Yes and no. I feel that not doing certain things is a very simple thing to "do" things - and I appreciate simplicity.
For example, I would probably prefer getting rid of half of the cars and trying to find ways so that people do not need to move that much, if they do not want to (for example making all possibly remote remote work actually remote) - as opposed to making all cars electric.
I do not own a car, I do not fly, I buy clothes used or new just once a year. My newest smartphone I bought five years ago (and I just bought a new battery for $10, so I can use it for five more years). My utility company stats says my energy consumption is overall very low. I eat meat once a month. I try to own as few things as possible. I try to write and deploy code in energy efficient languages (Go, Rust, C, ...) [1]. I self host some services on my arm board that consumes 0.79W when idle.
I do not feel, that it makes a large dent, but I feel that I'm way ahead what the average person does in order to reduce their environmental footprint.
Also, I do not feel I'm cutting myself short and I still depend on many of the niceties of modern civilization - but, I'd also be happier, if my footprint would be even lower.
> Was it always this bad and we just didn't see it before?
After about two hours of using FB in 2007 I had this crazy feeling that someone having access to all those people's photos and activities won't necessary end well; left and never looked back.
Question: How hard would it be to write create open source components to run a DSP collaboratively, e.g. with the power asymmetry created by aggregation
I support sourcehut, because I would like to see it evolve, and hopefully keep the no-nonsense UI approach. Having a nontrivial site with a usable interface running without JS in 2019 is beautiful on its own.
I managed complex datasets of a few TB with standard Unix tools just fine. If you are lucky, you have a sustained read and processing speed of 200M/s (or more with, if you are put in some work), which allows you to sift through a TB in a good hour.
Sometimes I am wondering, how people will rationalize the layer bloat in 5-10 years, when hardware went through a few more upgrades.
I once taught[1] the building blocks of git (basically bullet time view of a commit) and people found it a bit too theoretical - even though it contained all the elements, that helped me to understand and appreciate the simplicity.
There is a point, where you go from memorization (add, push, commit) to deduction (graph, objects and refs) but when this point is reached, depends on many individual factors.
> Maybe I am just naive, but this all seems terrible.
I do have most respect for TBH and I would consider everything he thinks and writes about, but this does not sound too good to me either.
The idea of linked data and semantic web has been around for almost two decades now and I have yet to see an application, technique or site that amazes me. On the contrary, most of the things in this space I have seen are bloated, unusable or simply unnecessary - whereas every paper sounds like revolution is around the corner. In that combination, it is the worst of both worlds: academic output, that claims practicality and fails to deliver.
Peter Norvig put it best, when he said: "The semantic web is the future of the web and always will be."
> What really freaks me out is the day Facebook die, what will happen to all of this data?
Interestingly, Facebook owns your data. I believe if they wanted to, they could close the company tomorrow and put a facebook.tar.xz of everything they collected on archive.org or somewhere else.
Luigi is a great Python library for building multi-stage pipelines with parameterized tasks. It easily extend to new storage types and targets.
We currently use it to build a moderately complex product from dozens of data streams (files, APIs and things in-between) and millions of records. At its core is a DAG and topological sort, which capture the essence of pipelines and execution.
A DAG is easy to visualize, even its development over time[1].
If it's not overdone, an animated walk-through can give you an impression of how simple something is - here a short animation can be worth a thousand words.
For more complex topics I find animations less helpful.
Yes, I used `load_data` and friends and it is certainly convenient. I was just curious about a kind of open source market place for models, where you could find niche models that people train in their spare time.