HNHacker News
TopNewBestAskShowJobs

gumby

65,359 karma · joined June 19, 2010

Currently working on removing excess methane from the atmosphere.

Previous lives:

gnu hacker (e.g. co-founder of Cygnus back in '89) Pharma developer (Talima) Distributed solar thermal systems (Terrajoule) MIT AI Lab/PARC/MCC/Rodin consumer tabletop AR (CastAR) Etc...

remove-everything-up-through-last-dash-gumby@brachiate.org

submissionscomments
gumby··on Small Inventors Are Being Squeezed by a Convoluted Patent Process
The patent system grew out of the European guild system. It was the opposite of regulatory capture: the local authority (typically a king) would be paid to sanction a monopoly (on shoemaking or whatever). So the shoemakers could keep their secrets but even if one leaked it didn’t matter because they controlled making shoe at all.

The US patent system was the first that worked differently: it was a bribe to the inventor for disclosure: if you describe your invention so anyone could copy it, the government will let you use the courts to enforce a temporary monopoly.

Sometime in the 20th century the “so anyone could copy it” part fell into abeyance and now one of the strategies is to write as vaguely and ambiguously as possible, both to make it harder to copy and so you could argue that someone doing something vaguely related or parallel was “infringing”.

Patents only really work in a few places. Pharma is one because the R&D and other up front capex is massive compared to production.

But few other domains even have that issue.

gumby··on Small Inventors Are Being Squeezed by a Convoluted Patent Process
I hate the typical use of NPEs but your point is correct. It’s like private equity companies: in principle quite reasonable but many people use them to extract the value of a business and leave a husk behind.

I don’t know of a way to fix these kinds of problems.

gumby··on Modern decor may be straining people's brains
Today’s style is a callback to the 60s, and we’re people making the same complaints then?
gumby··on Almost Always Unsigned
The "list of undefined behaviour" annex has been written and was discussed at the last meeting. As I said, some of what's in it are already being implemented by the big compilers (though TBF to your comment not yet in any released versions as far as I know).

But this case will be one of the early ones.

gumby··on New York City to ban deceptive subscription practices
These rules are great but “landmark” seems like puffery, as California has had such rules for quite a while.

Ironically that has meant it’s hard to unsubscribe from the New York Times except in California.

gumby··on Successful companies go blind
This is an internal form of the problem described in The Innovator’s Dilemma*: you optimize for what seems to work well and miss the big picture.

I thinkers a natural Hayflick limit for companies.

* you know: one of the 4 or 5 business books actually worth reading, that caused thousands of companies to start cargo-culting the word “innovation”

gumby··on Show HN: Analog Watch
“Halfway to” -> “half”
gumby··on Show HN: Analog Watch
Yes but it’s uncommon in English to simply drop a word from a sentence while pretty common in German casual discourse.

The only other English common case I can think of is the American “I could [not] care less” dropping “not” which is also confusing.

gumby··on Show HN: Analog Watch
This actually makes more sense!
gumby··on How to Write an Email
Best if the subject line is the conclusion and the message supports that.

Subject: feature X dropped from v4.4

Body: we all know this feature is delayed and will cause the release to slip. Marketing gave us the OK to defer it to 4.5

gumby··on No leap second will be introduced at the end of December 2026
The problem is future societies harvesting the engines for interstellar probes. This problem has been discussed in a series of books by Larry Niven.
gumby··on Show HN: Analog Watch
I’m a “quarter past” person but I’ve always been confused by “half ten” (which thankfully isn’t used in Australia). But in German, “half ten” means 9:30, which is make more sense to me (probably because I’m used to how German speech often drops words, which is less common in English)
gumby··on Almost Always Unsigned
> for instance C and C++ leave signed integer wrap undefined

Not in C++29, and I think the big 3 compilers (gcc, clang, MS) will have squashed this long before that version is officially approved.

gumby··on Is The Economist Always Wrong?
Every time I encounter this story I read it again because it’s so great
gumby··on Is The Economist Always Wrong?
His ability to focus and execute has been haphazard. But as you say he is a good writer and was apparently able to force his will upon most of the writers in the stable
gumby··on It seems that the age of reading might be a short anomaly in human history
Programming is even more recent
gumby··on Is The Economist Always Wrong?
If you’re a left leaning voter the Speccie is almost _never_ going to have an article for you.

And the “almost” is simply because often the articles are very well written and a pleasure to read even when the actual topic and argument are deeply deranged. Amazingly this was especially true when Boris was the editor!

gumby··on Acronym Fatigue Series Introduction: why I'm wary of acronyms
My company has a strict NTLA policy.

That’s No Three-Letter Acronyms

Instead we do name things after animals like Lamprey, Remora, Whelk, Axolotl, Tick (the last has not been approved)

gumby··on Should DayQuil Be Legal?
> … similar related drugs like dextromethorphan and ketamine and other NMDA receptor antagonists are innovative drugs to help prevent Alzheimer's.

Should read “NMDA receptor antagonists _may_ give rise to treatments that _may help prevent or ameliorate the symptoms_ of Alzheimer’s.

Nobody even knows how Alzheimer’s works at all — like most diseases it’s a description of some detectable symptoms, some of which could even turn out to be the body defending itself.

Thus compounds that may have a mechanism of action that affects some concomitant, visible symptoms might potentially be useful.

