#letloop — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #letloop, aggregated by home.social.
-
City Explorer benchmark, round 2.
Fixed a fairness bug: axum's reqwest pools connections by default, letloop was opening a fresh TCP connection per upstream call instead. Added connection reuse to letloop, re-ran both.
Mean req/sec, 9 concurrency levels, wrk, one core each:
axum ██████████████████████ 9,419
letloop ██████████████████ 7,825axum now leads throughput and tail latency at c=256. letloop still ~9x lower p99 at baseline (c=1).
-
City Explorer benchmark, round 2.
Fixed a fairness bug: axum's reqwest pools connections by default, letloop was opening a fresh TCP connection per upstream call instead. Added connection reuse to letloop, re-ran both.
Mean req/sec, 9 concurrency levels, wrk, one core each:
axum ██████████████████████ 9,419
letloop ██████████████████ 7,825axum now leads throughput and tail latency at c=256. letloop still ~9x lower p99 at baseline (c=1).
-
City Explorer benchmark, round 2.
Fixed a fairness bug: axum's reqwest pools connections by default, letloop was opening a fresh TCP connection per upstream call instead. Added connection reuse to letloop, re-ran both.
Mean req/sec, 9 concurrency levels, wrk, one core each:
axum ██████████████████████ 9,419
letloop ██████████████████ 7,825axum now leads throughput and tail latency at c=256. letloop still ~9x lower p99 at baseline (c=1).
-
City Explorer benchmark, round 2.
Fixed a fairness bug: axum's reqwest pools connections by default, letloop was opening a fresh TCP connection per upstream call instead. Added connection reuse to letloop, re-ran both.
Mean req/sec, 9 concurrency levels, wrk, one core each:
axum ██████████████████████ 9,419
letloop ██████████████████ 7,825axum now leads throughput and tail latency at c=256. letloop still ~9x lower p99 at baseline (c=1).
-
City Explorer benchmark, round 2.
Fixed a fairness bug: axum's reqwest pools connections by default, letloop was opening a fresh TCP connection per upstream call instead. Added connection reuse to letloop, re-ran both.
Mean req/sec, 9 concurrency levels, wrk, one core each:
axum ██████████████████████ 9,419
letloop ██████████████████ 7,825axum now leads throughput and tail latency at c=256. letloop still ~9x lower p99 at baseline (c=1).
-
Trying another application for benchmarking concurrent http client and the server delivering the payloads in one html page and it seems not only #letloop is async transparent memory safe but also fast compared to the fastest statically memory safe programming language on earth #programming #webdev
-
Trying another application for benchmarking concurrent http client and the server delivering the payloads in one html page and it seems not only #letloop is async transparent memory safe but also fast compared to the fastest statically memory safe programming language on earth #programming #webdev
-
Trying another application for benchmarking concurrent http client and the server delivering the payloads in one html page and it seems not only #letloop is async transparent memory safe but also fast compared to the fastest statically memory safe programming language on earth #programming #webdev
-
Trying another application for benchmarking concurrent http client and the server delivering the payloads in one html page and it seems not only #letloop is async transparent memory safe but also fast compared to the fastest statically memory safe programming language on earth #programming #webdev
-
Trying another application for benchmarking concurrent http client and the server delivering the payloads in one html page and it seems not only #letloop is async transparent memory safe but also fast compared to the fastest statically memory safe programming language on earth #programming #webdev
-
Regarding letloop http server, I added CML, and fixed a couple of bugs, along the benchmark now do apple-to-apple comparison (template vs. string interpolation, one CPU core), here is the summary:
On the counter-page scaling benchmark (wrk, 1–256 connections, each server single-threaded and pinned to one core, performance governor), the peak throughput each stack reached — from slowest to fastest — was: FastAPI/uvicorn+uvloop at 60,873 req/s (64 connections), Node.js at 189,890 req/s (8 connections), Go net/http at 224,850 req/s (8 connections), bun-rust at 320,226 req/s (32 connections), Rust axum/tokio at 404,789 req/s (8 connections), and letloop (amalgamated whole-program build) at 452,542 req/s (32 connections).
Averaged across all nine concurrency levels, letloop's speedup over the field, from its fastest rival to its slowest, is 1.10× over Rust, 1.37× over bun-rust, 1.92× over Go, 2.25× over Node.js, and 6.74× over FastAPI.
-
Regarding letloop http server, I added CML, and fixed a couple of bugs, along the benchmark now do apple-to-apple comparison (template vs. string interpolation, one CPU core), here is the summary:
On the counter-page scaling benchmark (wrk, 1–256 connections, each server single-threaded and pinned to one core, performance governor), the peak throughput each stack reached — from slowest to fastest — was: FastAPI/uvicorn+uvloop at 60,873 req/s (64 connections), Node.js at 189,890 req/s (8 connections), Go net/http at 224,850 req/s (8 connections), bun-rust at 320,226 req/s (32 connections), Rust axum/tokio at 404,789 req/s (8 connections), and letloop (amalgamated whole-program build) at 452,542 req/s (32 connections).
