Skip to content

Instantly share code, notes, and snippets.

Michal Migurski migurski

Block or report user

Report or block migurski

Hide content and notifications from this user.

Learn more about blocking users

Contact Support about this user’s behavior.

Learn more about reporting abuse

Report abuse
View GitHub Profile
kylemcdonald /
Last active Aug 29, 2015
The other reasons F.A.T. Lab died.

Today F.A.T. Lab shut its doors with this lovely note from Magnus Eriksson and Evan Roth.

But the "we lost to them" narrative is only one perspective. As a "virtual research fellow" since 2011, I had the opportunity to both look up to F.A.T. as an outsider for its first half, and get the inside scoop for the second half. But my first project appears on page 20 of the blog's 200+ pages. With that in mind, here are some other stories about why F.A.T. "lost":

  1. F.A.T. lost its rowdy juvenile edge when almost everyone got married, had kids, got (mostly) real jobs. The biggest producers from early F.A.T. effectively "retired" from producing F.A.T.-style work.
  2. There was not enough new energy to replace these retired members. To fix this, sometimes a member would propose adding someone. Inevitably, someone else would suggest they "weren't F.A.T. enough", or just thinking about how many inactive members we had, and how much of a boys club it was, would get us down and discussion woul
View Fiasco's suggestions (started by Marc)
View gist:9591491

Reciprocal Needs in the Employment Relation

We can look at two sides of the management coin: What do the individuals get out of it? And what benefit does the whole system derive from it?

I will disregard any benefits that accrue to managers just by holding the position of managing. Those are just circular logic. Circular logic abounds in discussions of management and hierarchy. For example, consider status reports. It will be said that status reports are necessary so managers know what their employees are working on. It’s

bensheldon /
Last active Dec 29, 2015
A summary of conversations I've had with current, former and future Code for America fellows.
  • Who's your customer? The department or the citizen. What’s your product? The negotiated-design process or the actual outcome itself. These things are interrelated, but ensure the entire team is aligned and in agreement on what it is you’re doing and what the outcome is you want. What do you want to say you built at the end-of-year summit? The earlier you can figure this out, the better your year will go. Project uncertainty is the mind-killer.

  • Don't negotiate with yourselves. My team would have achieved so much more if we had come to an agreement internally on how to proceed in our city design negotiations, and let CfA staff tell us if we were unreasonable. Instead, we watered down what we wanted, caved on major issues, and negotiated individually with our city partners rather than as a team.

  • Invest your time in one thing for the entire year. Your understanding of the civic problem-space may change and evolve, but creating a commitment (individually, as a team, and in your city-

FZambia /
Last active Apr 27, 2019
python in process memory cache based on ordered dictionary - with size limit and timeout support
# coding: utf-8
import time
from collections import OrderedDict
class Cache():
In process memory cache. Not thread safe.
jboner / latency.txt
Last active Feb 19, 2020
Latency Numbers Every Programmer Should Know
View latency.txt
Latency Comparison Numbers (~2012)
L1 cache reference 0.5 ns
Branch mispredict 5 ns
L2 cache reference 7 ns 14x L1 cache
Mutex lock/unlock 25 ns
Main memory reference 100 ns 20x L2 cache, 200x L1 cache
Compress 1K bytes with Zippy 3,000 ns 3 us
Send 1K bytes over 1 Gbps network 10,000 ns 10 us
Read 4K randomly from SSD* 150,000 ns 150 us ~1GB/sec SSD
MoritzStefaner / .block
Last active Jan 20, 2020
Force-based label placement
View .block
license: apache-2.0
You can’t perform that action at this time.