Lars Strojny

Broken window for software engineering: notices become warnings, warnings become errors, errors become outages, outages become bankruptcies.

LLMs are a great amplifier of the developer using them if you have taste, they let you scale your taste across your codebase faster if you're clueless, they let you scale your cluelessness across your codebase faster too

Hey #PHP, Casual Saturday morning ... 0.03 seconds on bare metal with avx2, 0.05 (if the timer can be believed) in your browser with wasm simd ... krakjoe.github.io/ort/ "The Solution" tab is where you run code ... #php #ai

> our implementation was not developed by Mozilla employees, but was contributed entirely by a single volunteer, André Bargull. Temporal is an absolutely massive spec and complicated implementation - it's been YEARS in the making. André out here just beating out billion dollar corps

@jessfraz Welcome to the faith!

for years everyone told me, "try nix, you'll love it" because I'm a weirdo who loves minimal base deps and the ability to clean shit up on the fly and i've come to report back I fucking love it

@jessfraz I’ve build a teeny-tiny abstraction so that I can have something I call universal apps in individual files, which is installation plus configuration. So hosts just select „install the git app“ and get git plus config incl. shell integration etc

@jessfraz Yes. On the Mac I also use it to drive homebrew installation. GUI packages on Mac are a bit better maintained in homebrew than in nix, so I mix and match between nix and brew as a package source. Brew doesn’t provide the same deterministic guarantees but that’s alright for GUI.

@davefarley77 And I feel „semantic diffusion“ puts it way too gently: given enough time, the worst interpretation of any good idea will prevail. The question is: why is that the case and what is there to be done about it.

@davefarley77 I’ve been thinking about exactly this a lot and there are countless more examples: object orientation came to mean writing classes, DevOps came to mean having a DevOps engineer, Observability means having a monitoring tool available.

"Semantic Diffusion" is everywhere and why this matters... a #Thread 1/19

@techspence This LAN is my LAN

Best Wi-Fi names, I’ll go first… WuTangLan 😆😆😆

I was talking with my friend @nealriley about how much fun I’m having with LLMs lately, including the project I've written about before to interpret screenshots of podcasts and YouTube videos. I made the observation that working with LLMs seems to boil down to a pattern of creating ETL pipelines, over and over again. He quickly corrected me, suggesting, "Don't you mean ELT pipelines?" I laughed at his curious and unexpected comment, which seemed unusually pedantic. However, I found myself thinking about his comment for days afterward. It reminded me of a quip from Rich Hickey, inventor of the Clojure programming language: "Information is simple. The only thing you can do with information is ruin it." In my experience, what Rich said is so true, and shows that Neal was absolutely correct. What I meant to say is "ELT," not "ETL." In other words, you should Extract, Load/Store, and then Transform data, in that order (ELT). Unless you love misery, you should not Transform before you Load/Store (ETL). Over the last many decades, I've ruined data in so many many ways, for me, the “T” in “ETL” stands for tarnish, trash, taint, etc. I’ve accidentally converted date-time strings incorrectly (turning them all into nils), I’ve discarded data fields that I thought were irrelevant, but years later, I wish I had kept them because it is priceless information, impossible to reconstruct. I’ve accidentally overwritten data, deleted data, and so many more terrible things, Thankfully, I've wiped those episodes from my memory, lest I start kicking myself again. This is because I was Transforming the data (or Trashing it) before storing it. One of the things I've learned from the Clojure and cloud community is the value of just storing data in its original form, often in storage buckets, deferring any transformation of the data to afterward, often at runtime (i.e., Extract, Load/Store, then Transform). I almost made this mistake last week: I mentioned how I use @revai and @deepgram and to create podcast audio transcripts. At first, I converted the audio transcripts to match the format used by the python-transcript-api program. That way, I could use the same code to render the transcript, regardless of what the transcript source was. This is when I once again learned that the only thing I could do to the data is to ruin it. So instead, I saved the entirety of the transcript in its original format, including the full API return payload. That way, I’m preserving optionality, enabling myself to use that data in new ways in the future. By discarding the data, I destroy that optionality. Today, I saw one such potential option: I realized that transcript diarization could be helpful (this is when transcripts detect when there are multiple speakers talking, and ideally labels them). The YouTube-transcription-api doesn’t support speaker labels, and I would have discarded that data to match it. My lessons: - Store your data in as close to its original form - Defer data transformation to after you load/store it

Last January, I noticed something peculiar in my 2yo’s bedroom that - after a year of obsessive reporting - led me to a profound cosmic revelation about what’s even possible in our universe. A 🧵.

ssh has secrets. Too many to share in one tweet. One of which is how it acts as a serial-line processor for secret keyboard functionality you probably never knew about. For example, why, when you press ENTER and then ~ immediately after, does the ~ not appear right away? Thread…

@notphanan Wohoo!

#HappyNewYear2024! No concrete resolutions but having German permanent residency (received on Dec 30!) means lots of restrictions just got lifted from my life and I’ll have much more freedom to focus on things I love the most this year.

@kicker hat Stephan von Nocks schon gekündigt? #bayer05 #carmenthomas

@GrowthMs Kapitalismuskritik am Limit

Jetzt auch noch die Gewerkschaften. "Fahimi kritisierte: »Die Finanzmärkte sind sehr volatil. Darauf kann man keinen Generationenvertrag aufbauen.«" Boah, Leute, ey 🙄

@TotherAlistair Any class has two interfaces muddled together. A “production“ interface and a “consumption“ interface. A constructor is part of the production interface but not of the consumption interface. Traditional OOP languages don’t offer a clear separation of the two interfaces

trying a poll: * If a class's constructor contains arguments, the constructor is part of the class's interface. true/false? arguments pro/con in the replies, but let's see what the poll produces.

If you are handling unicode user input you are likely doing it half wrong without realizing it until it’s pretty costly to fix. Congrats, you earned yourself a data migration project. This is what uffff solves, a PHP library to filter unicode user input uffff.readthedocs.io/en/latest/

@dan_abramov Next: feed your build process data from your web analytics data store to inform what to pre-generate

So there's a great Thai restaurant in my neighborhood called Kiin. Yesterday, I searched for their website to order some takeout. Here's the Google result.

@channingwalton It is! And he was a very economical composer, reusing patterns wherever he could, @chillygonzales blames it on his many kids 😀