HN user
I'll definitely be watching out for more built-in streaming support. Being able to throw the standard types directly over the wire and trust that the library will handle optimally utilizing the connection would make this the RPC library that I've been looking for all year.
My biggest complaint with library management tools, and I think this applies to Librari as well, is the lack of multi-user support. I have a lot of books in my "home library", spread out over a few rooms. But I'm not the only person interacting with this library. Their are books on the shelves that "belong" to other people in the house, and we all joint manage the books. Sharing logins can work, but misses out on so many things that we would like to have.
My dream tool for this would allow multiple people to be "members" of a library, and be able to belong to multiple members themselves. They could collectively manage things like metadata, like what books are on the shelves, but could have individual things like ratings or tracking what they've read.
Plex is actually a really good example of this. I hope some day to find a tracker like that for my books.
That's what I learned in my university data structures course. I don't know anything about C++'s std::deque though.
It doesn't use Chromium. I think that their point is that Firefox's rendering engine, Gecko, can only have an impact on the rendering engine space proportional to its user base, which they have argued is insignificant.
During our early exploration, someone made a data loader that returned an entire svg of a chart, instead of the data for a chart. I think it was a headless browser running Observable Plot, but I imagine there are lots of ways to generate charts in a data loader.
The docs for data loaders are here: https://observablehq.com/framework/loaders. The simple version is they are simply programs that write their output to standard out. Very Unixey. When those programs are referenced in client side parts of the JS, they are reactively run when in development, and prebuilt for deployment.
I don't think we have any full examples of using a database yet, but we have written a bit about using DuckDB via its Node bindings here: https://observablehq.com/framework/lib/duckdb
I imagine that either Malloy's CLI or its Python bindings would fit very well here.
Observable engineer here. I haven't looked into Malloy much, but Framework's data loaders are very flexible. If you can write a script or binary that uses Malloy and writes to stdout, it can be a data loader. For example, although we use SQL a lot in our internal usage of Framework, Framework doesn't actually have any specific SQL support. We just use our database's normal bindings.
Once you start typing the same icons should show up again. You can click on them to perform just that search with the selected search engine, or right click on it to make it the default.
I'd say that Firefox versions are mostly a unit of time now, since they aren't generally semantically meaningful in the way semver is. While working on Firefox, we often make plans that are delineated not in weeks, months, and quarters, but in Firefox releases and the nightly/beta/release cycle associated with them.
The recent change to having a releaes every 4-5 weeks is a small modification of the existing practice of having a new version every 6-8 weeks.
The 6-8 week schedule is about a decade old now. It's original goals were to help ensure that unrelated features didn't block each other from releasing, and make releases more consistent, which helps localization, testing, and project management. In that regard I think it's been very successful.
It's probably not worth nitpicking a decade old decision though.
This feature (source maps) has been improving a lot over time. A year ago it was basically unusable, and today I rarely have problems jumping from console to debugger. We're even working on a way to show the original variable names in the scope inspector when paused in the debugger now.
On the other hand, Firefox's layout dev tools, like the grid and flex box inspector, are way better than doing the same task in Chrome, in my experience.
Luckily, Firefox supporting Wayland doesn't hurt you at all. I'm also very excited about Firefox supporting Wayland. It isn't dropping support for X11.
What do you mean by this? If you mean ctrl-tab switching to tabs in most recent used order, there is an option for that now. It's actually on by default in new profiles in Firefox. The previous behavior (in order switching) was left for existing profiles, to avoid disruption.
This is a very one sided view of the situation. What features were added to the browser during this time period? What was there time to do because these features weren't taking up focus?
You're right that Focus is not a full fledged browser, though it is actively being revamped to use GeckoView instead of Chromium. I think the change is in beta right now.
Mozilla is also starting to put more resources into Android (GeckoView, etc). Hopefully we'll see some exciting things in this realm soon.
I've found that washing the Yubikeys, especially if they are on a keychain, can be helpful. Soap and water, and then let it dry thoroughly before using it again. The theory here is that build up on the contacts make them not able to conduct as well, and so touches aren't detected. It worked for, YMMV.
Yes, this is using the same system as the Mr Robot marketing gimmick last December. That isn't very worrying to me though, since that platform is basically a generic "do something in the browser" capability, where "do", "something" and "the browser" are fairly loosely defined. Both the Mr Robot promotion and the Pioneer program could be done through other avenues. Shield is the most convenient mechanism currently available for this sort of fine-grained deployment.
I don't think that load on the download servers is an issue now. CDNs do a great job of distribution the load. The concern is more that some bugs (especially GPU related) only show up "in the wild". A slow rollout limits the damage if there is a show-stopping bug in the release.
It may not be in the release notes, but this release has some performance improvements specifically on OSX.
This is called "Flash of unstyled content", and has been a bug since 57: https://bugzilla.mozilla.org/show_bug.cgi?id=1404468 . In 57 the delay before initial drawing was removed, which mostly makes things feel faster, but sometimes makes things show up naked for a moment.
In that bug you can see there are several things that can cause this, including website issues and certain extensions.
I very recently got a pair of QC35s. The noise cancelling was neat, but didn't seem that useful in every day situations. Then I took a flight with them, and they still didn't seem that interesting. Until I took them off for the first time after take off. I was amazed at how loud the plane suddenly become. Now I'm very impressed with how much of a difference they make, and will definitely be bringing with me on every flight.
I have the 9350 (6th gen, I think?). It's not bad, for me. The key is to reconfigure the touchpad to actually work like a Mac's. That is, one finger click anywhere on the pad is a left click, two fingers is a right click. The default when I got the machine about 18 months ago to make the bottom left and bottom right parts of the trackpad left and right click, respectively, which is terrible.
Palm rejection isn't really a thing I notice though, on any computer. I've had people tell me it is objectively terrible on trackpads I've had no trouble with.
The fact that you need a $100 dock to connect to external monitors is ridiculous
I think you're overselling this a bit.
I've got a USB-C dongle that with USB-A, HDMI, Network, and VGA, and a USB-C to DisplayPort cable that is entirely passive. I don't remember how much I paid for them, but it was much less than $100 combined.
Since on most platforms the tabs go all the way to top of the window now, there isn't room to grab the window to move it around. Those spaces act as extra drag targets.
There is a pref that controls the minimum width now (browser.tabs.tabMinWidth). The default changed from 100px to 76px in Firefox 57. You should be able to tweak this to change the minimum width.
One interesting thing to me is that you don't have to write an entire app using only WebAssembly or JS. You can take the classic approach of benchmarking to find hotspots, and then optimizing those. Migrating these performance sensitive sections to more performant code can be a win.
I know about this, which is a really interesting comparison, but I haven't found any benchmarks.
The details of these deals aren't generally public, so I don't expect to see any official answers about this.
I first used it a couple years ago. I think it is only for GSuite accounts though. I have it on my work calendar, but not my personal one.
I'd imagine it is down to the number of recharge cycles. A phone tends to go through an entire recharge cycle every day to two, whereas I'd imagine a Tesla wouldn't go through it's entire range as often.