Labnotes

Weekend Reading — The second page of Google search

Published on

image

On the Hunt for Japan’s Elaborate, Colorful Manhole Covers Street art.


Design Objective

How to simplify your design 👍 This article itself is well designed, with 21 illustrated examples and just the right amount of text.

1-LDHS0cnrBa78ruUoUjcx5Q

Rules for Autocomplete If you're doing any work on autocomplete UI, this is a fantastic list:

  • Exact matches always come first. If the user types in an option exactly, other options must always go below the one matching what they typed.
  • Besides exact matches, prefix matches come first. If I type “Fr” I want “Fresno” not “San Francisco.”
  • After prefix matches, it can fall back to substring matches. Starting with substring matches would almost always be the wrong thing to do since users start typing words at the beginning not somewhere in the middle.

Speak Human "Generate human centric microcopy for all purposes."

Screen-Shot-2019-03-30-at-6.59.49-PM

Dawnstar Australis 🤔

I believe I just overheard a software bug being described as an "unexpected user story"


Tools of the Trade

Announcing Lucet: Fastly’s native WebAssembly compiler and runtime The future of serverless is instant startup, and running code in the same metro area as the user:

Lucet is designed to take WebAssembly beyond the browser, and build a platform for faster, safer execution on Fastly’s edge cloud. WebAssembly is already supported by many languages including Rust, TypeScript, C, and C++, and many more have WebAssembly support in development.

Lucet can instantiate WebAssembly modules in under 50 microseconds, with just a few kilobytes of memory overhead. By comparison, Chromium’s V8 engine takes about 5 milliseconds, and tens of megabytes of memory overhead, to instantiate JavaScript or WebAssembly programs.

Karl Stratos Well, that explains it …

linear_dogs

Request’s Past, Present and Future Good bye request. You did great for Node developers everywhere. 👋

Jeff Forcier I feel ya:

Also: examining old image files marked as 'wallpapers' which are almost literally postage stamp sized on a modern (not-even-HiDPI!) display is, uh.

It makes a body feel old, is what.

clean slate True, true:

[inserting row in excel]
Excel: copy font format from the row above?
Me: no I’ll handle it
Excel: and copy border from below?
Me: no why?
Excel: idk :/
Me: *typing number* w-why did you make 31,320 a date?
Excel: it’s my birthday :)

Samsung just unveiled the widest computer monitor you can buy Apropos Excel, you can now view spreadsheets thiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiis wide!

D2w_U8dXQAAPWMQ


Web-end

Nicolas Goutay 👇 Web Performance metrics explained in a single tweet each. Good, quick reference.

amy nguyen "this whole interaction is why i am still on twitter"

D2TK5KVU8AEp8mX


Lines of Code

Sindre Sorhus I support this linting rule:

The prevent-abbreviations rule encourages you to write variable names like error instead of err and event instead of e. Of course, all of this is configurable. You can even use it as a general variable name blocker/replaces. Default replacements: https://github.com/sindresorhus/eslint-plugin-unicorn/blob/2f8f3d38e1c8058ac9bacb8bc318a66cda006307/rules/prevent-abbreviations.js#L17-L172

In the end, the biggest time efficiency happens between the ears:

The "writing" part of the code should not be the constraint. I personally spend much more time thinking about code than writing.

Jason Karns 😭 (the responses are funny as well)

Bought a mechanical keyboard with more resistance so my code will be strongly typed.

Sebastian McKenzie I'm going on 30 years …

This has been my debugging strategy for the past 10 years and I don't see it changing

D2Y_0H-UgAAQ1Lv

Meowlivia_ How to ace a job interview:

Interviewer: Tell me about a time when you had to solve a difficult problem

Me: I had a bug in my code and couldn't find any helpful answers on stackoverflow. I persevered and eventually found the solution from a link on the second page of Google search


Architectural

Kent Beck 👍 I couldn't agree more:

If there’s one lesson I would like the next generation of developers to learn, it is to spend less time doing hard things and more time making hard things easy. Customers benefit from the former. Customers and peers and we ourselves benefit from the latter.

Aaron Patterson The Big Sort:

I like to take bad code, mix it with good code, turn it in to AAA rated technical debt, then sell insurance on the debt

cyberglot "most tech discussions in a nutshell"

D26RAZrWkAA9BnG


Teamwork

Sophie Alpert How is that not "a product team"? Product needs ops and infrastructure to work, roadmaps and design specs by themselves don't deliver value to customers.

anyone have a good name for the combined roles of

  • software engineering
  • product design
  • UX research
  • product management?

“eng” sounds like just coders; “product” can sound like PM only (and excludes infra work).

“tech”? something else?

On second thought, this is a much better name jimthe.dev:

A group of these people is called a Jira.


None of the Above

Katie Mack "#TimeManagement"

D23fBrFW0AAtLGC

Kimberly Blessing ☎️ This is the friendly thing to do:

I practically weep anytime anyone uses the one tap format (DIAL-IN,,CODE#) in a calendar invite.

TrinaCharlotte "I hate everything"

D2nlynRW0AAo7q6

sophy wong "My new USB cable from @adafruit is reversible on BOTH ENDS! I could do this all day!"

ezgif.com-optimize

Your AirPods Will Die Soon "The surprisingly short life of new electronic devices" Mine barely hold charges, and incidentally, started failing a week before Apple announced the 2nd generation AirPods 😤

The first iPhone prototype A look at Apple’s red M68.

twarren_190308_3283_2280.0

The New Social Network That Isn’t New at All The NYT predicts personal email lists making a comeback. I certainly hope so, Weekend Reading readers. Meanwhile, Google wants to make email interactive.

00MIKE-TEAR2-superJumbo

Scientists rise up against statistical significance A must read for anyone doing research, whether into medical treatment or A/B testing link colors. Using P < 0.05 as a pass/fail test can easily lead to the wrong conclusion:

It is ludicrous to conclude that the statistically non-significant results showed “no association”, when the interval estimate included serious risk increases; it is equally absurd to claim these results were in contrast with the earlier results showing an identical observed effect.

d41586-019-00857-9_16551622

There's no silver bullet, we need to be statistically literate:

Our call to retire statistical significance and to use confidence intervals as compatibility intervals is not a panacea. … But eradicating categorization will help to halt overconfident claims, unwarranted declarations of ‘no difference’ and absurd statements about ‘replication failure’ when the results from the original and replication studies are highly compatible.

Lizzie Swann Epic:

Me: Husband, please stop leaving empty wrappers on the kitchen surface. Husband:

D2mx07lXQAMl95r