3 ms·

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

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.

bobobob420 4y ago
Can you elaborate more on the architecture of your single query system. Is everything in one source or are you searching multiple sources with one query
- Arainach 4y ago
  Searching multiple sources with one query. It's really just an internal search engine.
cormacrelf 4y ago
The question about the library isn’t really answered by “we have an internal wiki”. It was asking “Do you devote company resources to research, or do you think you already know everything necessary?” Internal docs/wikis are the expression of what your company already knows. We’re talking about recognising what it doesn’t know. The instinct to visit a library and read a book about something you don’t know about is not something everyone has from birth. A lot of people out there are pretending to understand a lot of stuff. You would only have a corporate library if the leaders understood anything well enough to know that sometimes you do need a book to go any deeper. So the question is also asking you “Does your company respect real depth of knowledge or does pretending suffice?” All that isn’t a comment on your company, which as you say, has a library. But I don’t want people to come away with the impression that any old “store of knowledge” qualifies as a library.
- Arainach 4y ago
  Almost no one is doing research these days. Not most companies, not most profitable companies, not most profitable teams on profitable companies. Most of business boils down to taking existing pieces and putting them together in the way that delivers the most customer value, or putting them together in an adequate manner and hiring a sales team to convince customers they deliver a lot of value. That's it. To be clear, I don't mean "no one is doing research and that's a shame". I mean "no one is doing research because there's no reason for them to". Telling a developer at nearly any company to go read a research paper on databases is silly; just go grab a database written by someone who specializes in that and use it. Reading about specialized 3D rendering techniques isn't going to save you anything that just using Unity/Unreal wouldn't. And so on and so on. It is dramatically more valuable to be able to answer "what are the available tree data structures in this language in our codebase" than to be able to answer "what's the algorithm to correctly implement my own red/black tree". If research is being done at your company it's probably being done by a handful (literally, fewer than 5) of people who know what they need to find and what they're doing. For most other folks it's a mistake. I saw multiple instances at large tech companies where some new college hire sat down, wrote a beautiful data structure that was algorithmically fine and well-researched.....and was immediately rejected and thrown away because we had equivalents already that they hadn't looked for or hadn't been able to find and no one wants to maintain multiple. Even at companies with real libraries (Microsoft etc.) the folks using the libraries weren't using them for research, they were using them for standard technical reading material.
flir 4y ago
When you said "virtue signalling" my mind immediately went to Knuth. You know, my copies are jammed so tight in the slip case I don't even think they come out any more.
- disgruntledphd2 4y ago
  Even just Volume 1 is amazing (as long as you use the new assembly language he wrote). I made it through this after I got laid off, and it was super helpful. I've only flicked through the other volumes though, to be fair.
tablespoon 4y ago
> 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). My company essentially deletes all that. They migrate from intranet portal to intranet portal periodically, and each migration is advertised as an opportunity to "clean up" all that documentation that hasn't been recently updated about legacy systems that we still need to run. The people who run the portal are technically in HR or something like that, and I think they have know idea how it's actually used, and mentally model it as a place for announcements.
somedangedname 4y ago
> R.V. Jones' Most Secret War Added to my reading list. Have you read Between Silk and Cyanide by Leo Marks? It's quite good.

bobobob420 4y ago

Can you elaborate more on the architecture of your single query system. Is everything in one source or are you searching multiple sources with one query

Arainach 4y ago
Searching multiple sources with one query. It's really just an internal search engine.

cormacrelf 4y ago

The question about the library isn’t really answered by “we have an internal wiki”. It was asking “Do you devote company resources to research, or do you think you already know everything necessary?” Internal docs/wikis are the expression of what your company already knows. We’re talking about recognising what it doesn’t know. The instinct to visit a library and read a book about something you don’t know about is not something everyone has from birth. A lot of people out there are pretending to understand a lot of stuff. You would only have a corporate library if the leaders understood anything well enough to know that sometimes you do need a book to go any deeper. So the question is also asking you “Does your company respect real depth of knowledge or does pretending suffice?” All that isn’t a comment on your company, which as you say, has a library. But I don’t want people to come away with the impression that any old “store of knowledge” qualifies as a library.

Arainach 4y ago
Almost no one is doing research these days. Not most companies, not most profitable companies, not most profitable teams on profitable companies. Most of business boils down to taking existing pieces and putting them together in the way that delivers the most customer value, or putting them together in an adequate manner and hiring a sales team to convince customers they deliver a lot of value. That's it. To be clear, I don't mean "no one is doing research and that's a shame". I mean "no one is doing research because there's no reason for them to". Telling a developer at nearly any company to go read a research paper on databases is silly; just go grab a database written by someone who specializes in that and use it. Reading about specialized 3D rendering techniques isn't going to save you anything that just using Unity/Unreal wouldn't. And so on and so on. It is dramatically more valuable to be able to answer "what are the available tree data structures in this language in our codebase" than to be able to answer "what's the algorithm to correctly implement my own red/black tree". If research is being done at your company it's probably being done by a handful (literally, fewer than 5) of people who know what they need to find and what they're doing. For most other folks it's a mistake. I saw multiple instances at large tech companies where some new college hire sat down, wrote a beautiful data structure that was algorithmically fine and well-researched.....and was immediately rejected and thrown away because we had equivalents already that they hadn't looked for or hadn't been able to find and no one wants to maintain multiple. Even at companies with real libraries (Microsoft etc.) the folks using the libraries weren't using them for research, they were using them for standard technical reading material.

flir 4y ago

When you said "virtue signalling" my mind immediately went to Knuth. You know, my copies are jammed so tight in the slip case I don't even think they come out any more.

disgruntledphd2 4y ago
Even just Volume 1 is amazing (as long as you use the new assembly language he wrote). I made it through this after I got laid off, and it was super helpful. I've only flicked through the other volumes though, to be fair.

tablespoon 4y ago

> 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). My company essentially deletes all that. They migrate from intranet portal to intranet portal periodically, and each migration is advertised as an opportunity to "clean up" all that documentation that hasn't been recently updated about legacy systems that we still need to run. The people who run the portal are technically in HR or something like that, and I think they have know idea how it's actually used, and mentally model it as a place for announcements.

somedangedname 4y ago

> R.V. Jones' Most Secret War Added to my reading list. Have you read Between Silk and Cyanide by Leo Marks? It's quite good.