It always amazes me that people who use IT extensively often end up trying to catalogue things in the same way they did manually. This naming convention for files is a perfec example. The simplest way is to develop a directory structure which is sufficiently flexible to pidgeon hole files and then give them names that relate to a date, a place, a person, or a project in some logical way. Of course the logical thing to do is use a database, either flatfile or relational, and insist that people register files they wish to retreive in the database. Then you can let your computer search for it! Such is the stuff document management is made of.
Hi all,
I am relying on your combined brain power and years of experience for the solution to the following question:
*Problem*: People complain they can never locate a document. Therefore two possible scenarios exist which require two solutions. (1) Where they know the filename (2) Where they don't know the filename
*(1) Where filename is known (ie reference is made to the printed doc):*
Currently a minimum of 5 document repositories are in use!..these range from using the mapped network drives to store files to placing them in Software Request Management systems to Test Management systems to Development Tracking systems etc.
Past Solutions tried but failed: The following file naming convention was put in place but there were several flaws with the solution:
PrimaryProject.SecondaryProject.TertiaryProject.SubProject.Software Release.SoftwareVersion.DocumentRelease.DocumentVersion. DocumentType.extension
*Flaws to the solution:* It doesn't always give you the location of the document as it sits in say Perforce, DevTrack or Test Director for example. The filenames are WAAAAY too long... The filenames become nonsensical in that you could have:
TRAIN.TRAIN.TRAIN.TRAINDEV.R1.03.V03.R1.00.V02.PIR.doc
Most of the documents generated are a snapshot particular to a project A document can span multiple projects (up to 3/4 at the most complex)
The developers have expressed an interest in keeping all documents in the Perforce Software Request Managment tool as this allows for versioning and tracking of differences etc. but not everyone uses/likes to use Perforce, especially the testers who use another Test Management tool.
*(2) Where filename is unknown* In this instance, I was thinking of using a table that sits behind the title page (can even hide it if it freaks people out) that details all the possible path/owner/author/developer etc information that can be inserted via docproperty fields. Then if someone needs to search, they can enter a text string and the document should surface.
The last thing management wants is to introduce yet another document management tool so I am a little perplexed by the situation. Any ideas? :)
Cheers Nikki
-- Rod Stuart 1/19 Thrall Street Innaloo, WA 6018, Australia (08) 9204 2957 <rod.stuart@xxxxxxxxx> (042) 813 5605