Yeah works well, as long as the code is rather simple and it occurred rather often in the training set. But I seldom use it currently (got a little bit more complex stuff going on). It's good though to find new stuff (as it often introduces a new library I haven't known yet). But actual code... I'm writing myself (tried it often, and the quality just isn't there... and I think it even got worse over the last couple of months as also studies suggest)
Yeah the whole situation is a little bit fucked, Google has too much power (over Mozilla) and browser engines in general.
But I'm optimistic about the open source community, e.g. brave integrates their own ublock compatible rewrite (in Rust ^^) into chromium. So I think sooner or later there are patched/forked chromiums and if Mozilla indeed makes the move towards WebDRM then also Firefox forks.
I really hope servo under the umbrella of the Linux Foundation proves to be a viable alternative in the future, that's a little bit more independent than Firefox.
yeah as nice as it is what you can achieve with trait-bounds there are definitely trade-offs, being compile time and error messages, and sometimes mental complexity, understanding what the trait-bounds exactly mean...
I really hope, that this area gets improvement on at least the error-messages and compile time (incremental cached type-checking via something like salsa)
Yep use a little bit more deeply cascaded generic rust code with a lot of fancy trait-bounds and error messages will explode and be similar as C++ (though to be fair they are still likely way more helpful than C++ template based error messages). Really hope that the compiler/error devs will improve in this area
Fortunately, there are other good browsers that don't follow this bullshit. I'm lucky that I haven't seen ads in ages (and I absolutely don't miss them). Websites/Services should just find different ways to monetize their work, ads are the worst way IMHO (collecting data, promoting often useless consumption in a world where we need to reduce consumption, annoying etc.).
Yet it still feels sloppy (and I couldn't find why to the defense of Jerboa, I skimmed through the relevant code but couldn't find immediate issues and there's an open issue: https://github.com/dessalines/jerboa/issues/445)
I don't get why, with so much hardware power we still have these issues...
yeah mostly (apart from big corporate and all the related issues), most of the stuff feels a little bit bulky/sluggy because of the overuse of web-technology (say Teams or VS Code (while being a great editor there are much faster ones)).
But Github itself is quite convenient for me to use for open source (and for work at that)...
Well I feel like all the "modern" SPA stuff that got recently added, certainly adds "some" kind of bloat on top, it's not as snappy anymore as it once was, but I think most of the stuff is improving QoL, so I think it's reasonable...
I'm very split between Github (currently) providing a really nice interface/collection/way to access all kinds of open source projects and the obvious 'out-of-control centralisation' by the mega corp Microsoft.
It definitely got a little bit bloated the last years, but I still think it has a generally nice interface (browse code/review stuff, simple issue/PR system, simple way for CI via actions etc.).
But I really hope something like https://forgefed.org/ takes off someday, I feel like if the barriers are much lower to get onto a different network with the same user (without registering etc.) decentralisation can lead to more innovation in this space (management of all the stuff that Git doesn't manage itself, like issues, PRs etc.).
The beauty of Git though is that it's decentralized, so you can just mirror it on Github while mainly using a different platform. If you want a bigger userbase interacting/contributing with your project you'll allow PRs and issues there and if not just add a link to the README that points to the platform you're using...
Yeah works well, as long as the code is rather simple and it occurred rather often in the training set. But I seldom use it currently (got a little bit more complex stuff going on). It's good though to find new stuff (as it often introduces a new library I haven't known yet). But actual code... I'm writing myself (tried it often, and the quality just isn't there... and I think it even got worse over the last couple of months as also studies suggest)