Tag: WebDev

I am building a Python service using asyncio and I am trying to keep my code fully async. However, my real workload includes CPU-heavy tasks like parsing large files and data transformations, plus some third-party libraries that are blocking and do not provide async APIs. I know I can use loop.run_in_executor or asyncio.to_thread for blocking operations, and I could also use multiprocessing. But in practice, how do experienced developers decide which parts should stay fully async, which should use a thread pool, and when it is better to isolate work into a separate process or service? I am looking for practical decision rules and patterns that work in production, not just toy examples. What are the tradeoffs?

I have this function: bool TmrwRenderImpl::copyTexture( std::optional<std::pair<std::unique_ptr<tmrw::opengl::TmrwPassRenderer>, std::unique_ptr<tmrw::opengl::TmrwPassRenderer>>>& pass_copier, const opengl::TmrwTexture& src_texture, const opengl::TmrwTexture& dst_texture, bool flip) { tmrw::opengl::TmrwPassRenderer* renderer = flip ? pass_copier->second.get() : pass_copier->first.get(); std::string uniform_texture = "src_texture"; //std::span<const std::string&> named_uniform_values = { uniform_texture }; return renderer->Render({ std::cref(src_texture) }, { std::cref(dst_texture) })); //, named_uniform_values)); } Here renderer->Render should accepc std::span object references, I tried to declare it in many ways: 1. virtual bool Render(std::span<std::reference_wrapper<const TmrwTexture>> src_texture, std::span<

I’m planning to add custom block components to Tiptap—specifically callouts (info, success, warning, tip), and later cards and accordions. Since Markdown doesn’t natively support these kinds of blocks, I’m looking for guidance on the cleanest long-term approach. Overview of what I’m trying to do Use Tiptap to author content with callouts and other custom block components Persist the content as Markdown (for portability and static rendering) Render the same Markdown consistently on hosted/static sites Current idea (high level) At a conceptual level, I’m considering representing callouts in Markdown using a custom, HTML-like wrapper, while keeping the inner content as standard Markdown. For example: Markdown content inside *italic* and **bold** On the rendering side, this would be transformed into a styled HTML block (e.g. a with classes), while ensuring the inner Markdown remains untouched and renders normally.

I'm recently working on a project that has an ESP32 as the client and Python FastAPI as the server. After the connection, there's an authentication process, which requires them to communicate using WebSocket (messages are in JSON format). The problem is that the ESP32 said it sent out the message: {"type":"auth_response","id":"CHY","data":{"hash":"f6bb6993e5a44bcde621d6bdd4a3dcb7fcfa5ac0b9a8137d88bed81510849fed","time":19726}} but the server only received {. The server didn't receive the whole message sent from the ESP32. I was expecting the server to receive the whole message, but it didn't. I've used Postman to test the server, and it's working fine, so I suspect the problem lies in the ESP32. This is the code for ESP32. void webSocketEvent(WStype_t type, uint8_t *payload

I need to bypass Iran's internet censorship to send emails using R's emayili package. The standard gmail() function fails because the regime blocks the default SMTP protocol. My goal: Use local HTTP/HTTPS/SOCKS proxies to route emayili's email traffic through a local proxy server (192.168.1.50:8080). Current setup that doesn't work: message_content <- "something" smtp <- gmail( username = sender_email, password = sender_email_pass # Gmail app password ) smtp(message_content) # Fails due to blocking What I've tried without success: # Direct proxy specification smtp <- gmail( username = sender_email, password = sender_email_pass, proxy = "https://192.168.1.50", proxyport = 8080 ) # Colon-separated format smtp

My context I have implemented an equals(Object o) method for a speed object (Vitesse here) that when it faces a comparand being in m/s when it is itself in km/h (in example) puts the comparand in km/h too before comparing the values. And I'm testing it in this test where it is triggered by the last assertEquals(...) statement: @Test @DisplayName("Ex E1, Benson Ch. 3") void ex_Benson_Ch3_E1() { LOGGER.info("Asafa Powell (Jamaïque) court 100m en 9.74s. Quelle est sa vitesse moyenne ?\nSerait-il en infraction dans une zone scolaire où la vitesse est limitée à 30 km/h ?"); Vitesse vitesseMoyenne = new Vitesse(100, 9.74); // Par défaut, en mètres secondes assertEquals(10.3, vitesseMoyenne.vitesse(), 0.1, "La vitesse moyenne du coureur n'est pas la bonne, en m/s"); // Vérifier les conversions, étape

I don't know how to convert Java string numbers to insert them into MySQL table. The field 'peso' in MySQL is declared as float. public void Conferma() { // Creo un oggetto Prodotto Prodotti prod = new Prodotti(txtCod.getText(), txtDescr.getText(), txtDicAdr.getText(), txtUdm.getText(), Float.parseFloat(txtPeso.getText())); // inserisce il record nel database try (Connection conn = DriverManager.getConnection("jdbc:mysql://localhost:3306/DDTFornitori", "root", "mA19720306#")) { // Crea l'istruzione PreparedStatement insStmt = conn .prepareStatement("insert into tblProd (idprod, descr, dicituraADR, udm, peso) values(?,?,?,?,?)"); // Specifico il valore dei parametri insStmt.setString(1, prod.getCodice()); insStmt.setString(2, prod.getDescrizione()); insStmt.setString(3, pro

I'm trying to use React Testing Library with Jest and I'm not sure how to address it. I'm trying to test React components with Jest and RTL and I keep getting caught with erasableSyntaxOnly errors. Here are my configurations: package.json: { "name": "06-jest-setup", "private": true, "version": "0.0.0", "type": "module", "scripts": { "dev": "vite", "build": "tsc -b && vite build", "lint": "eslint .", "preview": "vite preview", "test": "jest" }, "dependencies": { "react": "^19.2.0", "react-dom": "^19.2.0" }, "devDependencies": { "@babel/core": "^7.29.0", "@babel/preset-env": "^7.29

I have the following R code that makes a visualization using ggplot: library(ggplot2) set.seed(42) n <- 100 time <- 1:n baseline <- 10 + rnorm(30, sd = 2) intervention <- 10 + seq(0, 5, length.out = 40) + rnorm(40, sd = 2) post <- 15 + rnorm(30, sd = 2) values <- c(baseline, intervention, post) df <- data.frame(time = time, value = values) intervention_start <- 31 intervention_end <- 70 shaded_region <- data.frame( xmin = intervention_start, xmax = intervention_end, ymin = -Inf, ymax = Inf, Period = "Intervention Period" ) vlines <- data.frame( xintercept = c(intervention_start, intervention_end), Line = "Intervention Start/End" ) ggplot(df, aes(x = time, y = value)) + geom_rect(data = shaded_region, aes(xmin = xmin, xmax = xmax, ymin = ymin, ymax = ymax, fill = Period), inherit.aes = FALSE, alpha = 0.3) + geom_line(aes(color = "Observed Data"), linewidth = 0

I was fiddling with the Brainfuck esolang over the past few days, and tried to implement an increment operation on an N-byte-wide integer in big-endian format. Note that I am imposing a structural constraint on the data. Consider BF tape on the left and right side of this int to be infinite and all initialized to 0. The exact integer value is not known and the data pointer starts at the LSB. The very important additional constraints are that the bytes other than the N integer bytes must be reset before end and the pointer must end at LSB. I have managed to establish 2 algorithms which generate O(N) instructions for this increment, using ripple carry, which are structurally very different. Unrolled: We hardcode the int length, by opening N - 1 loops which after increment, check if the current position is zero (overflow occurred) and if then so, move forward, which we then close at the exact same position it was opened and from there we m