequence, I can certainly accept the accuracy of
the statement that individual dataSETs are intertwined, and I am led
to the conclusion that the protective curtain around each dataSET does
not exist. Is this conclusion consistent with your explanation of the
relationship between dataSET and dataBASE?
Marilyn Symonds
> You should keep in mind that TMG does not maintain your data in a dataBASE file. Instead it uses a dataSET to store data. A dataSET consists of many dataBASES that are all tied together by pointers and indexes. Thus the deletion of one of the dataBASE files causes the TMG internal tests to give an error and require the running of the Repair function. The internal tests are TMG's way of maintaining the internal integrity to insure the safety of your data. So when you delete a file (even if it only contains deleted data), the Repair function may have to "jump through hoops" to try to fix the problem because the dataSET is so inter-twined.
>
> Thus in your case, the idea of copying also includes some "packing" of data (optimizing), but on a more complete basis in that only valid (meaning not deleted) data is copied. This give you a new dataset with only current undeleted data in it.
This thread: