Puszta Browser Used AI

Last devlog before shipping! I finished the webpage, wrote the GitHub README and in the program itself I fixed the remaining few bugs and added support for h2-h6 as well as br because the lack of them was really bugging me (get it? bug-ing? :P).

Over all I am extremely proud of what I was able to accomplish; It' able to open basically all sites and they are surprisingly readable (makes you realize that under all that CSS and JS just how important HTML is...) and other than CSS it's mostly a fully functional (static) web browser! I am extremely burned out from this project though so don't expect a new update for quite some time (maybe I'll revisit in the new Autumn of Making?). I hope I get many shells for this project :D

Hi and once again sorry for the long devlog, I was busy with Jumpstart. This release is mostly of bufixes, graphical and QoL features, as most of the code is either complete or would need a bigger overhaul than I have the mood to do rn (I am very burned out from this project :[ ). New features are a completely new, much more modern UI with light AND dark mode, the program feels much more polished and less like a hastily thrown together thing now :D I also implemented browser history for going backwards, forwards and rewinding (I do not save it rn, I can't decide whether I should start saving it or leave it like this and call it a security feature :P), as well as fixed redirecting so that relative redirects can actually work now. I also started working on a nice little site for the project and it's looking good if I do say so myself :) I'll work a bit more on the site and fix up the remaining bugs, but then it's shipping time :D Also sorry for the picture, SoM didn't let me upload a video for some reason :/

Heya! Truly sorry for a long devlog, haven't posted in a while. I was busy trying to wrangle the HTML parser into existence which as you can see by my GitHub commits was... not easy so to say :/ In the end I decided to make a custom one semi-based on the tutorial, and I'm proud to say that (excluding some line height issues) the parser works! I'm getting really tired of this project tho so I'll probably ship after fixing these issues and implementing some QoL features (tabs, settings, history, yknow the gist). I plan to ship tomorrow (hopefully I can implement all these by then)! See ya next time :)

Hellohellohello! Didn't originally plan to submit a devlog this big but there was some weird bug with SoM where it thought I only ever had 2 minutes and did not let me make a devlog, that was weird :/ Anyway, HTML parsing is well underway now with a proper (although sometimes a bit glitchy) layout system and everything! A lot of changes aren't really visible since they're bugfixes or under the hood improvements, like the addition of variable fonts which was a PAIN IN THE ASS (average freetype2 moment). I feel really good about the progress I've made, the browser can already open and render basic sites great and I'm really close to finishing HTML! CSS will be really challenging tho...

verrrrrry quick fix - found what caused the segfault and patched it so we can actually run the program now! Still no text rendering with the new parsing system but hey, progress :)

Started actual HTML parsing today - here you can see one of the mid steps where I implemented different font styles and weights. Sadly I didn't have any more time today and I even got a yummy SIGSEGV :(

HUGE devlog: Added Unicode support, HTTPS, view-source, file:// loading, redirects work now (huge W for http links!), scrolling and started parsing HTML. The browser is starting to look really good :D

Shorter devlog today: Added error checks, and unsuccessfully tried to add HTTPS.

No real HTML parsing yet, only cleaning the text from tags, as I focused on other things like a much more robust URL handling (the old one randomly decided to just not update the URL for no reason), HTTP 1.1 support (User-Agents yippee!!!), and the pièce de résistance: Dear Imgui for the GUI; This means I can even support things like tabs in the future :D

The next step is to implement a better socket system as the current one does not support HTTPS which in 2025 immediately makes the browser not being able to open 90% of websites. After that (and a couple extras I want to add based on the browser.engineering exercises) we can finally jump into HTML parsing!

After a LOT of wrangling with FreeType, GLFW and sockets, I finally got the first HTTP request working and the first HTML. The fun part comes next: HTML parsing :D