mike_hearn

Joined 13 years ago

Karma
16,346
Submissions
6,329
Member since
Dec 16
Email me:

[email protected]: Hydraulic Conveyor makes distributing desktop apps as easy as a web app. Use it to bring your mobile apps to the desktop and avoid the need to maintain a web version, to use your language and toolkits of choice or to do things that you can't do in the browser. Integrated support for Electron, Flutter, JVM and native apps.

https://www.hydraulic.dev

--

[email protected]: Personal email. I've worked on anti-automation, was an early Bitcoin developer and then designed the Corda enterprise blockchain platform and the Conclave toolkit for writing SGX enclaves.

View on Hacker News

Recent submissions

Comment · 6 hours ago
IIRC BeOS gave every window its own thread rather than have one GUI thread per app. So in effect every app was multi-threaded by default.
Comment · 7 hours ago
I think they were planned to be capitalists from the first. They came out of the writers wanting a scary alien race to replace the Klingons. It shows how extremist the writers room had become that what they came up with was capitalists. They only scrapped the idea of the Ferengi becoming the new antagonists after screen tests gave them a reality check.

Their second attempt at this created the Borg though, so they did get it very right in the end!

Comment · 7 hours ago
> the system worked

Not really! There were several systemic failures:

1. It happened due to forecasting errors. Nothing remarkable happened that day, there were no sudden station trips or natural disasters. The margin was gone mostly due to planned maintenances combined with incorrectly forecasting demand. Whether the root cause is decentralized PV or not, the core issue is that the lowest frequency ever recorded happened on a totally unremarkable day. That can't be said to be the system working correctly.

2. The UK had to suddenly cut exports that had already been agreed and paid for, outside of normal systems. That puts it in a bad situation in future where allies may now do the same in reverse. A sudden unannounced cut in exports can be absorbed by the European grid but might not be absorbed by the British grid, which is smaller.

3. There were no EMNs issued and in the aftermath whistleblowers inside NESO claimed control room staff had been forbidden from taking notes or letting people know what happened. Information about this came out via newspaper articles in a drip drip months after the event. We're talking about this now and not in June because of this. That lack of transparency can't be said to be the system working.

4. The French were able to ship power hundreds of kilometers across France and under the sea to the south of England, but the UK couldn't move power the much smaller difference from its own north to south. That's another form of planning failure.

> Removing solar isn't helpful (that would just move the problem to earlier in the day.)

That depends what you think the problem is. The root of the event was forecasting errors, and the root of those come from the apparent difficulty of predicting how decentralized solar behaves especially at a sub-region level. Removing solar, or simply having never built it at all, actually would fix that problem.

All of this was predicted decades ago by conservatives so it's no good to just despair that they will "make hay" from being right (nobody will benefit from a big UK blackout, it will be bad news for everyone). They warned repeatedly that putting too much solar/wind into the grid will destabilize it and create huge price increases. Both of those warnings have come true in Europe. Solar/wind aren't reliable and so need thermal/hydro that's staffed, warmed up and ready to spin at a moment's notice but government policy ignores this and the stresses are now showing - both engineering and political (the apparent need for whistleblowers shows this).

N.B. the root cause of the Iberian blackout was a solar farm malfunctioning and desynchronizing from the grid, and when the control room tried to bring gas generation back online to stabilize things they couldn't do it fast enough because the plant was cold. It had been taken offline in the morning to make way for solar.

Comment · 9 hours ago
There were other issues too, no? They went all in on multi-threading of the GUI which is super hard to do well, so apps were often crash prone.
Comment · 9 hours ago
Yeah that's weird. I wonder if it's a bug. This particular permission prompt does appear way too often especially as it's hard to understand.
Comment · 9 hours ago
That hasn't changed. Rare is the writer who can just write a good story and not try to socially engineer the audience.

The problem with TNG wasn't so much the preachiness, which is still normal for TV. It's that the moral system behind it was actually quite poor. Picard's character was certainly not the paragon this thread is making him out to be - the more you look at Star Trek the worse it morally gets. It's one of those shows that works best when you don't think about the worldbuilding too much. And the Picard character only seems moral and superior on the surface because of the way Stewart acted him, and because his character is so unemotional, not because of what he actually does.

Consider the moral lessons of TNG and Picard.

Picard isn't married, has no family or children, and apparently no desire to change that. He's actually distinctly uncomfortable around children. Instead he married his job, which is all consuming. Right there that renders the character boring, overly intense and unrelatable for many people. When he gets assimilated by the Borg there's no emotional impact because it's not clear who cares about Picard being turned into an alien except his staff, whose only concern is professional duty. In fact when Riker decides to attack the cube with Picard still on board, he barely hesitates, and why would he? Picard keeps everyone around him at great emotional distance at all times. When the writers want to show Picard having an emotional breakdown about his traumatic experience, they have to show him running home to a brother he hardly ever sees or talks to, because that's the only character the writers have who might care about Picard as a person. Sad.

