keredson

I mean, if it worked for CVS... =D

and mandates that virtually all new projects not directly use borg/kubernetes.

you are right. made worse because there's an internal rule that you can't ever have two versions of a library in the monorepo ant the same time. it has to be one giant CL for the entire migration company-wide.

OS updates are important sometimes. Security and all...

or you have crappy code that can only handle a dozen RPS. [facepalm]

thanks! yeah, lower-right corner. i added a walk-through that highlights it to help. =D

Yep. There is more than one terrible way to run a business.

KR stands for "Key Result". They're definitely not "things you do to achieve your goal".

That's kind of the point of OKRs (in theory): To stop upper managers from dictating process to lower managers. (Which is a big problem.) Instead, ignore their process and evaluate them on how well whatever process they chose delivered. IE, give someone an objective, let them find their own solution, and evaluate them on only how well it worked.

OKRs in practice look suspiciously like waterfall software development. When you find yourself assigning ICs programming tasks by economic quarter, you've f-ed up.

That's in theory. In practice I knew of no ICs w/o OKRs assigned to them.

yep! https://en.wikipedia.org/wiki/Goodhart%27s_law

"Management [...] wants something that works and delivers economic value for customers and will help them continue to innovate and deliver EVA for them in the future."

You're confusing "owner" with "manager". Most managers just want to get promoted, which usually means increasing their HC/budget, making sure credit for a project's success is their's, failures are blamed on others, getting trendy internal projects, doing endless re-orgs (knowing they'll move on before any true costs/benefits are realized), etc. "Economic value for customers" is an afterthought at best.

(Speaking as a former very-large-company manager.)

looks like there's an impl already: https://github.com/cbschuld/uuid-base58

if you have a cryptographically secure random function, just use that. no need to hash it!

100%. also, why re-invent python's already pretty decent datetime/timedeltas?

and why not a proper python DSL?

  from redengine import minute, hour
  
  @app.run_every(hour + 20*minute)
  def do_first(): [...]
  
  @app.run_after(do_first)
  def do_second1(): [...]
  
  @app.run_after(do_first)
  def do_second2(): [...]
  
  @app.run_after(do_second1 and do_second2)
  def do_last(): [...]

that's very common in distributed databases. even traditional databases, it's very common to not have FKs on large tables, and just handle it in software. indexing billions or more of rows is non-trivial.

SSE supports gzip compression, and a gzip-ed base64 is almost as small as the original jpg:

$ ls -l PXL_20210926_231226615.*

-rw-rw-r-- 1 derek derek 8322217 Feb 12 09:20 PXL_20210926_231226615.base64

-rw-rw-r-- 1 derek derek 6296892 Feb 12 09:21 PXL_20210926_231226615.base64.gz

-rw-rw-r-- 1 derek derek 6160600 Oct 3 15:31 PXL_20210926_231226615.jpg

nice! i did similar, but used character frequencies in the remaining word sets to rank: https://github.com/keredson/wordle_solver

This was an afternoon project (when I got tired of the built-in animations of my new case LEDs). Thought others might enjoy. =)

That can be done in a non-sleazy way. For a long while there were effective caps on what you could earn as an IC, no matter how good. Everyone talks about 10x devs, but a decade ago the only real way to capitalize on it was to hold multiple remote gigs, each where you're accurately emulating the performance of say a $150-200k dev. 2-3 of those is very do-able, and a darn nice income if you're outside the major tech hubs.

As long as you're performing akin to your salary for each position, and you read your employment contracts carefully, I see nothing wrong w/ this approach.

Ouch, I'm sorry. When I get home tonight I'll check to see if GitHub supports loading=lazy.

I hadn't even thought of using small servos. Possibly that would have been a lot simpler! I went though a bunch of prototype EMs testing to get a working combo of voltage, windings, core size, etc. (Blew more than a few driver circuits too...)

Even more crazy were these "iron core for electromagnets" https://www.amazon.com/Soft-Ideal-making-electromagnets-inch..., 6" for $17.

Give me a break. Soft steel is >99% iron and I bought 4' for <1/2 that price from HD. Rip-off.

Hah, yeah, I think I saw them on Amazon for like $20. That seemed nuts; why I decided to roll my own.

On the dowel there are 5 steel nuts which are pulled towards the electromagnet. Gravity makes them return, but TBH it's not reliable enough. I'm working on a new version with small return springs and a wider swing for the striker.

The electromagnet is pulsed to control how hard to strike.

Thank you!

There's a little plastic hinge that goes on the end of the dowel / slots into the hanger. I don't have a picture, but the file is here: https://github.com/keredson/chimes/blob/main/striker_hinge.s...