cube00

A few stories in the news represents a beyond negligible fraction of billions of user accounts.

https://support.google.com/accounts/threads?thread_filter=(c...

How many times do you have to click "View more" to reach a post from last week? It's a problem.

The copy pasta advice from the "diamond product experts" says it all:

Account recovery is completely "self service". Google doesn't provide any form of live support for account recovery. If you can't recover your account using Google's automated recovery process, the account is lost. [1]

[1] https://support.google.com/accounts/thread/453660883/locked-...

You have to consider that the user base of Google accounts is in the billions.

Basing your security decisions on "it'll never happen to me" because there are billions of other users who will get burnt first is not a wise strategy either.

Why take the chance when there are so many other alternatives that let you own your vault or at least companies that still have some semblance of a support team.

Losing a decade of my Google Maps Timeline data even with backups enabled made me realise I may not be lucky enough to win the lottery but I am lucky enough for Google to pick little old me, hidden in the billions, to lose my data.

While this all sounds like a warm and fuzzy nostalgic trip, big music didn't like us doing this when we made mix tapes off the radio.

Spotify certainly doesn't like it and it's second only their prohibition on reverse engineering their own software in their User Guidelines.

  The following is not permitted for any reason whatsoever in relation to the Services and the material or content made available through the Services, or any part thereof:
  2. copying, reproducing, redistributing, "ripping," recording, transferring [1]

[1] https://www.spotify.com/us/legal/user-guidelines/#:~:text=co...

Being limited to only devices from a single vendor is horrible, and a firm no from a lot of people.

Maybe for a lot of tech people, I don't think the general public understands the risks enough to care.

Considering Google's track record of locking people out of their accounts it'd be very risky to trust them with your password manager.

According to the ads manager UI, unless I'm willing to allow up to $3 a click I'm warned "my bid may be too low to deliver reliably".

Considering I'm paying 20 cents per click elsewhere, this could get real expensive, real fast.

Are your customers clearly informed that you're sending their immutable fingerprints to an AI service?

I simply could no longer provide the effort (mental or time-wise) that the role demanded, and thus I was not performing my duties to an acceptable degree, facing severe burnout, and risks to my mental health.

[...]

We've proven that FLOSS works, and is in demand.

If "FLOSS works" maintainers shouldn't be "facing severe burnout", I've come across two more in the past week alone:

* Filebrowser (Mar 2026): https://hacdias.com/2026/03/11/filebrowser/

* axum-login / tower-sessions (last week): https://github.com/maxcountryman/axum-login/discussions/330

They've disabled reviews on all kinds of law enforcement locations for some time.

Examples of places that may have long-term restrictions include police stations, prisons, oceans, seas, and schools.

https://support.google.com/contributionpolicy/answer/1303613...

Oh well that changes things.

It's the same for any case where someone sends you something and wants it sent back with changes.

If you screw up their files it's never going to go well for you (family, friends, volunteer groups).

Nobody will ever thank you for messing up their work because it's now saved in an open format.

If they end up using a feature that triggers the format nag screen they'll likely end up saving it back into the native Microsoft format anyway.

Smaller annual opex is far easier to get approved then large multiyear capex with deprecation.

AWS was also cheaper in the beginning but has been slowly ratcheting up the pricing making colocation or on-premise viable again for some workloads which have steady compute demands.

send them back an ODT

The formatting goes wonky or the macros stop working and you get the blame for doing something "weird" to the document everyone else managed to pass around just fine.

As much as you're doing the right thing by using open formats, you still end up looking like the incompetent employee. No good deed goes unpunished.

I give the PR a proper title on creation. I don't see any reason to create a PR with a title of "test ci" and then rename it later to:

  Rewrite Bun in Rust #30412

it happened completely in the open

You don't create PRs like this when you're working "completely in the open"

  test ci #30412
  No description provided.

A breath of fresh air to pause new subscriptions rather than the Google approach of quietly nerfing the limits and hoping you don't notice you're getting less value for your monthly/annual subscription.

Limits may change without notice, including due to capacity constraints. When there’s a large increase in activity in Gemini Apps, we may change limits to maintain a high standard of quality. [1]

[1] https://support.google.com/gemini/answer/16275805

I'm guessing it's essentially just "Anthropic decides what gets done and accepted" today?

Similar to the Claude Code 60s timeout incident [1] it could just be "Jarade decides what gets done and accepted"

[1] https://news.ycombinator.com/item?id=48949942

Seems to be working just fine though?

As with all transpile ports, the true test will be how well it can be extended and maintained over time. Historically working with the output of a transpile is not pleasant and requires heavy rework to get it to the point where you can be comfortable enough to extend it.