Picard is supposedly a culture maven, but neither him nor his crew ever creates any. They're obsessed with replaying stories and music old even when the show was made like Moby Dick, Beethoven, Shakespeare, detective stories set in the 1940s.

Although the Enterprise has many crappy jobs on it along with many scarce resources (e.g. holodeck time), crew are unpaid. Picard says everyone on the Enterprise is there because Starfleet is prestigious and he seems fine with that idea. This arrangement could never work in reality so he comes across as delusional, and worse, we are left to assume there actually are mechanisms that incentivize or force people to beam down into severe danger, but Picard pretends they don't exist.

By the 1980s Roddenberry was a Maoist and the Federation appears to be some sort of communist dictatorship. We're told it's abolished money and we don't encounter operating businesses or conventional democratic politics. There is notionally some sort of elected leader in the Federation but they don't seem to control anything and Starfleet does whatever it wants. For example, TNG doesn't show Picard being given new orders because the elected government was changed, nor does Picard ever wonder how his actions might affect politics back home.

Picard holds the capitalist Ferengi in contempt and makes dumb diplomatic mistakes, like referring to "civilized societies" when talking to a Ferengi captain despite that the Ferengi fly around in advanced spaceships. This is because Picard views trade as dirty and beneath him.

Although he does occasionally disagree with Starfleet orders, he slavishly follows the Prime Directive even when the outcome would be absurd, like the entire civilization dying. He never even questions this rule despite the writers realizing that it leads to immoral outcomes! Starfleet is ambiguously presented as sometimes a sort of scientific research org and other times a military that in DS9 implements full blown martial law, suggesting the normal military/civilian separation doesn't exist. Picard never questions the structure of Starfleet or Federation society, which he seems to view as automatically superior (Sisko does a lot better in this regard).

Comment · 1 day ago
Are you always clicking Deny? I don't see this. I think denials might not be cached because Apple fear that otherwise the user won't know how to change their mind later.
Comment · 1 day ago
Are you arguing it's less work to rewrite your entire app in Rust than to use GraalVM Native Image? Because you can get C/Rust style startup time and less memory usage with the latter, but it should take an agent only a few minutes to get things working (unless you are using a lot libraries that spin bytecode, but there is a mode for native image coming that can fully handle bytecode generation and can run any Java app).
Comment · 1 day ago
That explanation sounds wrong to me. I've written my own implementation of macOS signing logic. TCC doesn't usually tie permission grants to paths, it ties it to the designated requirement of the app bundle, which in turn is usually an indirection through the code signing certificate identity. The whole macOS architecture is designed to let you drag apps around and treat them as files, it's not a UNIX where installation path is meant to matter.

I just dumped my TCC database and the entries are mostly not path tied. The Chrome entry is as expected a bundle ID+designated requirement. Unless the local network permission inexplicably uses a different permissions engine, that statement is a misdiagnosis of what's happening.

     $ sudo sqlite3 "/Library/Application Support/com.apple.TCC/TCC.db" "SELECT hex(csreq) FROM access WHERE client = 'com.google.Chrome';" | xxd -r -p | csreq -r /dev/stdin -t

    (identifier "com.google.Chrome" or identifier "com.google.Chrome.beta" or identifier "com.google.Chrome.dev" or identifier "com.google.Chrome.canary") and anchor apple generic and certificate 1[field.1.2.840.113635.100.6.2.6] /* exists */ and certificate leaf[field.1.2.840.113635.100.6.1.13] /* exists */ and certificate leaf[subject.OU] = EQHXZ8M8AV
In other words, the permission grant applies to any Chrome channel as long as it's signed by Google's team ID.
Comment · 1 day ago
Don't know about Adobe but the issue with Claude Code is when Claude does a recursive find or grep on your home directory. You might be able to fix it just by putting some warning against doing that in your global CLAUDE.md

CLI/developer tools in general often lose or forget permissions because they aren't shipped as proper signed Mac app bundles, so the OS has no stable identity to attach the permission grant to. You might find that even if you click allow, that gets forgotten too sooner or later.

Comment · 1 day ago
I think this S-1 is about as close as you can get to announcing such a policy. It doesn't literally commit them to never returning capital, but it appears to block off or rule out pretty much any chance of them doing so via various means.
Comment · 2 days ago
You're arguing with the claim that else is archaic, but this sub-thread is about whether it came from a German translation.
Comment · 2 days ago
But in reality people hire dog walkers, nannies and care homes all the time. What people might want to do isn't always the same as what they can do.
Comment · 2 days ago
The document in question specifically says it is a rushed translation of a German document, what's in dispute here exactly?
Comment · 2 days ago
Rely completely on LLM moderators. But it's problems largely stem from its reliance on volunteer human moderators, who often self-select for strange and problematic characteristics (e.g. they are attempting to advance a political agenda by manipulating what people can read on the site).

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton