Former President Jimmy Carter has died at the age of 100. Easily the most impactful former President in US history and, too rarely for that powerful position, a good man. https://www.theguardian.com/us-news/2024/dec/29/jimmy-carter-dead-longest-lived-us-president
"To all of the young people in this nation and for anyone in search of what it means to live a life of purpose and meaning – the good life – study Jimmy Carter, a man of principle, faith, and humility. He showed that we are great nation because we are a good people – decent and honorable, courageous and compassionate, humble and strong."
- President Biden
Will the "no" win this poll?
Boost if you like formal logic.
-
yes 54% (784 votes)
-
no you voted for this answer 45% (652 votes)
The Wormhole is considered a VITAL Asset for the Federation. A secure Wormhole is crucial for Commerce & rapid deployment of Starfleet, between the Alpha & Gamma quadrants. It was given to Bajor to manage. The fees being charged by Bajor are a complete “rip-off” and will stop…. #StarTrek
Miles: You're the love of my life and my best friend. I would do anything for you.
Benjamin: I want you to eat three meals a day and have a decent sleep schedule.
Miles: Absolutely not
TIL pandas are one of the few animals with opposable thumbs - but the "thumb" develops from a wrist bone, so when looking at pandas, it looks as though they have six fingers
https://evolution.berkeley.edu/examples-of-analogies/when-is-a-thumb-a-thumb/#:~:text=Studying%20the%20anatomy,finger%20in%20pandas!
#til #todayilearned
https://www.reddit.com/r/todayilearned/comments/1hmyscd/til_pandas_are_one_of_the_few_animals_with/
A reader found her father in a photo on the back cover of Isabel Wilkerson’s book Caste. The photo was of a crowd of people at the 1963 March on Washington. https://www.instagram.com/p/DD-f5CWyeyc/
This hotel in Monterey Bay, CA has a stick library available for dogs. “Every morning my dog would carefully pick one out to take to the beach.” https://www.instagram.com/reel/DDcQuBVJVlO/
Hi, friends!
The incredible @dannotdaniel made a PartyAlice emoji!
@theropologist added it to beige.party, so everyone on beige.party can use it.
If you're not on beige.party, you can ask your admin to add it to your instance.
Also, if you're willing, will you please boost this post because nothing would make me more happy than seeing a bunch of PartyAlices floating around the Fediverse.
I wasn't allowed to watch "A Charlie Brown Christmas" as a kid because of my Peanut allergy.
Al Green’s cover of R.E.M.’s “Everybody Hurts” is as great as you’d expect it to be. https://kottke.org/24/12/al-greens-cover-of-rems-everybody-hurts
A Thirty-Year-Old Man Today vs. a Thirty-Year-Old Man in 1884. “Today: Is working toward paying off a two-bedroom condo by age seventy-five, if all goes well. 1884: Owns a drafty, four-hundred-square-foot cabin by the crick. Loves it.” https://www.newyorker.com/humor/shouts-murmurs/a-thirty-year-old-man-today-vs-a-thirty-year-old-man-in-1884
I finally turned off GitHub Copilot yesterday. I’ve been using it for about a year on the ‘free for open-source maintainers’ tier. I was skeptical but didn’t want to dismiss it without a fair trial.
It has cost me more time than it has saved. It lets me type faster, which has been useful when writing tests where I’m testing a variety of permutations of an API to check error handling for all of the conditions.
I can recall three places where it has introduced bugs that took me more time to to debug than the total time saving:
The first was something that initially impressed me. I pasted the prose description of how to communicate with an Ethernet MAC into a comment and then wrote some method prototypes. It autocompleted the bodies. All very plausible looking. Only it managed to flip a bit in the MDIO read and write register commands. MDIO is basically a multiplexing system. You have two device registers exposed, one sets the command (read or write a specific internal register) and the other is the value. It got the read and write the wrong way around, so when I thought I was writing a value, I was actually reading. When I thought I was reading, I was actually seeing the value in the last register I thought I had written. It took two of us over a day to debug this. The fix was simple, but the bug was in the middle of correct-looking code. If I’d manually transcribed the command from the data sheet, I would not have got this wrong because I’d have triple checked it.
