For those interested in a lua example, from this article [1], I found & cherry-picked bits from here [2]. The key ingredients (to aid searching) are init_worker_by_lua_block & balancer_by_lua_block.

[1] https://medium.com/@sigil66/dynamic-nginx-upstreams-from-con...

[2] https://github.com/sigil66/nginx-consul-cookbook

VPN companies don't care too much about customers that are concerned about privacy - too small a percentage of their customer base. Catering to those trying access geo-blocked content is where the money's at.

Yes? I think we get more done in the short-to-mid-term but the fruits of our labor are less likely to survive long-term. There's no value judgement there - there are pro's & con's. At 40 years around the sun, I've got one leg in the "Get off my lawn!" and the other leg in the "Damn, these kids are wicked smart!" camp. I find that people over-focus on the seemingly "attention deficit" critiques of modern day, but even if that is the case, I think it's a small price to pay for having orders of magnitude more people in the space - there will be a lot of failures but there will be more successes relative to prior generations.

If we want to blame younger generations for not "thinking deeper" or "solving the hard problems", then we should shift our focus towards the economic drivers that are favoring quick/quantity over quality. I know how to develop software but if an employer is given the choice of fixing a buggy tool at the cost of maintaining a fork (GASP!) vs spending unbounded man-hours working around said buggy tool, my money's on that they will choose the latter since maintaining tool X is not a core competency. Meanwhile, tool X is hemoraging production data or producing buggy results and nobody flinches.

For that flavor, maybe try:

    python3 -m pip install -t .pip ...
    export PYTHONPATH=".pip:$PYTHONPATH"

Use whatever directory name you prefer instead of .pip. Mucking with PYTHONPATH is a bit dirty.

ZK 3.5 brings this.

I tell people to `ipython; %magic`. For all of python's faults, that REPL is the gold standard and makes for a nice way to learn the language.

All the yak shaving around: room history, notifications, security.

Also, doesn't the +/- sigma rule only apply to a normal (gaussian) distributions?

Yes, that rule:

   31%  outside +/- 1 standard deviation  ("2 sigma"), 
   5%   outside +/- 2 standard deviations ("4 sigma"),
   0.3% outside +/- 3 standard deviations ("6 sigma"), 

etc. However, the more general Chebyshev inequality states that for any distribution, you have at most 1/k^2 outside +/- k standard deviations, so (at most):

   100% outside +/- 1 standard deviation, 
    25% outside +/- 2 standard deviations,
     9% outside +/- 3 standard deviations, 

etc.

https://en.wikipedia.org/wiki/Chebyshev%27s_inequality

Are there any tools or sites where crowd-sourced dictionaries is a key feature? As a developer of an app, I would like to have basic actions (eg open, save, login, cancel, ok) translated to all the locales.

I think context/scale/application is important. Classical Newtonian physics takes us pretty far pretty accurately until we get to more micro or macro scales. For example, if wanting to model a physics engine in software and the Newtonian models are considerably more performant, then could make sense to sacrifice the accuracy.