The use of definitive sentences about unknown results is how we end up with wellness and some “biohacking” nonsense.

gumby··on Decline of PhD Admissions Could Imperil a 'Generation of New Talent'
In the physical sciences at least, there are also corporate jobs.
gumby··on Real time map of France's rail network
The German equivalent uses much less bandwidth as it’s simply a static png.
gumby··on Is The Economist Always Wrong?
Odd how? I’ve been a reader since 1985 and am satisfied.
gumby··on Reducing Assumptions, Exploding Your Code
> …agents are predisposed to write extremely defensive, odd-case-handling code… the SNR ratio is very low. You get a spaghetti that is unlikely to crash, but really hard to extract the gist of. And finding more global bugs can be difficult because what should be structural impossibilities can be coerced into silent skips.

Have you tried asking for a summary of the logic being relied on, and asking LLM to simplify the code?

I’m 85% joking, of course, given the current state of automated code generation and the terrible summaries you usually when asking for a summary of text. But in theory that’s what they are supposed to do.

gumby··on California bans 'sell by' labels, hoping to cut food waste
The taxes are collected one way or another, since places need roads and such. OP might find income taxes to be better than property taxes or sales taxes, depending on where they choose and what they get for their money.
gumby··on California bans 'sell by' labels, hoping to cut food waste
> left California for a zero tax state.

Where, pray tell, is this state with zero taxes? Próspera?

gumby··on The End of North America
I think 0 people will say “I may be suffering but at least my countryman is doing fine.”

But I think you’re accidentally arguing a position it doesn’t sound like you support. Macroeconomics has moved into an abstract region where “homo economicus” is their spherical cow, and actual human behaviour is left by the wayside. Straight up utilitarianism is used to calculate benefit. From your comments I believe you consider that bad — and I do too.

But you ignored the part of my comment that addressed that when you wrote:

> His job moving to Ontario does not have these virtues, and no amount of repeating

> > It’s economically superior to integrate these three countries and beneficial overall

> is going to change this.

My wife is taking some medicine that makes her throw up so she takes another one to counteract that. I said that the US chronically fails to attempt the equivalent, instead assuming that humans are fungible equivalents and that the magic hand of the market would take care of things.

Even where attempts are made to ameliorate economic impact they usually involve trying to tech 50 yo coal miners to code. The actual human nature of humans is ignored.

Whet I didn’t like about your comment is that you deny the global (national in this case) benefit — also received by most of the dislocated — for the short term benefit of the few. That’s is just as abstract, and is like arguing that the automobile should have been suppressed because of all the people whose job was to drive horses.

The right hung to do is to address the disadvantaged, as humans. But that attitude (in more than just macroeconomics policy) is considered anti-American and “socialist”

gumby··on The End of North America
> He instead says that the US has more auto jobs because of the cross-border integration than otherwise. The decline of the US auto industry began very soon after the start of the auto pact.

His topic is NA as a large single economy, and on that basis auto production (in both employment and GDP) has grown and productivity has grown since NAFTA was signed (and has gyrated wildly as well due to idiotic auto industry management, which is unrelated to NAFTA).

The jobs had indeed moved around to the detriment of regions in the US, Canada and to a lesser extent Mexico. But if you look at the US alone, you see that process happening as well (e.g. aircraft and auto mfg moving to the non-unionized south). It’s economically superior to integrate these three countries and beneficial overall: but the US has been the laggard in helping those regions who have suffered.

NAFTA is also a condemning example for the EU: a free trade region based on propinquity where different languages and systems are used yet the free trade works.

gumby··on The End of North America
> Also as a Canadian I suspect the new agreement most likely will be even less beneficial to Canadians.

I think Carney is unlikely to let that happen.

Australia’s most recent FTA with the US was lopsided because AUS still had a policy of hoping the US would come to her aid if necessary and so placated the giant, and the GWB administration took advantage of that.

These days that’s unlikely to happen because people understand that that era is over and the current government has more spine.

gumby··on Factories are just rooms
> When I was a kid I was amazed that people like Edison and Tesla had labs not filled with gear from some lab manufacturer, but stuff they built themselves from first principles.

Most chem labs re still like this — ours certainly is. Certain instruments you might buy (gas chromatograph, micrometer) but it’s hardly unusual to find them adapted as well. There’s lots of blue tape in the fume hood too.

And our prototypes are all hand made because it’s simpler. Once we’ve worked out the physics of the device (with lots of tweaks along the way) we can then start to think about DFM. You’ve seen a picture of one of those prototypes: the boards were sent out but mostly we did the welding, bending, assembly, firmware etc all ourselves.

And as a startup a small team can move faster and more flexibly than a bigger organization when making these kinds of things.

gumby··on Valve New Employee Handbook (2012) [pdf]
The reasulting reality of the managerless approach hasn’t been good. As the they say, “if you don’t have any managers you have politics”.

I have several friends who used to work at Valve none of them hate the place, they still have friends there, etc. But they tell similar stories as to why things that normal companies do successfully are impossible at Valve. Perhaps it’s best summed up by something one friend said about her year and a half at Valve: “I first learned who my boss was on the day she fired me.”

Google tried this, notoriously dense grating and then firing basically all the managers at an all-hands. That didn’t work out well at all... And now they have over-steered in the opposite direction!

← PreviousPage 12 of 34Next →