there’s plenty of tech thinkpiecing right now which is like “if a web scraper uses libxml2 to process web pages does that mean that contributing to libxml2 is being complicit in web scraping” and i don’t buy that, i think (1) the people who are complicit in the web scraping are the people providing the physical infrastructure (machines, wires, electricity) required to carry it out, and (2) libxml2 isn’t even that good, and the fact that the web scrapers can Just Use It instead of writing their own XML parser might save them a little bit of time and intellectual work in the short term, but in the long term they would just write their own XML parser, it’s not like nobody knows how to do that, secrecy regarding the methods and computer code is not a meaningful impediment

programmers seem very committed in the present moment to thinking of code as concrete and material but infrastructure as abstract and ephemeral and i would like to suggest that maybe that is exactly backwards

but i still would not contribute to rails

@Lady On one hand, I get it. On the other hand, you use Rails, and I use Rails, and I think we're pretty okay?

@aschmitz i wish all the pretty okay people who used rails would get together and fork rails

i know this won't happen but i wish it

@Lady genuinely curious. What's wrong with Rails?

@wallhackio two things:

• guy who makes rails (dhh) is a techbro with terrible opinions about a lot of things, including (imo) technical decisions regarding the direction of the platform

• rails is really good at enabling small teams to build a minimally viable product quickly, which is why i work with it daily and why it is everywhere in library technology (lots of small underresourced teams in library tech). this is, more or less, good. however it also means it is often the framework of choice for ethically dubious startups trying to get some quick investor cash, which is not good. i want my code to benefit other libraries, but i don't want to do free work for the ethically-dubious startups

@wallhackio speak of the devil and he shall appear: https://hachyderm.io/@pat/112216636273524747

@Lady I dunno, I don't buy that code isn't concrete and material also.

Like similarly, in what way is a curated database not infrastructure?

@gaditb i think there is a categorical difference between a set of instructions on a page and the actual enactment of those instructions. i view “infrastructure” as perhaps everything EXCEPT the instructions themselves which allows the execution to take place; i think lumping the instructions (an abstract product of human knowledge) in with all the things which actually are processing and doing and performing the execution is a mistake. but to be clear, this is referring to computer code as intellectual product, not computer code as bits on a disk. likewise, the information in a database is not infrastructure; the actual physical database that you can query and perform tasks on is

this distinction is necessary because the intellectual property of computer code (or database information) is not the only requirement to having an actual usable infrastructural piece. for databases, you need to have a computer which can store database entries, for one. then you need to actually load the database entries into the database. you need to connect it in such a way that it can be queried. only after you have done these tangible steps do you have a database-as-infrastructure. just the information, without performing those steps, is no more infrastructure than a cookbook is grocery

or, you might need recipes to make good food, but you also need ingredients. you need code to execute tasks on a computer, but you also need something to load the code onto and something to execute it. when i say “infrastructure”, i mean the latter items, because code which is not loaded onto anything or executable by anything is not actually a usable computer program. or, for that matter, meaningfully distinguishable from any other output of human intellectual pursuit, e.g. books

(if your argument is that all human knowledge is infrastructure i feel like you are severely weakening the utility of thw term)

@Lady My argument isn't that all human knowledge is infrastructure, but against the idea that it categorically CANNOT be infrastructure. I don't think being an abstraction prevents something from being infrastructure (I also think that there's a lot more abstraction at varying levels than is useful to categorize "abstract vs. concrete" in universal-across-all-scopes terms).

I think "infrastructure" is a relationship of a particular kind of dependency, not an isolated property.

And so, a piece of human knowledge CAN be in the "being depended on" role in that type of relationship. (Or, indeed, a human -- someone with a particular role or set of skills or who takes on a particular task for the community.)