Chris Wehner

@milan_milanovic You have to have the prereqs down of having a background in distributed systems for this or you’re drawing dead from my experience

𝗪𝗵𝘆 𝘆𝗼𝘂 𝗺𝘂𝘀𝘁 𝗿𝗲𝗮𝗱 "𝗗𝗲𝘀𝗶𝗴𝗻𝗶𝗻𝗴 𝗗𝗮𝘁𝗮-𝗜𝗻𝘁𝗲𝗻𝘀𝗶𝘃𝗲 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀" 𝗯𝗼𝗼𝗸 I read it twice, and here's what people who didn't read it are missing. After two decades in software, I thought I had a solid understanding of distributed systems. Then DDIA rewired how I feel about data. The book doesn't just explain technologies; it explains why they exist. Why does Cassandra use LSM-trees while PostgreSQL uses B-trees? DDIA shows you: LSM-trees optimize for write-heavy workloads through sequential writes, while B-trees optimize for reads through in-place updates. Simple trade-off, huge implications. Why does MongoDB default to single-leader replication instead of multi-master? Because write conflicts in multi-leader setups are brutal to resolve. The complexity rarely justifies the benefits. The real value isn't learning what these systems do. It's about learning how to reason about trade-offs: consistency vs. availability, latency vs. throughput, and simplicity vs. performance. Every architecture decision becomes clearer when you understand the fundamentals. Who should read it: 🔹 Mid-career engineers designing systems 🔹 Anyone preparing for systems design interviews 🔹 Architects who want to move beyond cargo-cult engineering Who might struggle: 🔸 New developers (assumes distributed systems background) 🔸 Anyone looking for quick tutorials (it's pure concepts) DDIA isn't light reading; it's 500 pages of dense material. But it's the difference between copying patterns and understanding principles. I wrote a complete breakdown covering what I loved, what I didn't, and the key takeaways that changed how I design systems. 👉 Read the complete analysis here: newsletter.techworld-with-milan.com/p/what-i-learn… Have you read DDIA? What was your biggest insight? #softwareengineering #programming #systemdesign

@BalesBets @catehall I suspect the status quo bias causes more damage than risk seeking strategies do. I’m guilty to this phenomenon myself.

I really liked this excerpt from @catehall in her post "In praise of quitting" usefulfictions.substack.com/p/in-praise-of…

@therosieum Agree, it’s very common. I think it often develops in corporate culture, and then follows the person into other pursuits. Once adopted, seems uncommon for someone to break it.

@elonmusk I visited the TSMC manufacturing plant this year, and their narrative was they cannot be dethroned.

Samsung’s giant new Texas fab will be dedicated to making Tesla’s next-generation AI6 chip. The strategic importance of this is hard to overstate. Samsung currently makes AI4. TSMC will make AI5, which just finished design, initially in Taiwan and then Arizona.

@howdoibuildthis @thiccyth0t Glad you liked it! I am still working on shifting from indexing highly just on the mean outcome ($ev) of events, to also giving proper consideration to the median outcome, case by case

@ChrisWehnerPLO @thiccyth0t Quality share ty

@mathsppblog @s_gruppetta you guys are all-stars. I found reading Rodrigo’s “little book of itertools” followed by Stephen’s “the itertools series (with Yteria)” articles to be a super effective (and complementary) learning path. Thx!

@treyhunner I find writing lots of asserts w/ LBYL to often require too much up front effort to justify what it serves. Only thing I like about this style is that they are silent when they work, at least. I prefer to just use try/except everywhere with EAFP.

How do you usually discuss EAFP vs LBYL in the context of #Python? Asking out of curiosity, as I expect perspectives might differ.

@s_gruppetta "why let mere facts get in the way of a good story" 💙

Here's the link to the article: thepythoncodingstack.com/p/python-duck-…

Today's article has a duck calling out… It discusses a bit of duck typing, and talks about callables Let me know what you think. It's more of a thought experiment than full-fledged code today link in replies, as usual

I just completed "Historian Hysteria" - Day 1 - Advent of Code 2024 adventofcode.com/2024/day/1 #AdventOfCode

now that the cost of intelligence has crashed, focus on agency

@EcZachly Is the top table on the right side the signal table?

Writing data to production is a contract! The contract states: 1. I’ve tested this data set for quality bugs in an automated way. This doesn’t mean bad data will miss production every time. But it does mean a large majority of quality errors will be stopped before reaching production. 2. I’ve validated the approach, metrics and dimensions with downstream consumers. Publishing high-quality data that nobody had say in defining is almost as bad as publishing bad data. Include your customers in these conversations. Don’t publish non-useful data to production! When pattern do you prefer? Write-audit-publish or signal table? #dataengineering

@pawjast Does writing apps with the GUI libraries lean heavily on writing in an OOP style?

@pawjast What main things should we consider when deciding whether to build a GUI app vs a browser-app?

@driscollis @pawjast So when a library is “built on top of” another library, it is importing some pieces of, but not the entirety of, the parent library?

@ChrisWehnerPLO @pawjast I don't think every function or method in PIL uses NumPy, so not exactly. PIL will import it when it needs it

@s_gruppetta_ct Why is it that it seems we can assign a counter attribute to a function object but not to some other objects, such as a list object?

Do you want to know what I ate while in Rome on holiday? No? I thought so. Instead, do you want to see some quirks that are a result of functions being objects? ("Everything is an object in Python"–how often have you heard that?!) Here's a video highlighting some of these…

@mathsppblog @dannyreigns015 This is how a dict ensures uniqueness per key and a set ensures uniqueness per item? Nice…I presume it is less efficient to compare the dictkeys and set items in their original formats without converting them into ints via using hash?

@dannyreigns015 Under the hood, an object is hashable if you can call the built-in `hash` on it. `hash` is a built-in that takes an object and converts it to an integer that is “almost unique”. It's used to implement efficient dicts and sets. en.wikipedia.org/wiki/Hash_func…

I'm working on a set cheatsheet. This is a little section on `dict` vs `set`. What am I missing?

@s_gruppetta_ct A couple ones that I think *are* sugar(?) @my_decorator my_slice = s[1::2]

Most things people call “syntactic sugar” in Python aren’t really syntactic sugar

@pawjast A primary benefit of OOP is managing state? Very helpful reframe