HNHacker News
TopNewBestAskShowJobs

pierrebai

1,310 karma · joined October 13, 2011

submissionscomments
pierrebai··on Neal Stephenson was prescient about our AI age
So, the fantastical plot argument is that one can insert a singular hard-to-verify subtle error among a see of factually correct statements? That's not novel, it is a fact on about any decently sized compendium, \that there will be a typo, an out of date fact or error. That's supposed to be scary?

One of the underlying assumption behind these sci-fi ideas is that people believe weird things because of carefully constructed misleading chain of reasoning, when in fact they simply believe in things that are convenient to support their general world view. Climate denial, for example, is not about carefully planted false evidences, it is about adhering to a view point that avoids forcing you to reconsider your lifestyle.

pierrebai··on We switched to Java 21 virtual threads and got a deadlock in TPC-C for Postgres
Probably the author thought that this typical AI-generation quirks were a funny wink to the concept of virtual threads. The elephants, of course, represent Postgres.
pierrebai··on We switched to Java 21 virtual threads and got a deadlock in TPC-C for Postgres
I thought the blog was great but "in summary" conclusion bad.

The summary merely stated that Java virtual thread are great. I expected a summary of the problem and solution, for example something like:

When using Java 21 virtual threads, you can end-up starved of carrier threads due to all carrier threads waiting on a pool exhausted resources with no thread available to free such resources. The solution is to wrap those resources in a virtual-thread aware object. In our case, we solved our problem by wrapping connections in semaphores.

pierrebai··on Python 3.13 Gets a JIT
The copy-and-patch also assumes the compiler will generate patchable code. For example, on some architecture, have a zero operand might have a smaller or different opcode compared to a more general operand. Same issue for relative jumps or offset ranges. It seems the main difference is that the patch approach also patches jumps to absolute addresses instead of requiring instruction-counter relative code.
pierrebai··on Scheme modules vs. whole-program compilation: fight
Worked on a similar problem (when looked at from a very high level), and the solution was to alternatively invoke each algorithm until everything is done. That is, do as much as possible in one form, then do as much in the other form, rinse and repeat.
pierrebai··on Liero – Sling'n'shoot Worms Game
Looks like the classic Soldat [1], except uglier, only 1v1 and less interesting maps...

[1] https://store.steampowered.com/app/638490/Soldat/

pierrebai··on IDEs we had 30 years ago
The fact that it is easy to confuse the server with the client in X, it does not change the fact that the XServer and XEmacs are running on different computers, so each interaction is a round-trip.
pierrebai··on Stern-Gerlach experiment used to probe the clash of quantum theory and gravity
But the two lines they get on the pin-head-sized region is physical and has a particular physical orientation? How can the spin not be a physical orientation and yet the experiment result is all about the physical location of the lines?

I'm sure the issues are with the details of how the experiment is explained, but I still don't understand.

pierrebai··on Stern-Gerlach experiment used to probe the clash of quantum theory and gravity
I don't understand the experiment: even if there are only two orientation for the spin, the atom, itself, can be in any orientation?
pierrebai··on Unified versus Split Diff
Your examples don't explain how both side can possibly be aligned when the differences are large. If they don't align, then the two view are independent and it's hard to see the correspondence. If they do align then... I'm not sure what is the difference between your suggestion and split diff, except th eleft side has less highlights.

So, as I understand it, you are advocating split diff with SUBTLER grey highlights in the current version of the file?

pierrebai··on Unified versus Split Diff
"everyone seems to be happy reviewing diffs, rather than the actual code?"

The split view does show the actual new code, just with extra empty lines... so sure it is a bit distracting if a lot of lines have been modified, but it rarely comes that far.

My beef is more with some team's habit of doing too much work in a single pull request. Large diff are just a symptoms of that. I find that it is much more likely to introduce bugs and hard-to-find changes in behaviour in large changes. If you later find that something broke, finding what caused it in a large PR can be a real time waster.

pierrebai··on The Paperweight Dilemma
How don't grasp why megi would close their github account plus prevent pull. Seems like not only a very unsympathetic move (to be very euphemistic) but quite against the spirit of open source. Did they write the entirety of the Linux kernel? If not, then whatever their contribution might be pales in comparison to what they're building upon. To treat their own work in such a egoistical manner seems out of touch.
pierrebai··on Making Hard Things Easy
Well, I'm going to disagree with her solution to bash. Bash is terrible, end of story. Just the *.txt is actually an endless pit of trap, badness and wrong behaviour.

I've not written a bash script in years. I always use Python and in my experience professionally, most people automating things nowaday use Python. I only use bash to write one-line to invoke the underlying Python script. For the record, I do the same for batch file on Windows, with a simply .bat file to invoke the underlying Python script.

Trying to improve the bash experience with tools and knowledge is throwing bad time at a problem that has a well-known solution.