Another case it had inverted the condition in an if statement inside an error-handling path. The error handling was a rare case and was asymmetric. Hitting the if case when you wanted the else case was okay but the converse was not. Lots of debugging. I learned from this to read the generated code more carefully, but that increased cognitive load and eliminated most of the benefit. Typing code is not the bottleneck and if I have to think about what I want and then read carefully to check it really is what I want, I am slower.
Most recently, I was writing a simple binary search and insertion-deletion operations for a sorted array. I assumed that this was something that had hundreds of examples in the training data and so would be fine. It had all sorts of corner-case bugs. I eventually gave up fixing them and rewrote the code from scratch.
Last week I did some work on a remote machine where I hadn’t set up Copilot and I felt much more productive. Autocomplete was either correct or not present, so I was spending more time thinking about what to write. I don’t entirely trust this kind of subjective judgement, but it was a data point. Around the same time I wrote some code without clangd set up and that really hurt. It turns out I really rely on AST-aware completion to explore APIs. I had to look up more things in the documentation. Copilot was never good for this because it would just bullshit APIs, so something showing up in autocomplete didn’t mean it was real. This would be improved by using a feedback system to require autocomplete outputs to type check, but then they would take much longer to create (probably at least a 10x increase in LLM compute time) and wouldn’t complete fragments, so I don’t see a good path to being able to do this without tight coupling to the LSP server and possibly not even then.
Yesterday I was writing bits of the CHERIoT Programmers’ Guide and it kept autocompleting text in a different writing style, some of which was obviously plagiarised (when I’m describing precisely how to implement a specific, and not very common, lock type with a futex and the autocomplete is a paragraph of text with a lot of detail, I’m confident you don’t have more than one or two examples of that in the training set). It was distracting and annoying. I wrote much faster after turning it off.
So, after giving it a fair try, I have concluded that it is both a net decrease in productivity and probably an increase in legal liability.
Discussions I am not interested in having:
- You are holding it wrong. Using Copilot with this magic config setting / prompt tweak makes it better. At its absolute best, it was a small productivity increase, if it needs more effort to use, that will be offset.
- This other LLM is much better. I don’t care. The costs of the bullshitting far outweighed the benefits when it worked, to be better it would have to not bullshit, and that’s not something LLMs can do.
- It’s great for boilerplate! No. APIs that require every user to write the same code are broken. Fix them, don’t fill the world with more code using them that will need fixing when the APIs change.
- Don’t use LLMs for autocomplete, use them for dialogues about the code. Tried that. It’s worse than a rubber duck, which at least knows to stay silent when it doesn’t know what it’s talking about.
The one place Copilot was vaguely useful was hinting at missing abstractions (if it can autocomplete big chunks then my APIs required too much boilerplate and needed better abstractions). The place I thought it might be useful was spotting inconsistent API names and parameter orders but it was actually very bad at this (presumably because of the way it tokenises identifiers?). With a load of examples with consistent names, it would suggest things that didn't match the convention. After using three APIs that all passed the same parameters in the same order, it would suggest flipping the order for the fourth.
If you're struggling, seek support.
Mindline Trans 0300 330 5468 (Fridays 8pm-midnight)
TRUK Listens 0800 009 6640 (8am-midnight daily) https://truklistens.org/
Switchboard 0800 0119 100 (10am-10pm daily) https://switchboard.lgbt/
Samaritans 116123 (24hrs every day) https://www.samaritans.org/
Mastodon #research 🚀
I’m working on my master’s thesis at the University of Vienna, and I’d love your input! My research focuses on sustaining long-term user activity on Mastodon, contributing to a more active Fediverse. If you have a few minutes, please take my short survey:
👉 https://sosci.univie.ac.at/MastodonActivity/
Your thoughts will contribute to meaningful research. Feel free to boost this post so we can reach as many users as possible. Thank you so much! 🙌
Today during our daily stand-up, someone on our team referred to our Scrum Master as "The Master of the Scrum".
Now I'm imagining a Broadway musical about a software development team, with the finale of the first act being Master of the Scrum.
Defying Gravity from Wicked, Non-Stop from Hamilton, One Day More from Les Misérables are all pretty good. But Master of the Scrum from Silicon (title is a work in progress) will be a real show-stopper.
Does @[email protected] use the Mastodon API or should it work with any ActivityPub instance?
I guess this should tell me.
I'm not crying, you're crying.