The existing community is now either going to need to learn Rust or new Rust developers are going to have to join the project and they may not invested enough to refactor the transpiled output when they could work on something else like Boa.

There has to be a second PR approver. There's no way a company as large this can allow an individual the ability to push to production without a secondary person involved.

When there's at least two people involved you can then start to look at more systematic factors that go beyond any single human mistake.

When both the submitter and approver miss that there's no changelog entry for a PR does that mean a checklist is needed? Should the changelog be automated with the commit message which can be improved for external consumption?

I'd find it very hard to believe that this high profile change just happened to be the only change missing from the changelog.

Bugs are always going to happen but housekeeping like writing something to a changelog when a commit is merged can be quality assured.

Not every feature will necessarily appear in the changelog

This was such a frustrating part of this incident, along with Anthropic's refusal to explain why the changelog is no longer a complete record, what else is going out? [1]

Boris Cherny's only participation in the thread was to delete "extreme danger" from the GH issue title [2]

I guess we should be thankful they added an option and disabled it by default. OpenAI is standing firm on their decision to not allow their 60s timeout to be disabled, [3] however more of the Codex harness is open source so customers have been able to fork it to add the option themselves.

[1] https://github.com/anthropics/claude-code/issues/73125

[2] https://github.com/anthropics/claude-code/issues/73125#event...

[3] https://github.com/openai/codex/issues/28969

I didn't have time to recreate the entire comic before publish date.

It's depressing that even a blog post about open sourcing a two decade old piece of software has such a hard deadline the author feels pressured to publish before they're ready.

SLM was at version 1.5 by 1988 and looking at chapter 5 suggests it had strong version number and external release management [1]

[1] https://fpga.org/wp-content/uploads/2023/09/SLM-1.5-Guides.p...

  v1.0-pre and v1.0 share the same internal version number (rup 206, "Beta 2") but differ in ~99 of 111 shared source files [1]

[1] https://github.com/microsoft/comic-chat#:~:text=v1.0%2Dpre%2...

[2] https://devblogs.microsoft.com/oldnewthing/20251028-00/?p=11...

I wonder if their use of a "buy" button would potentially weaken their case regardless of the language they put in the EULA.

I used to think I had to use a component library, but for some cases it seems all I really needed was CSS to give my webapp that material feeling everyone seems to feel more comfortable with.

I now have a more critical eye and look into potential UI components code to decide if it really does need the custom code or if styling alone can get the job done.

Shoelace upending everything to Web Awesome has given me an incentive to revisit this rather then blindly find/replacing all the sl- prefixes to wa- given how much larger it makes my bundle.

I hate the video links try to open in the app if you're not careful!

Uninstall (disable) the app, YouTube on Firefox mobile is fine.

--system-prompt ""

Doesn't the model need at least a basic system prompt to understand what tools are available?

  You can request deletion of your Google user data by emailing the official support channel. [1]

I'm logged in with my Google account over a secure connection and yet I can only delete over insecure email.

Not that it really matters when there's no support email address anyway.

  For privacy-related questions, please contact the application operator through the official support channels listed by your team. [1]

There is a privacy popup setting for:

  Marketing - These cookies are used to deliver relevant advertisements and track their effectiveness.

  Ello has specific instructions to talk to the child about the learning activities.

There really needs to be more published technical detail on what safety systems you have in this because if trillion dollar companies can't stop AI going off the rails, I feel you're overselling the safety systems you've built.

Reading further it then goes beyond "learning activities" into LLM generated content.

  Some (not all) of Ello’s stories and illustrations have been created with AI technology, and AI allows children to create their own stories and experiences through guided and safe scaffolding.

He's been teasing this blog post everywhere (in commit messages, multiple times on X, here on NH [1]) so I wonder how much of it was building hype compared with it legitimately taking two months to write.

I wonder if the delay of Fable has also been a factor and maybe they didn't want to release this blog post while they couldn't allow customers to use Fable and waste the advertising opportunity.

[1] https://news.ycombinator.com/item?id=48133519

the assumption that the rewrite was happening with an Opus-like model, and not with Fable

I thought the same thing. Looking back, I was probably mislead in May when Jarred was explaining the pattern to "Rewrite every .zig file to .rs" as if it was something I could have done in May following his pattern. What he wasn't telling us was he was using pre-release Fable. [1]

A possible signal for next time is when we see an Anthropic owned company disabling the Claude Co-Authored-By trailer. [2] In an IPO year they have to take every chance to promote Claude unless it was something (Fable) that we weren't supposed to know about back in May.

[1] https://xcancel.com/jarredsumner/status/2060050586024743376#...

[2] https://github.com/oven-sh/bun/commit/23427dbc12fdcff30c23a9...