• Distributed Systems | Embedded Systems | O11y | Data Science | ML
• Finance | Fitness | Music | Movies
• YouTube University student
• aiming 1%: Atomic Habits
@debasishg Congrats! 🎉
Today morning, I saw the announcement that Seligman Ventures has invested in this company.
Rote looks like an interesting idea. 👍
Firefox on macos. https://t.co/TgQB7mSCD9 loads in "light" mode.
No "dark" mode?
I discover by accident that it's available!
The button is almost hidden in bottom left corner.
That's like an easter egg!🥚
@marimo_io,
is that by design? 😀
#UI
@vivekgalatage Beej's network programming, #Unix IPC guides are quite popular. 🙂
But I was not aware this one. Thanks for sharing! 👍
It even has a chapter on Worktrees! These days, #Git worktrees are gaining lot of attention due to #GenAI agentic workflows.
This chart is a good reference for understanding the compiler behavior w.r.t. special member functions.
--
How I Declare My class And Why - Howard Hinnant
https://t.co/tcK2yHIY17
#cpp#cplusplus
2/2
Highly informative presentation about std::move & special member functions by Howard Hinnant, co-creator of std::move, unique_ptr, <chrono> etc.
https://t.co/OfzbCBvjRa
#cplusplus#cpp
1/2
@christinaqi Good info. Thanks for posting. 👍
Exchange (source) -> databento (provider) -> customers (consumer)
Q. What is the average latency for these hops?
Trying understand how fast the data moves as an engg challenge.
As the title suggests
1. CPU cache behavior is a major factor
2. contiguous, cache-friendly data structures tend to perform better than fancy data structures
Another key takeaway is that benchmarking is a must.
2/2 #cpp
https://t.co/fpQFacbMTR
In this @cppcon 2025 presentation, Michelle D'Souza starts with a bold statement:
Replacing unordered_map with vector resulted in 10X speedup! 🚀
1/2
@mgill25 Interesting approach. Thanks for sharing!
How often the 2 windows are evaluated?
Do you compare current data vis-a-vis historical data (say, a week or a month ago) to check for trends?
@MarcJBrooker@russellrcohen@golang 2. Context Switch Delay
JVM uses OS threads and relies on the OS kernel for scheduling.
Go implements its own scheduler and sticks the mostly idle goroutines on their own OS thread and the small # of active goroutines are scheduled by a separate thread.
@MarcJBrooker I followed some links and found @russellrcohen's excellent blog post explaining the reasons for scale difference between @golang vs #Java threads.
https://t.co/dkUlK4bCyf
@MarcJBrooker@russellrcohen@golang The blog talks about two main differences.
1. Stack size
In a 64-bit environment, the JVM defaults to a 1MB stack/thread, whereas Go has dynamically sized stack.
A new goroutine will have a stack of only about 4KB !
1/2
@jeremyphoward@levelsio@jeremyphoward Agreed 💯.
I mean, these days, you can just ask any LLM of your choice to generate a custom config file (.tmux.conf) with say, vi or emacs style key bindings, mouse support etc. 🤷
In the last few years, Forward Deployed Engineering has emerged as a new role.
In this blog post, Vlad Shulman @ @baseten offers a nice overview of what FDE is and whether you should have a FDE team. 👍
https://t.co/yGzlwascbT
1/2 #AI#ML
A #FDE team sits within the Engg & works directly with your customers’ engineers.
Consultants, solutions architects etc. often operate at the boundary between the product & the customer.
FDE expands the product's perimeter.
2/2 #AI#ML
In the last few years, Forward Deployed Engineering has emerged as a new role.
In this blog post, Vlad Shulman @ @baseten offers a nice overview of what FDE is and whether you should have a FDE team. 👍
https://t.co/yGzlwascbT
1/2 #AI#ML