Peter Schuller
@scode
Human. I'm KZY2KZZH on Threema, scode.01 on Signal.
Was dir gefallen könnte
The never ending creep of "needs a reboot" to more and more basic devices has achieved a new low: Needed to power cycle my thunderbolt hub to unwedge it.
Still broken as of today. Chose mail-based PIN instead. cc @CalFTB
Made the mistake of trying to pay an adjusted fee today to the CA FTB (cc @CalFTB). Which they require me to do online or face penalties. (Online is fine if the system works.) Here is the insane list of broken parts of this process that ultimately resulted in not being able to do…
macOS Tahoe seems to have gone very round and beveled. We already had like a couple of years of GUI staffing going to "omg rounded corners" several years ago. I hope we don't do this again.
Google has broken the 98% reason I ever open the Home app. Which is to unlock my lock. Now the app "loads" for many seconds when starting instead of immediately working. Conveniently, the "feedback" feature in the app is also broken so I can't even tell them.
Made the mistake of trying to pay an adjusted fee today to the CA FTB (cc @CalFTB). Which they require me to do online or face penalties. (Online is fine if the system works.) Here is the insane list of broken parts of this process that ultimately resulted in not being able to do…
We are alarmed by reports that Germany is on the verge of a catastrophic about-face, reversing its longstanding and principled opposition to the EU’s Chat Control proposal which, if passed, could spell the end of the right to privacy in Europe. signal.org/blog/pdfs/germ…
I find it highly amusing how everyone is talking about "AI slop". Meanwhile my biggest annoyance is human slop - all the vapid "5 prompts that are awesome" and "I tried X and Y assistant and one is better" crap that humans are spewing out. Or at least I think so - maybe that's in…
Given the alleged rules supposedly serious journalism is meant to follow, can we add something that bans the use of "associated" and "linked" in favor of terminology that explicitly includes whether or not a causal claim is being made? And mandatory linking/hovers to explain what…
My single most commonly use exact chatgpt prompt: "re-assess, think hard".
Retweeting this too since I retweeted the previous claim which turned out to not really be true (seems like some overzealous extrapolation occurred). So everyone can go back to arguing over implementation details of autonomous vehicles instead of breathing a word about…
After doing some more digging, I think this news piece that I shared yesterday is partially incorrect in its conclusions. The June 13th and Sept 4th NHTSA announcements (linked below) do indicate that the NHTSA is modifying specific standards to better align with ADS vehicles,…
Imagine if all the effort that went into arguing over style and tabs/spaces (prior to auto-formatting becoming the norm), had been spent teaching everyone to ask "Can you please add a comment explaining why we are doing it [in] this [seemingly weird/non-obvious] way?".…
If magic isn't real how do you explain how putting a cable in a cupboard and just letting it sit for a year somehow tangles it beyond all reason.
Serious question. If I'm logged into a company's site (in this case AT&T) and I use chat support. Why don't they make support agents immediately see who I'm logged in as, instead of having them ask me for account numbers and phone numbers to "verify" who I am? What is going on?
I very much hope the Agent Client Protocol / ACP catches on: zed.dev/blog/bring-you…
Thread: Tesla’s traction control is next level. Not talking ESP / yaw control, although they do some interesting things I haven’t seen before, strictly talking wheel speed control Here’s why Tesla is different (btw BMW claims to now be doing this with their new platform EVs 👌)
"Do as I do, except when I do it wrong." :P GPT-5 also makes the classic human mistake of putting comments in the to-be-committed code which address the reader as if they are reading the diff, instead of actually making sense while reading a snapshot of the code. I find this…
Well, looks like GPT-5, like Sonnet 4, also suffers from duplicating production logic in test-only code and then unit testing that code instead of the production code path.
Haha. Cursor's agent web interface made it so hard for me to copy and paste out of the chat history that I took a screenshot and asked AI to transcribe it into text. I don't know whether to be happy at AI or angry at the AI agent interface.
OAuth experts: I want to validate my understanding is correct (after spending considerable amount of time trying to verify through documentation which is implicit). If I have a *DESKTOP* oauth client in a Google Cloud project, it seems to be documented/standard practice to embed…
United States Trends
- 1. Aaron Gordon 8,622 posts
- 2. Wentz 23.4K posts
- 3. Vikings 49.3K posts
- 4. Chargers 53.9K posts
- 5. Pacers 21.4K posts
- 6. Brosmer 3,707 posts
- 7. #criticalrolespoilers 6,678 posts
- 8. Shai 20.7K posts
- 9. Herbert 15.8K posts
- 10. #Skol 5,181 posts
- 11. $UNI 3,166 posts
- 12. Quentin Johnston 2,134 posts
- 13. Al Michaels 1,257 posts
- 14. #TNFonPrime 2,360 posts
- 15. Cam Johnson N/A
- 16. JJ McCarthy 2,198 posts
- 17. #LOVERGIRL 10.5K posts
- 18. Ajay Mitchell 2,126 posts
- 19. Justin Jefferson 4,055 posts
- 20. Jokic 13.6K posts
Was dir gefallen könnte
Something went wrong.
Something went wrong.