Is it time for a new browser engine?

Sorry in advance for the "browser thread". I have been wondering why there are not that many browser-engines. I kinda wanna make one. I think that we need more competition in that area. I don't even know the lang that I want to create that project in that well but I want this to succeed and want to learn. I am fine with adding more maintainers. I understand that this is not an easy job. Git: https://gitlab.com/RobinWils/unnamed-browser-engine What do you think of this idea?

>>15637 >I have been wondering why there are not that many browser-engines. Basically because to implement fully what's in a mainstream browser-engine takes a large professional team and the average consumer doesn't benefit from the work being done in house rather than by google/apple, and because the average consumer isn't willing to use a browser-engine which doesn't implement everything a mainstream browser-engine does (like links, lynx, w3m, dillo, netsurf, etc). >What do you think of this idea? I think first of all it's exceedingly ambitious. >I personally would like to use a LISP dialect but the language isn't decided yet. >The nature of LISP should make the browser engine easy extendable. This is correct, it also means you can implement the mark up languages (and js if you get there) in the form of reader macros and normal macros elegantly (probably). Lisp is the family of language for writing DSLs. >Steel Bank Common Lisp (SBCL) is the first dialect that I will try. Common Lisp is a good pick, put please only start to use SBCL specific functionality after you've got something working and you dislike your performance. If you don't know Common Lisp you should go learn the language first. >The idea behind this browser engine is that it has a configurable list of options. my suggestion would be to change this to be simply a lisp file which is evaluated after the functions and constants not in the main loop, this way the entire browser-engine would be programmable. Here would be your example: (setq js t) (setq css t)

Dillo and Netrunner have their own engines. You could count Servo as well.

Thanks for the great feedback. I should indeed start learning Common Lisp first. It will take a while before any actual code gets written unless someone wants to contribute/maintain.

>>15659 It seems there are quite a few lains on /tech/ closer to the dawn of their programming journey. Many here would likely benefit from a schedule and discussion if you're willing to organise. Perhaps you could encourage the others to join your project in full starting either with 'Practical Common Lisp' or one of the many wonderful Scheme books like 'The Little Schemer'. The admins would likely allow you to make a irc channel on the lainchan server for these purposes.

>>15662 That sounds awesome! Keep in mind that I don't know LISP that well yet though. The email that I used is my real email, so feel free to mail me.

Its funny lisp was suggested even though the idea before js was to use scheme as the browser language.

>>15665 Maybe that'll be a more rewarding experiment? sexps for html and scheme (or what other dialect you're into) for js?! The reason why there are only three major implementations is this web of soykaf adhoc specs that you'd need to follow and do it real fast, in order to appeal for them internet users. Just look at the html parsing section in html spec - and it's only tip of.. no not iceberg, rather pile of stinking soykaf.

>>15665 >Keep in mind that I don't know LISP that well yet though. The email that I used is my real email, so feel free to mail me. I'm a bit selfish so I'm only willing to point you in the right direction, and I'd prefer to simply discuss here (or in a new thread meant for this topic) as I'm quite a private person. Attached are the two books I mentioned, 'The Little Schemer' is less relevant but with better pedagogy and a simpler language, and 'Practical Common Lisp' is more relevant and well practical but perhaps more difficult and with the quirks of Common Lisp. I suggest you briefly look of them both and see which you'd prefer. Regardless of which you pick a chapter a week would be conservative, but might be a good pace for a group activity like this. The weekend some time would clearly be best for meeting (most folks here are (east) US or (west) EU for time-zones). You can ask about setting up a channel for your topic at: irc.lainchan.org:6697. Once you sort out the details you can make a new thread here on /tech/ with the book you choose attached as I've done here. Even if you decide organizing this group activity is too much these are the books I'd suggest and the minimum schedule I'd suggest. Best of Luck.

>>15672 I think that I will go with "Practical Common Lisp", I have read a bit of the book before. It was indeed not the easiest book but I think that it should be doable if you do the exercises. I will hop into the IRC when I find the time.

I am not on the IRC yet since a bouncer might be useful for the voice rights. I might be on the IRC soon though. My nick is rmw usually.