Averaged across all nine concurrency levels, letloop's speedup over the field, from its fastest rival to its slowest, is 1.10× over Rust, 1.37× over bun-rust, 1.92× over Go, 2.25× over Node.js, and 6.74× over FastAPI.
-
Regarding letloop http server, I added CML, and fixed a couple of bugs, along the benchmark now do apple-to-apple comparison (template vs. string interpolation, one CPU core), here is the summary:
On the counter-page scaling benchmark (wrk, 1–256 connections, each server single-threaded and pinned to one core, performance governor), the peak throughput each stack reached — from slowest to fastest — was: FastAPI/uvicorn+uvloop at 60,873 req/s (64 connections), Node.js at 189,890 req/s (8 connections), Go net/http at 224,850 req/s (8 connections), bun-rust at 320,226 req/s (32 connections), Rust axum/tokio at 404,789 req/s (8 connections), and letloop (amalgamated whole-program build) at 452,542 req/s (32 connections).
Averaged across all nine concurrency levels, letloop's speedup over the field, from its fastest rival to its slowest, is 1.10× over Rust, 1.37× over bun-rust, 1.92× over Go, 2.25× over Node.js, and 6.74× over FastAPI.
-
Regarding letloop http server, I added CML, and fixed a couple of bugs, along the benchmark now do apple-to-apple comparison (template vs. string interpolation, one CPU core), here is the summary:
On the counter-page scaling benchmark (wrk, 1–256 connections, each server single-threaded and pinned to one core, performance governor), the peak throughput each stack reached — from slowest to fastest — was: FastAPI/uvicorn+uvloop at 60,873 req/s (64 connections), Node.js at 189,890 req/s (8 connections), Go net/http at 224,850 req/s (8 connections), bun-rust at 320,226 req/s (32 connections), Rust axum/tokio at 404,789 req/s (8 connections), and letloop (amalgamated whole-program build) at 452,542 req/s (32 connections).
Averaged across all nine concurrency levels, letloop's speedup over the field, from its fastest rival to its slowest, is 1.10× over Rust, 1.37× over bun-rust, 1.92× over Go, 2.25× over Node.js, and 6.74× over FastAPI.
-
Regarding letloop http server, I added CML, and fixed a couple of bugs, along the benchmark now do apple-to-apple comparison (template vs. string interpolation, one CPU core), here is the summary:
On the counter-page scaling benchmark (wrk, 1–256 connections, each server single-threaded and pinned to one core, performance governor), the peak throughput each stack reached — from slowest to fastest — was: FastAPI/uvicorn+uvloop at 60,873 req/s (64 connections), Node.js at 189,890 req/s (8 connections), Go net/http at 224,850 req/s (8 connections), bun-rust at 320,226 req/s (32 connections), Rust axum/tokio at 404,789 req/s (8 connections), and letloop (amalgamated whole-program build) at 452,542 req/s (32 connections).
Averaged across all nine concurrency levels, letloop's speedup over the field, from its fastest rival to its slowest, is 1.10× over Rust, 1.37× over bun-rust, 1.92× over Go, 2.25× over Node.js, and 6.74× over FastAPI.
-
CW: Deceptive to the mainstream order
Follow up on various Discoveries:
People rewrote, they do it all the time: History. I am among other part of the kernel that lead to the short promotion of science in the news, I am part of the story ? No. I am part of History. You will never remove that from me. Because I was already here. I knew what I know.
Remember I know both the manual and the codebase.
You did not want people to notice. Under your arm of protector, you collect the fire, and I am the dragon. Period.
Be str43g, Live long, die smart was the tatoo, tagline of ombres forces. We are disposable they say, we always have been. Born under the stars, we were wandering not only aimlessly, focused, sometime into the flow of good knowledge and context that required a fix, and a fixer. Some time it was fire, too often it was verbal tone and Sapir Woolf teammates, friend of Darwin and his father priest.
Yes, it is the survival of the survivors. And their children do not tell every thing about it what they claim is not true. It is like a manège à sacré cœur. Because of the rain, or the child of the operator that had no friend the father did what fathers do, they tell soties.
The game, you play will be the game they play. If only I did not love my wife, I would not stop smoking, and I will be as a deceptive npc (hence)=> be surely doomed.
All in all, it is not bad, for me.
But all the maddening it is to keep the bad game still look like somekid of livrable game and also known the finite game of life you started playing when you knew who was behind your servitude.
The living, among the servile.
