Mark D'Arensbourg

68 posts

Mark D'Arensbourg

Mark D'Arensbourg

@mdarens

web developer in new orleans. opinions include "accessibility is good" and "we should improve society somewhat." https://t.co/4Uu80K28KO

New Orleans Katılım Ocak 2012
82 Takip Edilen22 Takipçiler
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
the real target customer for the apple vision pro is somebody who's been putting off refactoring a 3k loc javascript function
English
1
0
0
96
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
if you want to drive a developer to the brink of madness, give them a bunch of unit tests with bugged assertions
English
0
0
0
48
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
to add to this, people have an inclination to misuse it to produce implementations they assume to be correct/exhaustive. but one area where it's already infinitely valuable is education. it may not be a full fledged engineer yet, but it's quite the tutor.
English
0
0
0
31
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
ChatGPT is extremely impressive, yes, and has a tendency to slip insidious oversights past you even as you continually steer it. I would pay real money for a version which iterated responses based on runner output from a unit test prompt until it detected a pass.
English
2
0
1
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
there are a lot of things you can do to reduce stress. exercise, meditation, etc. but the biggest one by far is to write failing tests before modifying legacy code.
English
0
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
request waterfalls and time to first byte are almost certain to dwarf every other perf concern in every web app. start there
English
0
0
1
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
"react is slow" is the new "x programming paradigm is slow." for most domains, the biggest perf gains come from scheduling the biggest chunks of work for when they're needed. a really good model will yield great performance on even mediocre tooling and runtimes.
English
1
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
is anybody working on using the es record literals+deep property syntax proposal for serialization and schema definition?
English
0
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
in type-fest this is called LiteralUnion. mystery solved
English
0
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
apparently a "weighted union" is actually already a thing for finding nodes in a tree structure, which this is not. better names welcome (bonus if it's already a name for precisely this from type/set theory)
English
1
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
i'm not against writing specs with ascii charts but i think they should be in a parseable format like mermaid charts
English
0
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
"i barely feel this formal methods gummy dude" one hour later:
Mark D'Arensbourg tweet media
English
0
0
0
0
Mark D'Arensbourg
Mark D'Arensbourg@mdarens·
no way i'm getting any sleep for the next 36 hours or so probably
English
1
0
0
0