Edit: to be clear, the presentation itself is awesome. I'm just disagreeing that bash needs to be better explained. I'm ready to admit that there are existing scripts out there, most of which are probably only working in the trivial, normal case and are just one snag away from exploding, and having a known source of help can help. But please, take you're bash scrip to the shed and upgrade them to Python.

pierrebai··on Linear code is more readable
The linear version is hard to test. The split-function one is much more testable. There is also that thing called complexity, which increase with function length and has been proved to correlate with bugs count.

The problem with the example is that it is both extremely artificial and shows a single use case. Even with the artificiality, one can easily imagine baking a calzone instead, which could reuse all the factored-out oven functions in the split version.

(The comment about pizza vs baked pizza is one about using typing to encode your logic, but is separate from the issue that your functions should do one thing.)

pierrebai··on Cops still take more stuff from people than burglars do (2021)
Calling it civil forfeiture is in itself problematic. The assumption is that the person committed a crime and profited from it. Kinda of a weird mix. You must prove you did not commit a crime in a civil case.
pierrebai··on Curl Command Line Variables
I really feel this is reinventing the wheel and the resulting wheel is not quite round.

Would it not have been both simpler and more powerful to have a .curlrcscript file that gets executed and the output used as the config contents?

Even the original use case looks somewhat fishy: the user has secrets in their environment, but they can't have a bash script or function wrapper around curl to call curl with the environment variable baked? Again, more general and powerful than a limited variable-replacement scheme using yet-another curl-specific syntax.

I can think of many examples where the direction curl went will backfire: people will want variable based on other variables. People will want variable based on conditions (if/else), etc. All things a general scripting solution would fulfill in a standard way.

pierrebai··on Coroutines for Go
The examples given prompt me to say: if all you have is Rube-Goldberg hammer, everything looks like an Escheresque nail.

Sieving primes by turning functions into coroutines, parsing text by yielding characters, all with unnatural functions and state management... that;s an improvement over what?

pierrebai··on Writing prettier Haskell with Unicode syntax and Vim
I think the fact that the author somehow thinks that Greek letters are prettier and easier to understand hints at an enormous bias.

Many people would not know their names, most people are probably unfamiliar to Greek letters beyond beta or gamma, with some exceptions for pi, which has the dubious property of being confusingly ambiguous if the constant is meant or a simple high-brow variation on a variable named p.

It's my conviction that there are a lot of things in computing that are made harder than they necessarily need to be by bad naming and chief amongst them is the insistence on borrowing mathematical notations and jargon.

In my mind, it is mostly a question of elitism and exclusion.

pierrebai··on It’s not wrong that "🤦🏼‍♂️".length == 7 (2019)
... but maybe it simplifies and speed-up the internal processing? I haven't looked at Python 3 C implementation of strings, but that is a guess. Also, IIRC, Python 3 has the ability to keep different internal representations of strings and uses the most compact one. If all character are 7-bit ASCII, it uses bytes representation. That's what I remember from Python dev discussions long ago.

But the overall tone of the article is still bashful. Caring about the internal representation of strings and bashing UTF-32 feels lame and angry. (Especially if I'm right about the multi-rep nature of Python 3: their choice is good for most text, and they could add a UTF-8 internal rep in the future, although that would probably break enough code that expect the UTF-32 value for len() that it is not worth it.)

pierrebai··on What are the enduring innovations of Lisp? (2022)
I've read the same, almost word-for-word claims repeatedly, which is annoying given how misleading it is.

Concretely...

To know the context of the conditions, the conditions must give the information. Otherwise, the handler would need to know intimately the implementation details to be able to retrieve the filename, line number, etc. If you can provide the information to the condition you call fill a throw exception with the exact same data. The exception can carry the filename, line numbers, token being parsed...

Second, the example itself is ludicrous. The caller of a file parser providing replacement data for a malformed file? In what world does that ever happens? How could it handle every possible ways a file might be malformed?

Third, in every language, the same can be implemented with a callback. In C++, the standard is now to use std::function for this, which supported free functions, members, lambdas... pretty much everything. The only advantage of List is that the declaration and registration of the callback is a language feature.

pierrebai··on I didn’t learn Unix by reading all the manpages (2022)
I learned Unix through its manpage. I learned abotu sh, awk, grep, I learned about sockets and tcp, I learned about semaphore and shared memory. When I started working on Unix, in the early 90s', it was on a true-to-form physical VT220. I quickly learned to run make in the background so I could read the manpage while compiling.
pierrebai··on H. G. Wells interviewing J. V. Stalin (1934)
It is interesting to remind ourselves how intellectual were disconnected with the reality of their time. All theoretical debates without ground in how political, economic power really worked.
pierrebai··on All you need is data and functions
I disagree with the post. The goal of traits, or any similar mechanism in other languages, is not to implement a given functionality. It is to describe and enforce a common way and API to expose the functionality.

