Print

Print


Looks like you are having a fun time!

 

Of course Taxonomies and Fileplans are different things.  There are, of
course, close relationships but (and this is the killer) they are not one to
one.  I suspect that any use of one as the other will indeed defeat the
purpose.  Indeed I suspect that the manifest unsuitability of the taxonomy
as a file plan will not only be ignored as a fileplan but affect the
credibility of the taxonomy by association with the failure.

 

Getting a coherent overall fileplan adopted is quite enough of a challenge
without this sort of nonsense.  More people outside the Information
Management community will see the value of a good fileplan than will ever
buy into a taxonomy.

 

Regards

 

Jim

 

==================================================== 
J.S.M.Whitaker 

Mercia Information Ltd
Cocksian Cottage                 
Banks Green                    07798 702428 
Nr Redditch                      [log in to unmask] 
Worcestershire                 Skype: J.S.M.Whitaker
B97 5SU 

 

From: The UK Records Management mailing list
[mailto:[log in to unmask]] On Behalf Of DCSA DST-IM4
Sent: Tuesday, March 27, 2007 5:39 PM
To: [log in to unmask]
Subject: Taxonomies and fileplans - any help appreciat ed!

 

Hello all, 

Firstly, apologies for those of you on more than one group that get multiple
copies of this. 

I am looking for some advice on a tricky issue.  We have developed a new
functional fileplan for our organisation, which will use a taxonomy as the
metadata.  As part of a merger we are now working with others to combine our
fileplan with theirs, to have a consistent one across the organisation.  

My belief is that taxonomies and fileplans are different things, with
different functions and capabilities - they are not interchangeable.  One of
the members of our working group is determined that we use the taxonomy, as
it is already there, and even though it isn't very good and isn't fit for
purpose we should use it, because it's there.  I have also been told that
anything we implement won't work, so why bother doing any hard work when
it's going to fail anyway!

Does anyone have any arguments or evidence that prove one way or the other
which is the more sensible approach.  I would prefer ones that support my
argument, but if there is overwhelming evidence that using the taxonomy is
the best way forward, I am prepared to be open-minded about it.

Any advice or guidance would be much appreciated, and if I get enough
responses I will be post them back to the lists 

Thanks 

Megan 

Megan Roberts 
Information Manager 
DCSA Directorate Strategic Transition 
Room 004 
Building 111, 
Basil Hill, 
Corsham. 

Tel. (Mil): 94382 8602 
Tel. (External): 01225 81 8602 
Tel. (Mobile): 07785337671 
E-mail (Internal): DCSA DST-IM4 
E-mail (External): [log in to unmask]