trickypr

I think it does take away from the point. I could not be bothered trying to filter the information out of whatever this llm layout is.

That seems like a horrible idea:

1. Do you really want the rust compiler to run at the speed of an llm?

2. Compiler optimisations are already extremely unpredictable with deterministic compilers[1], I hate to think how unpredictable your compiler would be.

3. What if someone else wants to build the software, do they have to decide on optimisations now? What if the optimisation depends on your features not available on old generations of CPU? (There is a reason we don’t compile with -march=native)

4. Compilers already have “unsafe” optimisations, but people rarely enable them (-ffast-math)

[1]: https://faultlore.com/blah/oops-that-was-important/

The anubis docs website is used as an example of how anubis works.

Also, if you are not using a big CDN (e.g. Cloudflare), you still need to pay for the traffic (even if you don't pay as much for the CPU). This is doubly true for smaller CDNs like Bunny.

I think it is still compute-heavy if not compute-hard, and stuff like WASM SIMD makes it less compute heavy (not really available in JS). And also, for memory bound problems, you probably don't want the JS GC needing to clean up afterwards when WASM gives you that control.

I think it is more likely they are just drinking their own koolaid (and their friends are too). They are paid to bring in investor dollars, which requires lofty promises.

Bold of you to assume that “current economic inventives” will remain around for 500 years, especially given the American-style neo-liberal economy is only like 70 years old.

I think their point is that paint shouldn’t have any GenAI features (or any new features other than compatibility)

I think the copyright should go to the tool creator in that case, if the tool is acting as an extension of the creator.

But if the tool is created from collective human creation, the copyright should belong to all humanity, not the person who triggered the tool.

If you trained an LLM entirely on your own input, I think you should own the output, but that is not the case for any widely-used llm.

What if it was a human who took the photo, would the copyright still be owned by the equipment owner / person who set it up?

What if this applied in a photography class? The instructor owns the equipment and helped “set up” the photo. Does the instructor own the copyright?

I think there is an easy test: if you didn’t write it, it’s not yours.

If we go back 10 years and your friend says “I have an idea for an app, here it is,” and you build it, you own the copyright because you wrote it.

You give an idea to the pile of math calculated of the stolen work of humanity, the math owns it (which it can’t, so no one owns it).

No matter how detailed of a conversation you have with a friend, I don’t think they have justification to claim copyright over code written by you.

In the same way “smoking causes cancer” was a minority opinion in the 50s and 60s.

Australia has a similar issue, in part because we import a lot of US domestic vehicles and advertising, and in part because the ato has tax exemptions that only apply to overbuilt utes[1].

[1]: https://www.drive.com.au/caradvice/ute-tax-breaks-exemptions...

Cloudflare isn’t just as simple as “bot” or “not a bot”. They give customers categories of how likely a request is to be automated (or a number for some enterprise plans)[1] and the customer decides what happens to your request (nothing, challenge, block, run a worker, etc).

[1]: https://developers.cloudflare.com/bots/concepts/bot-score/#b...

> incremental old-tree reuse were removed

Maybe I am missing what they meant here, but isn’t the entire point of tree-sitter that you can reuse old trees to make edit updates faster? It is seems quite probable that all the performance gains came from optimising for fixed files with no error recovery, at the cost of how ts is actually used.

So your benchmarks are primarily just “how fast is go’s c interop” rather than any algorithmic improvement on tree-sitter?

Edit: yep, you are just calling a c function in a loop. So your no-op benchmark is just the time it takes for cgo to function. I would not be able to get any perf benefits from e.g. rust

Do you have an equivalent of TreeCursors or tree-sitter-generate?

There are at least some use cases where neither queries nor walks are suitable. And I have run into cases where being able to regenerate and compile grammars on the fly is immeasurably helpful.

At least for my use cases, this would be unusable.

Also, what the hell is this:

> partial [..] missing external scanner

Why do you have a parsing mode that guarantees incorrect outputs on some grammars (html comes to mind) and then use it as your “90x faster” benchmark figure?

I think a lot of your examples are flawed. Google didn’t build the initial version of YouTube, they bought it.

A lot of Apple’s apps predate the App Store. The apps that came later had limited use until Apple spent a lot of time refining them. Think Apple Maps.

Microsoft released Word for Mac a year before they released Windows 1.0, so Windows was “down the stack” for them.

We are currently on Firefox 112.0 (latest major, not latest minor). Updates are applied inside of `gluon.json`, not via pnpm.

https://github.com/pulse-browser/browser/commit/b9bf6b439a13...

Mozilla has landed a partial fix for this[1], which is currently on the nightly branch and should be available with 108.

[1] https://hg.mozilla.org/integration/autoland/rev/04e6abff2792

Currently it is in the "tech demo" phase. The only people I would recommend using this are people willing to contribute in some form. I would not recommend it as a daily driver hence the "alpha" tag on all the builds.

To be honest, the only part that I would recommend is our tooling that we use for soft-forking Firefox/Gecko, which is where we have the most outside testers and where I am focusing most of my upstream (mozilla-centeral) patches to improving.

I think you are miss interpreting me. I don't want to focus on privacy above all else. To do that well would compromise on the usability of the browser. For example, if I were to advertise it as "privacy focused", I would not feel comfortable unless I included strict anti-fingerprinting, which would break websites.

It is a balancing act, and I won't be collecting or selling your data, but I won't be competing with Librewolf or Tor for the "privacy" market either. However, at the bare minimum, we will be more "private" than Firefox by extent of including uBo and disabling Mozilla's non-critical telemetry by default.

We try to keep it up to date with the latest version of Firefox. Currently it is bound to 105.0.1 but we will soon be on 106.0.5 (once PR#135 merges).

Our website is a touch (read very) out of date. To clarify, we are working on newer features, like the sidebar tabs. Currently, it is a combination of a tech demo and a set of personal preferences (features, extensions, preferences) that I and the other developer would prefer. If you have ideas for new features, I am interested to hear them.

Hi, I am a dev on Pulse Browser. I want to clarify our current development direction (I haven't significant updated the website recently). We have been moving further away from privacy and towards developing/iterating on features faster than Firefox is willing to, along with providing sane defaults.

For example, we have been experimenting with a opera/vivaldi-like method of accessing Firefox's sidebar. Whilst you could technically achieve this via patching `omni.ja`, you would need to repatch it every update. At that point, it might as well be a custom browser.

Sorry about Vercel's insight stuff, not sure how that got enabled, but it should be disabled now (or whenever Cloudflare invalidates your cache).

If you want further reading into some of the specific parts of our code that we are putting most of our effort into, here are some links: - https://github.com/pulse-browser/browser/blob/alpha/src/brow... - https://github.com/pulse-browser/browser/tree/alpha/src/brow...