Anders Bo Neilsen is the Senior Adviser on Digital Preservation at the Danish National Archives
Thursday was yet another busy and versatile day here at the section of Digital Preservation of the Danish National Archives. As usual there were the daily audit reports and the results of the quality assurance of the ingested SIPs which once again were spit out (pun intended) by our QA system. The producers of the SIPs were notified and given a new deadline for resubmitting SIPs which we can ingest and digest. Almost all of the rejected SIPs were produced by national authorities, but one or two were actually produced by a colleague. A taste of one's own medicine can be bitter and hard to stomach. The errors were the typical ones: lack of context documentation, missing explanation of code values, broken referential integrity and poor conversions to TIFF.
Having dealt with the ingest problems we turned our focus on the next item in the process, the packaging and storage of the AIPs. We are in the process of storing five AIPs from five similar authorities ranging in size from two to eight TB. At first we could not understand the huge size of these AIPs produced from ordinary digital case and document management systems. It seems that many incoming documents are an order of magnitude larger than the outgoing. Apparently, quite a few citizens seem to reply to these authorities by printing out the documents they receive, adding handwritten comments on them, taking pictures of all the pages using their smart phone, and emailing them to the authorities. That is how an outgoing black and white document is transformed into an incoming document in full colour - and full size.
While packing and storing we were also chasing bit errors on our storage media – optical BD-R and magnetic LTO-6 - by running integrity checks. None were found on the optical media using random checks, but a few on the tapes using full checks! The bit errors grew into byte errors, nothing more, but still we were alarmed, and checked that the other two copies were intact, and began recreating the third copy. Once again the culprit seems to be software, i.e. drivers and firmware that fail under very rare circumstances, not the hardware and the media itself. Having more than 100 TB of digital archives bit errors are bound to happen, but we get very uneasy when they show up almost every decade.
A thing that shows up a least every decade is a national strategy for digital archiving. Actually, this Thursday my colleague Phillip and I had planned working solely with our next national strategy for digital archiving (and the detailed plans for its implementation) as well as our international, well at least European, DLM Digital Archival Standards. But sadly, as the day progressed, both problems and performed task became more and more local, and we did not get much work done on neither content type SIARD nor the DIP specification. Phillip did manage to do a little dissemination at https://digitalbevaring.dk/om-sitet/about-us/ both otherwise the world around us was not much disturbed by our presence.
That very same evening the Archivist Union (these days solely functioning as a society) celebrated its centennial anniversary with the publication of the book ”Documentation in digital times, DCDM, archives and proper public administration practice” - along with a glass of champagne.
Sadly, having to relieve our families at home Phillip and I could not attend, we will simply have to wait for the bicentennial anniversary, if it was not for - archives persist, archivist perish.