To take the example given for gleam, there is nothing enforcing that the conversion to string is called to_string, what are its arguments, how formatting works. So you're left to just-so compatibility for each type. They only happen to be the same by accident, and nothing enforces consistency.

Programming language and code in general is about communication between programmers. When I design a trait, or an interface, I'm telling everyone else what are the requirements.

pierrebai··on F2-NeRF: Fast Neural Radiance Field Training with Free Camera Trajectories
IANAL, but this seems easily circumvented by setting up an independent entity. (It's not an affiliate if it's not owned by the competitor.)
pierrebai··on For the first time, the Fed is losing money
That wiki graph does not show what you say. There are multiple place in the curve that don't follow your theory. And by places, I mean for example the stretch from 1980 to 2000.
pierrebai··on For the first time, the Fed is losing money
Or you simply are inverting cause and effect. Or at least cause and multiple independent effects. When the inflation goes up, both interest rates and unemployment go up.

It's akin to saying that coughing causes lung cancer. Smoking causes coughing and lung cancer. One cause, multiple effects.

pierrebai··on New ATLAS result weighs in on the W boson
This is ludicrous, but this kind of thing have been going on in science and physics in particular for a long time. Let's call the phenomenon maths side-blinders.

One experiment finds the mass of thw W boson to be 80370 +/- 19 MeV, another 80434 +/-9 MeV. Clearly, both results are incompatible, their rnage don't overlap. Of course, these are statistical ranges. But even with 95% certainty, their differences is many times the uncertainty, so it's not just that they're a bit off. IOW, we can be 100% (not 95%) sure that at least one, if not both, are incorrect.

Yet they are boldly reported with those uncertainty ranges, even though, clearly , those ranges cannot be correct. And then ATLAS double down by "apply more statistical analysis" to narrow their uncertainty range!

There should not be work on "improved stats analysis", but more work on finding where the systemic error between the two experiment lies. I truly don't see the point of retreading the same data set to change the value and uncertainty range when clearly there is something wrong with the data, the science, the experiment or all of the above.

PS: what I'd like to see, is the labs to say something along the line: given result A and B being incompatible, statistically there is a (say) 99.99% chance that one or both experiment has a hidden flaw or that there is a major flaw in the standard model.

pierrebai··on Go claims telemetry objectors arguing in bad faith and violating Code of Conduct
Too bad the linked article arguments are of poor quality and unconvincing.

There was a bug on MacOS that ended-up requiring XCode to be install in order for the Go compiler to work? That example is bad because:

1. it was going to be very obvious real quick 2. it did get discovered and fixed without telemetry.

... and both of my points above are irrelevant. Why? Because the problem raised here is not if telemetry is useful or not. It's not even if telemetry is good or bad. It's that discussions about telemetry silenced people who were against it.

Your point is that people should not be allowed to raise arguments that have not been approved as valid. I mean, the simplest reformulation is that "if we think you're wrong you will not be allowed to speak."

For real? Being wrong means you're not allowed to participate? (And by wrong, I mean "deemed wrong by the people in power to ban others.")

pierrebai··on Weakening TLS protection, South Korean style
No.

What is needed is less centralization, not more. I'll take the US society as an example: having congress, senate, president, courts, military, police, states, free press (with multiple participants), university etc means that to compromise democracy, you need to subvert many elements simultaneously.

Likewise, having DNS and PKI separate is a plus, as the cert is linked to DNS and finding a web server is done through DNS. To completely fool an attentive user, you would need to both fool their DNS and their pool of trusted certificate (or a single CA).

To increase security, we need more cross-linked layers that each require the other. We need to have more steps and more elements that a crooked player needs to control simultaneously in order to achieve an exploit.

For example, we could require multiple certs, so that you'd need to compromise multiple CA. Although, if you can add one CA to the trusted pool, then you can add 3, so this does not protect against actors that have access to the user's OS. But, on that front, if you got root access to a user's OS, you can do anything. Once you MITM them, you can make a user divulge their credentials. Only a physical fob with its own certs, DNS resolver and hardware-level private key can protect against that.

pierrebai··on An AI lawyer was set to argue in court – real lawyers shut it down
You are saying exactly what I'm saying, only with different words. As life goes on, this is a trend I get more and more acutely aware of: people bias, opinion and values make them re-interpret what others say to make it fit in their already made up mind about what the conversation is about.

I was saying that not knowing the law is not a defense yet the laws are so complex that an expensive expert that dedicated all their study to the law is practically (as in, yes, you have the option to not use a lawyer) required in court.

It is not the same for many profession. In no other profession are non-experts expected and required to know the matter.

← PreviousPage 2 of 15Next →