Timeline for Why is short polling considered an antipattern for event-sourced systems?
Current License: CC BY-SA 4.0
25 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 25 at 16:13 | comment | added | guest271314 | No, you made this absurd claim alluding to what you and other people do: "Real people write apps to make money, not for kicks." I also do consulting myself for people who want primary source research done. They wouldn't ever hire you, becuase you'd stop at it's not standardized or stabilized, yet. | |
| Oct 25 at 16:12 | comment | added | freakish | @guest271314 I asked you about what you do for living? Interesting. Did you try consulting a professional? | |
| Oct 25 at 16:04 | comment | added | guest271314 | "Real people write apps to make money, not for kicks." Haha. Too funny. I hack for sport. For money I do other things, like build multi-million dollar estates for folks who happen to strike it rich with one of these little toy applications that gets suckers sprung. I don't care about your little beans. You gotta come see me to build your house, like one of the Microsoft executives is doing right now, 'cause you ain't doin' it. | |
| Oct 25 at 16:01 | comment | added | freakish | You completely miss the point. It doesn't matter what you can do NOW in a browser. Specifications and standards are there to stabilize reality and not rewrite stuff every time a backwards incompatible change happens. Noone cares about what expirements Firefox or Chrome does at the moment. Real people write apps to make money, not for kicks. Anyway, I'm done. This discussion isn't even funny anymore. | |
| Oct 25 at 16:01 | comment | added | guest271314 | I have no idea why you are talking about WebTrasnport specification becoming stable. If you want to make WebTransport connections right now in Chromium-based browsers and Firefox you can do that right now. WebTransport is HTTP/3 over QUIC. Here's some instructions for doing so github.com/guest271314/… | |
| Oct 25 at 15:54 | comment | added | guest271314 | Huh? Specifications are at best a guide that folks try to agree on about how to do stuff. I'm not under any obligation to wait around for specification authors to agree. WebTransport is already shipped in Chromium-based browsers and Firefox. In Chromium based browsers it's possible to make direct UDP connections, and it's possible to use TLS as a client in the browser github.com/guest271314/sockets/blob/fetch-webrtc/assets/…. So whatever ideas you have about what can't be done are outdated and stuck in basic user thoughts; not software engineering thinking. | |
| Oct 25 at 15:51 | comment | added | freakish | @guest271314 sure you can do what you want. By why bother with draft specs when you can simply recompile the browser with any features you want? | |
| Oct 25 at 15:44 | comment | added | guest271314 | I just demonstrated I can do what you say can't be done in the browser. Software engineers and hackers don't wait around for specifications and stabilization. "TCP and UDP sockets in the browser" github.com/guest271314/sockets. Clearly you are not using WebTransport at all. As for WebSocket, I've written a WebSocket server that runs in Chromium-based browsers, and in Node.js, Deno, Bun, and txiki,js github.com/guest271314/direct-sockets-http-ws-server | |
| Oct 25 at 15:43 | comment | added | freakish | @guest271314 ahh, another draft spec. That's cool. Once it gets stabilized, together with WebTransport, then talk to me. Otherwise I couldn't care less. | |
| Oct 25 at 15:40 | comment | added | guest271314 | No, You're not reading and experimenting enough. We can make UDP connections from the browser. See the direct-sockets branch github.com/maceip/quinn-wasm/tree/direct-sockets "This fork extends quinn-wasm to support the Direct Sockets API, enabling QUIC connections directly from the browser without requiring a WebSocket relay server." | |
| Oct 25 at 15:39 | comment | added | freakish | @guest271314 "It forwards UDP packets over a WebSocket connection to a relay server", is that how you define "I can do anything"? Another joke. | |
| Oct 25 at 15:35 | comment | added | guest271314 | github.com/maceip/quinn-wasm "This demo runs the quinn QUIC implementation compiled to WASM in the browser." UDPSocket interface wicg.github.io/direct-sockets/#udpsocket-interface | |
| Oct 25 at 15:26 | comment | added | guest271314 | You are talking to the wrong person about what I can't do in the browser. I can do whatever I want in these browsers. WebTransport is literally HTTP/3 over QUIC. We have UDPSocket in the browser in an Isolated Web App that we can connect to from arbtritrary Web pages over WebRTC Data Channels. Maybe you can't make direct QUIC connections in your browser. I can. | |
| Oct 25 at 7:27 | comment | added | freakish | @guest271314 and of course you can't use Quic from browser. The fact that you don't know the difference between Quic and WebTransport, and you keep repeating those mistakes, is a joke as well. WebTransport to Quic is almost exactly like WebSockets to TCP. It lives on top of HTTP. And btw the design of WebTransport isn't even finalized yet, and still subject to major changes. So I don't even know what you are talking about. But hey, feel free to use unstable tech like this. Again: I don't care. | |
| Oct 25 at 6:52 | comment | added | freakish | And btw, if you think that these are all just tools than go back to writing assembly. Or better: punched physical cards. These are just tools right? There is no better and worse. I know that dumb people like you keep repeating this foolish statement. Entire progress is about making better and better things. But no, lets ignore this. Fkin joke. | |
| Oct 25 at 1:19 | comment | added | guest271314 | "Yes, it does matter how many people use a tech." No, it doesn't. There's a whole bunch of old fools. I'm not a follower. "Regardless, WebSockets are not only used more, but more importantly are vastly superior to SSE." What's up with the "used more" and "superior" jargon? As if a a whole bunch of people doing something has inherent value. It doesn't. I use SSE, WebSocket, WebTransport, WebRTC, UDPSocket, TCPSocket, fetch(), and other means to do peer to peer and server-client networking. Use the appropriate tool. There is no "superior". They are all just tools. | |
| Oct 25 at 1:02 | comment | added | guest271314 | "You can't open Quic connection from browser." Yes, you can. Read my answer. Everything is comparable. Depends on the use case what tool to use. | |
| Oct 24 at 14:50 | comment | added | freakish | @guest271314 Quic is superior to WebSockets? These aren't even comparable. One is a transport protocol, the latter is not really (requires HTTP). You can't open Quic connection from browser. | |
| Oct 24 at 14:49 | comment | added | freakish | @guest271314 I think your opinion on Windows doesn't really mean anything. Yes, it does matter how many people use a tech. But that's not the time for such discussion. Regardless, WebSockets are not only used more, but more importantly are vastly superior to SSE. | |
| Oct 24 at 14:45 | comment | added | guest271314 | So, you think if a lot of people use something that makes it a viable software solution? A bunch of people use the garbage that is Microsoft Windows. That doesn't mean anything. The HTTP/3 folks would say QUIC is "superior" to WebSockets, even given the broad usage of WebSockets. | |
| Oct 24 at 14:44 | comment | added | freakish | @guest271314 I didn't mean it is unsupported. I meant that hardly anyone uses it anymore, since WebSockets are simply superior in every aspect. | |
| Oct 24 at 14:18 | comment | added | guest271314 | "SSE seems to be a dead tech anyway." Huh? EventSource is not "dead". It's still defined in modern browsers. | |
| Oct 24 at 8:58 | history | edited | freakish | CC BY-SA 4.0 | deleted 10 characters in body |
| Oct 24 at 8:25 | history | edited | freakish | CC BY-SA 4.0 | added 55 characters in body |
| Oct 24 at 8:18 | history | answered | freakish | CC BY-SA 4.0 |