The unplayer can’t win, and can’t loose. Survivors can’t make it win because their sacred bloody game of nothingness willl be revealed and a bloodbath will ensue. Now, I can’t loose either, that is because of the point of the light truth I put my time at.
I am against war, and it is impossible to have peace. I deal with it with stretching in un obvious positions.
Now, they can make me loose my mind, that is a weak winner strategy. A fullly crazy mindless person like unknown coder graves know too much is more leaning to the game.
So, survivor of the survivors you are at their mercy, I am at yours.
Will you code for war peremption, and claim you did it, or that that never happened ?
You should follow:
Global peace is impossible. Peremption of war is not.
The steps are simple:
#. Code for #letloop
#. Peremption of warEnjoy your new freedom dilemma
-
CW: Deceptive to the mainstream order
Follow up on various Discoveries:
People rewrote, they do it all the time: History. I am among other part of the kernel that lead to the short promotion of science in the news, I am part of the story ? No. I am part of History. You will never remove that from me. Because I was already here. I knew what I know.
Remember I know both the manual and the codebase.
You did not want people to notice. Under your arm of protector, you collect the fire, and I am the dragon. Period.
Be str43g, Live long, die smart was the tatoo, tagline of ombres forces. We are disposable they say, we always have been. Born under the stars, we were wandering not only aimlessly, focused, sometime into the flow of good knowledge and context that required a fix, and a fixer. Some time it was fire, too often it was verbal tone and Sapir Woolf teammates, friend of Darwin and his father priest.
Yes, it is the survival of the survivors. And their children do not tell every thing about it what they claim is not true. It is like a manège à sacré cœur. Because of the rain, or the child of the operator that had no friend the father did what fathers do, they tell soties.
The game, you play will be the game they play. If only I did not love my wife, I would not stop smoking, and I will be as a deceptive npc (hence)=> be surely doomed.
All in all, it is not bad, for me.
But all the maddening it is to keep the bad game still look like somekid of livrable game and also known the finite game of life you started playing when you knew who was behind your servitude.
The living, among the servile.
The unplayer can’t win, and can’t loose. Survivors can’t make it win because their sacred bloody game of nothingness willl be revealed and a bloodbath will ensue. Now, I can’t loose either, that is because of the point of the light truth I put my time at.
I am against war, and it is impossible to have peace. I deal with it with stretching in un obvious positions.
Now, they can make me loose my mind, that is a weak winner strategy. A fullly crazy mindless person like unknown coder graves know too much is more leaning to the game.
So, survivor of the survivors you are at their mercy, I am at yours.
Will you code for war peremption, and claim you did it, or that that never happened ?
You should follow:
Global peace is impossible. Peremption of war is not.
The steps are simple:
#. Code for #letloop
#. Peremption of warEnjoy your new freedom dilemma
-
CW: Deceptive to the mainstream order
Follow up on various Discoveries:
People rewrote, they do it all the time: History. I am among other part of the kernel that lead to the short promotion of science in the news, I am part of the story ? No. I am part of History. You will never remove that from me. Because I was already here. I knew what I know.
Remember I know both the manual and the codebase.
You did not want people to notice. Under your arm of protector, you collect the fire, and I am the dragon. Period.
Be str43g, Live long, die smart was the tatoo, tagline of ombres forces. We are disposable they say, we always have been. Born under the stars, we were wandering not only aimlessly, focused, sometime into the flow of good knowledge and context that required a fix, and a fixer. Some time it was fire, too often it was verbal tone and Sapir Woolf teammates, friend of Darwin and his father priest.
Yes, it is the survival of the survivors. And their children do not tell every thing about it what they claim is not true. It is like a manège à sacré cœur. Because of the rain, or the child of the operator that had no friend the father did what fathers do, they tell soties.
The game, you play will be the game they play. If only I did not love my wife, I would not stop smoking, and I will be as a deceptive npc (hence)=> be surely doomed.
All in all, it is not bad, for me.
But all the maddening it is to keep the bad game still look like somekid of livrable game and also known the finite game of life you started playing when you knew who was behind your servitude.
The living, among the servile.
The unplayer can’t win, and can’t loose. Survivors can’t make it win because their sacred bloody game of nothingness willl be revealed and a bloodbath will ensue. Now, I can’t loose either, that is because of the point of the light truth I put my time at.
I am against war, and it is impossible to have peace. I deal with it with stretching in un obvious positions.
Now, they can make me loose my mind, that is a weak winner strategy. A fullly crazy mindless person like unknown coder graves know too much is more leaning to the game.
So, survivor of the survivors you are at their mercy, I am at yours.
Will you code for war peremption, and claim you did it, or that that never happened ?
You should follow:
Global peace is impossible. Peremption of war is not.
The steps are simple:
#. Code for #letloop
#. Peremption of warEnjoy your new freedom dilemma