It depends on how abstract you consider a "library". My employer does have a library of physical books, but they're almost never used. On the other hand they have an enormous amount of
digital internal data - Wikis, API docs, design docs, Stack Overflow-esque Q&A, and more - that's well indexed and can be collectively searched via one query. Those assets are used constantly every day by everyone (and even moreso if you consider code search part of your library).
Books have a higher cost to entry - you have to buy them, you have to store them, you have to get out of your chair and go get them, and you have to find the relevant data point in them. They're not suited to topics where the answer or the state of the art changes every few months.
I have a library in my home office, but there's nothing in it that I consult every week or even every month. Broadly, my library can be broken down into a few themes:
1. Books that I found valuable and lend out often (but don't necessarily need to reference) - things like "The Design of Everyday Things", "Working Effectively With Legacy Code", "Staff Engineer", "Peopleware", etc.
2. Books that contain information that is not easily and immediately available in an internet search (most often non-technical history books)
3. Subjects I reference on occasion but need detailed information from (Skiena - Algorithm Design Manual, CLRS, etc.)
4. Things that bring me joy to reread or even just look at and reminisce about reading (this is for anyone a personal list - for me it's things such as Ignition!, R.V. Jones' Most Secret War, the works of Neal Stephenson and Robert Caro, and more)
There are of course exceptions, such as the unspoken category 5 (virtue signaling) - I'm probably never actually going to read those Knuth tomes or even finish that Dostoyevsky, but they look impressive on the shelf behind me in video calls - but over time I've gotten better about eliminating those.
Books have a higher cost to entry - you have to buy them, you have to store them, you have to get out of your chair and go get them, and you have to find the relevant data point in them. They're not suited to topics where the answer or the state of the art changes every few months.
I have a library in my home office, but there's nothing in it that I consult every week or even every month. Broadly, my library can be broken down into a few themes:
1. Books that I found valuable and lend out often (but don't necessarily need to reference) - things like "The Design of Everyday Things", "Working Effectively With Legacy Code", "Staff Engineer", "Peopleware", etc.
2. Books that contain information that is not easily and immediately available in an internet search (most often non-technical history books)
3. Subjects I reference on occasion but need detailed information from (Skiena - Algorithm Design Manual, CLRS, etc.)
4. Things that bring me joy to reread or even just look at and reminisce about reading (this is for anyone a personal list - for me it's things such as Ignition!, R.V. Jones' Most Secret War, the works of Neal Stephenson and Robert Caro, and more)
There are of course exceptions, such as the unspoken category 5 (virtue signaling) - I'm probably never actually going to read those Knuth tomes or even finish that Dostoyevsky, but they look impressive on the shelf behind me in video calls - but over time I've gotten better about eliminating those.