I try to develop primitives for data structures and/or algorithms - starting with e.g a doubly linked list node / binary tree node… moving on to e.g. breadth first search implementation, etc.
Being older, for me this has the benefit of refreshing some compsci fundamentals while exploring the new language.
I have a largish white board (thanks home office) on which I summarize a (sub)section of a book / online course in my own handwriting / symbols / diagrams after reading / watching it first and doing the summarizing on a second run (after which I take a photo and save it to a doc on the topic).
This method has helped me commit quite a lot of conceptual information into long term memory - but - it is time consuming.
If I ever need to refresh my memory on a topic, I pull up the photo and redraw it - sometimes adding or changing things a bit - then take photo again. Rinse and repeat.
You are assuming edible plants to have been readily available at high enough concentrations at the hunter gatherer stage of human evolution.
It is energetically more expensive to roam a large area to gather x calories than finding and killing an animal (herbivore) that did that for you - never mind the ability of the herbivore to release calories contained in leaves and grass (something humans cannot do)
Then in winter this ratio increases even more.
I don’t know if you’ve ever been to the African savannah, but early human evolution here was definitely not driven by the abundance of edible vegetation.
Interesting to see the relatively small percentage of PUT requests.
A (generously opinionated) observation of the technical debt / re-invented wheels in the wild?
For example a gateway / proxy can much more efficiently route a PUT payload to the required upstream (with “INSERT” permission and tuning) than deciding by analysing the payload, or even worse, not having to because the entire stack just uses POST as some franken analogue to “UPSERT”
Goes a bit deeper. Can you replace a source of data with another in a few lines (besides the new implementation) and not break a whole chain of dependencies and associates unit tests? (apart from "does someone knows what's going on") - i.e. code at scale.
Agreed. Good application code often contains edge case handling, build time checks, unit tests and defensive flows that handle the unexpected so that users don't wake you up at night. Why can Ops not do the same? Why can Dockerfiles / Orchestrators / CI / playbooks not also implement sanity checks on deployments?
"Ooops... deployment failed. While deploying your artifact we found the following:
- Nothing is listening on the nominated port
- Your deployment is utilizing 100% CPU while idling
- We detected an abnormal volume of write operations to the mount
Please fix these issues and re-trigger the pipeline at your earliest convenience.
Considering global wealth distribution and the "1%", how often do 95% of people make a EUR10k purchase and of those, how many prefer the safety and guarantees offered by conducting the transaction through a financial institution?
This article is biased, opinionated and littered with unsubstantiated facts, which I guess makes it an opinion piece.
The facts are that the UK has almost no manufacturing base, lags behind even smaller EU countries in various modern intellectual property endeavors, and greatly benefits from the free flow of goods and services from the continent.
The author is correct in pointing out that Brexit won't be the disaster everyone makes it out to be, but it will cause a systemic and gradual decline in the GDP and living standards of the British people - and cosying up the US will come with its own set of problems.
The author assumes that software (or "code" as the article puts it) is a "problem" that has been - or is close to being - solved, and so can reliably be abstracted into an easy to use and understand UX.
However, software, the algorithms they encode, and the applications with relation to the hardware they run on is constantly evolving. By the time you have a reliable drag and drop interface for a use case, things have moved on.
Thus, these "no code" approaches will always lag considerably behind what is required to remain competitive, resulting in the requirement for the "DSLR" crowd.