Carlos A. Escobar

Hola, sería posible que en en la app de @Bancolombia añadan la posibilidad de visualizar nuevamente un comprobante de transferencia?. Ya q si lo por alguna razón no lo alcanzas a descargar pierdes el comprobante y no es visible en el historial de transferencias de la sucursal web

cc @OVHcloud

@OVHcloud I was charged twice and need help with this, thanks you...

Hoy formalicé ante el Presidente @petrogustavo nuestra solicitud de tener una mayor participación accionaria en las empresas públicas generadoras de energía, Urrá y Gecelca. Córdoba merece recibir un verdadero beneficio por la explotación de sus riquezas naturales.

@nacordr7 Gracias concejal, por defender su comunidad.

🦾 #Caribe #Nomasabusos

La @DIANColombia se ingenio el sistema más complejo de facturación electrónica, convirtiéndolo en un negocio. En vez de crear un API publica para los comercios, tienes que pagarle a un intermediario. Y se queja que los comercios no han podido implementar correctamente el sistema.

@PovilasKorop I think the key is in proper testing: SQLite in local, MySQL in prod, run tests against the db in local, same in stage with prod env.

My new Laravel video of the day. Upcoming Laravel 11: SQLite as Default Database - YouTube youtube.com/watch?v=CH7_Ys…

Disfrutando del #viernes de teatro 🎭, en la hermosa Ronda del Sinú. Gracias Alcalde @kerguelenhugo por activar estos espacios de recreación y cultura, la invitación es a las familias Monterianas para que asistan y se diviertan. #LaRondaVive #Montería

@ianlandsman Forgot to mention it's a SPA

@ianlandsman I'm using a SaaS that downloads 32Mb of resources, wtf, it can be done with Livewire with even less than 10% of that.

@fredldotme If the feature I need it's out of my area of expertise or highly complex I usually will find a external dependency useful. For example OCR. You don't do that from scratch lol.

As a developer, where do you draw the line between adding a dependency and rolling your own?

@JoshuaJBouw @RetroTechDreams Yup me too, lots of hours in the Ao1 server

@RetroTechDreams Played the heck out of Delta Force Black Hawk Down back in the day. Warhammer Fire Warrior and Dawn of War were pretty great too.

What are you playing first?

@cachecrab @flaviocopes You're using Mailchannels right?

@flaviocopes The world must work in mysterious ways because I literally just figured this out the other day. I’m building a newsletter product that requires the user to install a cloudflare worker I provide so they can send 100k emails free a day. Infinite emails for $10/m here we go

This space recording is really full of interesting bits

@taylorotwell Can I bring my own @OVHcloud server/VPS to Forge?

Forge is ready to deploy your Octane + FrankenPHP applications via the Application panel. 💪 Just a toggle away. ❤️ forge.laravel.com

@flaviocopes That day we will finally be able to use IPv6

What would you do if the internet was destroyed 😅

@bradleybernard @codewithantonio @PlanetScale Hello, I'm using it with servers in @OVHcloud Canada DCs. Everything is good so far.

@codewithantonio @PlanetScale Curious what you host your site with: VPS or serverless and what provider! I was demoing PlanetScale and trying to see if anyone out there is doing non AWS/GCP servers due to latency from app server to DB

I recently had 4 anomalies in my @PlanetScale database. It was my first time seeing this alert, and here is how I approached it: 1. Understanding my database baseline I am very well aware that my database activity will fluctuate depending on whether I just released a new video, or if its just business as usual with regular people visiting the site. Knowing the fact that I just released a new video, where I explicitly instruct viewers to view my website, I knew it could just be a deviation from the baseline activity. 2. Understanding what is an anomaly My first instinct was to query these spikes and match by rows read and query latency to see what was going on. To my suprise, they were high, but not higher than a week before that which did not trigger any anomaly at all, so I was a bit confused. Thankfully @PlanetScale has a great link to "Learn more about anomalies" where the following line is written: "Increased query latency can be indicative of an anomaly, but not always." 3. Determine what is my slowest running query / most not-optimized query at the time of the anomaly For me, this query was the leaderboard, which sorts all users by their XP and ID, and excludes those who want to stay anonymous. This is also publicly available to logged out users, so it is especially expensive. I've been fighting with this query for some time now, specifically because adding an index for "XP" is (at least from my research) not a good idea, since it is a very often changing column. So what I did was the following: - Move away from offset based pagination - Implement cursor based pagination for infinite load instead. This definitely helped a lot, since users now cannot jump from page 1 to page 700. But it still did not stop it from being my most expensive query. So I did the following: - Disable pagination all together - Load only top 10 users - Display currently logged in user rank, points and level at the top of the page so they know their status - Cache the result This stopped it from being my most expensive query, and greatly improved load speed of the leadboard page. 4. Continue to monitor the query After I improved my slowest query during the time of anomaly, As you can see in the screenshot, the problems went away. This confirmed to me, that at least in some part It could've been the cause. Most of my career I've spent as a frontend dev, so I rarely had experience fixing database-related anomalies. It is insane how well @PlanetScale's DX is, and not to mention the amount of information they share on their youtube and docs. And now im wondering, would @aarondfrancis do anything different? 😁

@kettanaito Omg I can remember when everyone was handling raw http requests in nodejs. How things have changed.

This is a little reason why I dislike Next's API choices. Because they always encapsulate back on "next". This is how I want to declare this function instead: function getIp(request) { let forwadedRef = request.headers.get('x-forwarded-ref') } A subtle difference, right? But the implications are HUGE. 1. I can use "getIp" with any Fetch API request instance. 2. I am familiar with Request, so accessing headers is no feature to me. Use the platform and get better at JavaScript, not Next. 3. "getIp" has a clear input. Using "headers()" from Next is magic. Where does it comes from? How does it know which request I'm handling? I dislike magic in APIs, it makes them unpredictable and robs them of any educational value at its core. 4. Suddenly, this is standard JavaScript. There's no need for Next. Maybe that's the issue. It's the subtle differences that make a huge impact on API design.

@DanielTurus @flaviocopes I think Laravel has some libraries that can do that

@flaviocopes It's fair to say that your statement is correct for a particular usecase and not as a general rule. Could you build a complex/interactive dashboard like page by sending html instead of JSON? Would the user experience be better?

Switching to sending HTML over the wire instead of JSON has been eye-opening

@PlanetScale Thanks, this is really useful for us newcomers

💲 pscale ping planetscale.com/blog/pscale-pi…