Hi Prasad, thank you for your interest in DROID and PRONOM,
We aim to create a single PUID for each distinct format we add to the PRONOM registry. A PUID may contain information such as file extensions, mime/media types, descriptions, links to specifications and most importantly from our perspective, signature sequences for identification.
On occassion we do have need to update certain entries within a PUID. This is because our research is not static and we may discover problems, or improvements we can apply to a particular entry. Often these are identified and contributed by members of the wider DROID/PRONOM user community. It may be that a particular signature sequence is too weak, so identification is clashing with a different format, or that our initial understanding of the construction format was flawed - most often if a file format technical specification is unavailable and we have therefore had to make certain presumptions about an identifiaction mechanism.
We do not take such changes lightly, and endeavour to do this only where such a change improves that accuracy of the registry.
Since March 2010, we have recorded these changes on our 'release notes' page on The National Archives website -
http://www.nationalarchives.gov.uk/aboutapps/pronom/release-notes.xml. for more significant changes (i.e. those which may fully alter the identifation of a particular format, then we will write a more detailed explanation. We're always happy to answer any questions about why we may have made a change.
I hope this helps. Please let me know if you have further queries or any concerns.
David Clipsham (Support Engineer at The National Archives, presently with lead responsibility for the PRONOM registry)