Why Card Catalogs Matter (2017)
smithsonianmag.com
smithsonianmag.com
Not long after this, the library moved to a brand new building. The card catalog did not make the move to the new building. I may have been the last person to have updated it.
What is the excuse? "The company that printed the cards went out of business" I bet there is a lot of companies that could print a batch of such cards upon order. How hard would it be to set up a printer that filled and updates the cards and the it system to be in sync?
I remember using libraries without computerised catalogs for a long time. When I used electronic search for the first time I thought it is huge. I could find stuff in minutes instead of hours (or asking a librarian to do the search for me). However, it t is extremely short sighted to just discard or let these old catalogs to rot.
We're seeing more and more problems arise because of the lack of a failover plan. All it would take is a small printer to generate a new car from the digital system whenever a new item is added to a library's collection, and for someone to file the card just like they do the book (or do they still have physical books in libraries? ).
This is such a trivial problem to solve. In fact, libraries are really good for distributed storage so have a peer to peer replica all over the world.
Libraries are trustworthy and stable so good spots to distribute 100 durable replicas of the catalog data.
If you’re worried about surviving Armageddon or something then that’s a different problem. Not that the books will likely be around, but I suppose you could dump the data out to records or something and bury it under the library and that’s more likely to survive whatever takes out the library.
But I think a digital copy of the catalog distributed around the world and into space is more likely to survive.
That might be a fun project, to broadcast the data out into space at something that will reflect it back in 10-100 years so we’ll have a copy coming back to us in the future. Is that possible, astronomically?
Backups are not a failsafe.
Firstly, backups are not an emergency plan. Restoring backups is the emergency plan, and that needs to be tested and verified on a regular basis. Making backups without validated restoration is just theatre,
Second, a backup is not a failsafe. A failsafe is something that lets you keep working in the event of a failure. It's not something that lets you keep working after you recover. If a self-driving car had the failsafe of stopping dead in the middle of the railway tracks until you restore it from backup you're going to have a lot of damaged trains.
No amount of computer database backup will help you when the power goes out. Fires, floods, tornadoes, hurricanes, earthquakes, they happen and power can go out for weeks (in the first world) or longer (in most of the world). With today's connected world, all it takes is for a network switch to go out and hundreds of thousands of commuters get stuck in the downtown during rush hour because the commuter trains have no failsafe while freight trains keep trundling on the same tracks because they do [0].
Making backups is not a failsafe and it's not good enough.
[0] ttps://www.cbc.ca/news/canada/toronto/go-transit-system-problem-1.6985282
However, printed catalog cards as a backup to the digital catalog have practical problems. For one thing, they are heavy. A full card catalog of a large public library could easily weigh tons [2] and take up space that would be better used for other functions. For another thing, the expense would detract from other library services, and many libraries already suffer from budgetary challenges.
A proper backup strategy is the way with the risk of losing the digital catalog. Besides, libraries have more fundamental problems to deal with regarding the permanence of electronic resources. For example, the fact that libraries largely do not and cannot _own_ the ebooks they circulate causes numerous problems.
[1] https://www.allpurposeguru.com/2021/03/henriette-avram-and-t... [2] https://blogs.loc.gov/preservation/2022/10/moving-card-catal...
"Why does a new index cost disk space?" "It's a whole new wooden cabinet"
"Why is re-indexing expensive?" "Look at all those cards! Each one needs to be regenerated"
And so on.
Although this was before the internet and the Internet, there was a large, single purpose, DOS-based computer with a searchable encyclopedia and a Library of Congress catalog.
Dewey Decimal Classification represent.
Like a totem, a card catalog isn’t really needed anymore but it represents something and may as well have something like that in the “real world” as the interface to the actual info.
I used to wear a mechanical watch not because it told time but because it was a physical thing that represented complexity and architecture and time. etc etc
As a kid, I read these books by Gordon R Dickson, the Final Encyclopedia. And the it was a giant place that had all the information in the universe and an AI to sort and deliver it. But it was neat because it was a place, like a temple. This was before the internet so it really could be distributed, but the idea of having limited interfaces stuck with me as a kid.
Speaking of card catalogs, every time this topic comes up, it reminds me of a book I really enjoyed about index cards.
> Why the card catalog—a “paper machine” with rearrangeable elements—can be regarded as a precursor of the computer.
Paper Machines: About Cards and Catalogs, 1548-1929 - https://web.archive.org/web/20220531211447/https://mitpress....
Yes, computer search is better in many ways, but I miss the tactile experience of riffing through the cards and the minor archeology of finding "ancient" cards of books that had been in the collections for decades amid crisp new cards for recent releases.
He’s convinced he’ll eventually find a surplused one for nothing. But I think many others feel the same.