Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
BenoitEssiambre
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
BenoitEssiambre
1y ago
To add to this, there are fundamental information-theoretic principles that support inlining code and components. It's about reducing code entropy, reducing length and referential distances. https://benoitessiambre.com/
32.
▲
by
BenoitEssiambre
1y ago
I'll agree that the network layer adds more problems to microservices, but even with a perfect network, they are problematic. Everything being out of sync, (if they are stateful microservices which queues imply), is one big issue. Thin
33.
▲
by
BenoitEssiambre
1y ago
To add to this. There's fundamental theoretical reasons why microservices or bad. They increase the entropy of code ( https://benoitessiambre.com/entropy.html ) by increasing globally scoped dependencies. They are the gl
34.
▲
by
BenoitEssiambre
1y ago
Asynchronous queues make your data out of sync (hence the name) and inconsistent one of the main downsides of microservices. Their use should be minimized to cases where they are really necessary. A functional transactional layer like postg
35.
▲
by
BenoitEssiambre
1y ago
Exactly. I think of them as Markov Chains in grammar space or in Abstract Syntax Tree space instead of n-gram chain-of-words space. The attention mechanism likely plays a role in identifying the parent in the grammar tree or identifying oth
36.
▲
by
BenoitEssiambre
1y ago
I love this stuff. This 256 byte video clip is a great example using similar principles: https://www.youtube.com/watch?v=sWblpsLZ-O8
37.
▲
by
BenoitEssiambre
1y ago
Lit is amazing, one of the most concise framework out there. For deepening my understanding of concepts around Lit, I once built LittleLit, a 37 lines of code framework around Lit-html that helps understand how simple and elegant the moving
38.
▲
by
BenoitEssiambre
1y ago
This is good advice but it can be difficult to define what simple means. The only technical way I was able to make sense of it is by targeting reducing code entropy and scopes (Inspired by how language models try to minimize Solomonoff/
39.
▲
by
BenoitEssiambre
1y ago
Though maybe avoid USD or USD bonds while the gov is trying to get rid of Fed independence.
40.
▲
by
BenoitEssiambre
1y ago
I feel this article misses the main reason (unless it falls under "Build a better wheel") which is to build a wheel that is tailored and stays tailored to your purposes. How often do I see people metaphorically trying to use a car
41.
▲
by
BenoitEssiambre
1y ago
Solomonoff induction says that the shortest program that can simulate something is its best explanatory theory. OpenAI researchers very much seem to be trying to do theory building ( https://x.com/bessiambre/status/
42.
▲
Evaluate software and code like AI models do it internally
(benoitessiambre.com)
2 points
by
BenoitEssiambre
1y ago
|
0 comments
43.
▲
by
BenoitEssiambre
2y ago
Here's a somewhat related post where I argue that logic that _can_ have small pointers has lower entropy and is more optimal as bayesian model of the domain: https://benoitessiambre.com/entropy.html
44.
▲
by
BenoitEssiambre
2y ago
These mostly align with the idea of reducing code entropy: https://benoitessiambre.com/entropy.html
45.
▲
by
BenoitEssiambre
2y ago
Ah that's an interesting tidbit. I have a Wealthsimple account through my (YC company) employer. The Desmarais (Power corp folks) gave me a scholarship back in the day. I hadn't made the connection.
46.
▲
by
BenoitEssiambre
2y ago
Yeah there's definitely that aspect too. Though nowadays, led lighting is so efficient that you can get a lot of brightness on reasonable wattage. That's not enough for the light to feel natural though, you also have to have the p
47.
▲
by
BenoitEssiambre
2y ago
What I'm trying to achieve is lighting a bit like in these "Artificial Skylight" concepts: https://architizer.com/blog/practice/materials/let-there-be-... They claim their technology is about
48.
▲
by
BenoitEssiambre
2y ago
I'm in a basement and had fairly good success with certain "led shop lights" to make the space feel more like it has natural light (though I could improve things further). I do find however, that using diffuse lighting and mi
49.
▲
by
BenoitEssiambre
2y ago
Yup, I hear the top candidates to replace him are Chrystia Freeland or Marc Carney (maybe known to this community since he's on Stripe's board). These top candidates might not accept to play this round though as the chances of win
50.
▲
by
BenoitEssiambre
2y ago
Indeed and breaking out logic into more global scopes has serious downsides if that logic needs to be modified in the future, if your system still needs to support innovation and improvements, downsides not totally unlike the downsides of u
51.
▲
by
BenoitEssiambre
2y ago
"large enough" and "bath" is doing a lot of work here. I don't think it's necessarily about size. It's about that size and configuration implying the particle states are difficult to know. If for example,
52.
▲
by
BenoitEssiambre
2y ago
It is though. Temperature is an aggregate summary statistic used when the observer doesn't know the details of individual particles. If you did know their position, speed and velocities, you could violate the laws of entropy as Maxwell
53.
▲
by
BenoitEssiambre
2y ago
> Fundamentally, entropy depends on the probability distribution, not the observer. Right but a probability distribution represents the uncertainty in an observer so there is no inconsistency here (else you're falling for the Mind P
54.
▲
by
BenoitEssiambre
2y ago
I definitely agree that models are difficult and contextual. I do think that minimizing distances and entropy does reduce the difficulty of doing migrations and refactoring of the data model. Doing migrations across spread out, distributed
55.
▲
by
BenoitEssiambre
2y ago
Also Poet: https://cmrr-star.ucsd.edu/static/shannonsite/pdfs/poster.pd... ( maybe following in the footsteps of James Clerk Maxwell https://allpoetry.com/James-Clerk-Maxwell like other Shan
56.
▲
Leveraging the lightcone around the source of truth with Postgres
(benoitessiambre.com)
9 points
by
BenoitEssiambre
2y ago
|
5 comments
57.
▲
by
BenoitEssiambre
2y ago
Here is a followup post for those interested: https://benoitessiambre.com/integration.html
58.
▲
by
BenoitEssiambre
2y ago
What I'm advocating here is only coincidentally a performance optimization. Readability and maintainability (and improved abstraction) are the primary concern and benefit of (sometimes) keeping things inline or more specifically of red
59.
▲
by
BenoitEssiambre
2y ago
Here are some information theoretic arguments why inlining code is often beneficial: https://benoitessiambre.com/entropy.html In short, it reduces scope of logic. The more logic you have broken out to wider scopes, the more
60.
▲
Leaning into complexity by reducing code entropy
(benoitessiambre.com)
2 points
by
BenoitEssiambre
2y ago
|
0 comments
More ›