** Reply to note from [log in to unmask] Wed, 13 Aug 2003 15:38:20 +0100
> Charles wrote:
> "When you have the new system delivered you could migrate ALL the data and
> keep the old system running whilst you satisfy yourself that the new system
> is working properly. However much you test with dummy data you will HAVE
> TO put the real data on and still check it"
>
> I would classify this as parallel running/implementation. Agreed that you
> have to do it - but it should only take place after the system has been
> thoroughly tested using specially created test data.
Yes agreed. I was thinking of a system as being fully tested (by whom?) by the
time it is delivered. The perils of developing apps for which I am the primary
tester.
>
> Stuart mentioned:
> "appropriate volumes of data and/or combinations of data values"
>
> To take the latter first, ensuring all combinations of data are tested is
> more easily done by generating test data - because you just don't know,
> without (time consuming) searching, which live records contain the
> combinations you are after. As regards volumes, de-personalisation should
> not affect volume testing so there is no need to use data which can
> identify real people.
Yes.
Charles
==============================================
Charles Christacopoulos, Management Information Officer,
Planning & Information, University of Dundee, Dundee, DD1 4HN,
Scotland, United Kingdom. Tel: 44(0)1382-344891. Fax: 44(0)1382-201604.
http://www.somis.dundee.ac.uk/
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
All archives of messages are stored permanently and are
available to the world wide web community at large at
http://www.jiscmail.ac.uk/lists/data-protection.html
If you wish to leave this list please send the command
leave data-protection to [log in to unmask]
All user commands can be found at : -
www.jiscmail.ac.uk/user-manual/summary-user-commands.htm
(all commands go to [log in to unmask] not the list please)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|