vvillena

The exterior design language is the same as their current newer models like the C-HR or even the GR GT. The interior is like any other concept car. I don't like it either, but it looks like a normal Toyota.

The platform compromise is a new thing. One of the reasons people used Twitter was its open nature, enabling its use as a bulletin board where anyone could post a message and being sure that the message link was free for anyone to see. This isn't the case anymore, as X forces people to make accounts to see messages.

> it's up to you who you follow and what exactly you read on Twitter

The Twitter experience is a highly algorithmic feed where tweets from the people you follow are not guaranteed to be in chronological order, or even shown to you at all. The default experience also mixes it other tweets "you might be interested in", that is, things they want you to see.

There's also the fact people could freely see tweets and follow discussion threads, that's now locked behind an account requirement.

It depends a lot on the config language. Gradle has the uncanny ability to be a "worst of both worlds" option, where there are not many upsides to its using a full language for its configuration. Build configuration is also a space where people and even some tooling seem to forget about good engineering practices, giving birth to fragile builds that people are afraid to modify and improve.

Isn't the shader stuttering an issue with games refusing to compile shaders ahead of time?

I also lived in Madrid. 1h15 means you were not in Madrid, but in one of the surrounding towns. 30 minutes in a car is enough to get from any point in Madrid to any other (assuming clear traffic).

It's obvious that people outside of the areas where transport is dense enough won't enjoy all the benefits. This is a usual complain about all dense public transport areas, but it simplifies to the fact that you were not in the area with dense coverage. In the case of Madrid, Metro is only dense enough in areas surrounded by the Circular line. There's also a very dense bus network that covers everywhere within M-40 and even some areas outside it (some bus lines can be hindered by traffic though, there are bus lanes in most critical spots but there's always some unpredictability).

Also, you're wrong on the closing time, it's been 1:30 AM for a long long time. That's the time where the last train departs on each line, so it's possible to catch trains at almost 2 AM depending on where you are. By Madrid standards, if you are out and about at 1 AM it means you are in a bar or in a party, and it's time to decide whether to continue through the night and get back when Metro opens again at 6 AM, to use a taxi to get back home, or to use the night bus network if you're lucky to be near a stop and willing to wait for it.

Because public transit, done well, is the fastest way to move around. It gets you everywhere, even to places personal vehicles can't reach. It's a lot cheaper. It generates hubs of activity that keep cities lively and relevant. It doesn't get stuck in traffic. It doesn't need to be parked.

The benefits of good public transport are so mind blowing that it's difficult to explain unless you have lived on a city that has it.

Gnutella fell in popularity before Youtube existed. The big reason the decentralized P2P file sharing networks went away was their vulnerability to spoofing. Downloads were slow so it was important to be sure of the quality of the files being downloaded, which meant grabbing links from a website or some sort of forum. And once there's a forum, for some of the sites it was not much of a jump to add a BitTorrent tracker to the infrastructure. Additionally, for those sites that were primary sources of uploaded material, BitTorrent was a much better fit.

Their own greed is causing their issues. They could be doing a million different things to reduce demand, but they don't want to dampen their current growth and have opted to continue scaling up at the cost of quality.

The behavior of the iOS keyboard also showcases how there must not be many decision-making people who communicate in multiple languages.

I have a Emacs + Spacemacs setup only for Magit. The base Spacemacs config works well, so I never had the need to tinker with it. Nowadays I don't care much about the rest of Emacs. It stays out of the way, and I keep happily using Magit.

The free rectangle area is a 16:10 screen, the space around the notch is a freebie.

A high end amp+speaker system from 50 years ago will still sound good. The tradeoffs back then were size, price, and power consumption. Same as now.

Lower spec speakers have become good enough, and DSP has improved to the point that tiny speakers can now output mediocre/acceptable sound. The effect of this is that the midrange market is kind of gone, replaced with neat but still worse products such as soundbars (for AV use) or even portable speakers instead of hi-fi systems.

On the high end, I think amplified multi-way speakers with active crossovers are much more common now thanks to advances in Class-D amplifiers.

Compiler errors got a lot better during the Scala 2.12 era.

Avoid making memories. Keep your brain pristine.

This is not an issue with Java and the other JVM languages, it's simple to use GraalVM and package a static binary.

The screen size advertised by Apple measures the "full screen" area, the undisturbed 16:10 rectangle of pixels. I just took measures on both a 14 and a 16 inch Macbook Pro. The screen we get is indeed slighly larger.

If you want to avoid the extra space, it's as easy as using a 16:10 resolution size. The menubar will drop down to the 16:10 space.

Reader mode.

Funny that the language famous for enforcing standard formatting allows freedom for this one particular thing.

The Spain blackout was caused by a multitude of reasons. Lack of stability was one of the factors, but there were other causes, such as energy generation facilities disconnecting while the oscillations were still under a nominal range, or a generator ordered to become online to induce stability, that started driving the load in the wrong direction. All this was compounded by a distribution network unable to redistribute or at least isolate the problems to individual regions, resulting in a complete blackout.

All in all, it's several things that need to be reinforced. The distribution network needs to be smarter. The energy generation facilities need to be tested through their entire voltage range, so they can be counted upon. And there has to be more voltage inertia available in the network.

It's important to note that the infamous Dropbox comment was not just misguided. It was wrong.

The proof is that multiple competitor products have been launched since, and all of them have had sync issues at some point, with different degrees of severity ranging from sync delays, through data conflicts, up to loss of data in all synced devices. To this day, I still trust Dropbox more than its competition. This includes custom rsync scripts.

This is impossible by design. Decades ago there were some distributions that had this as a goal (e.g. Mandrake, Suse), they included an application similar to the Windows Control Panel to manage everything. But such applications can never reach into all the corners, unless the distribution is severely locked down. The example of this extreme is... macOS. And still, there are some cases where dropping into the command line is the better or even the only option.

Back on Linuxland, the userbase realized this about two decades ago, when Ubuntu launched. Having a nice default experience was considered better than having easy tweakability, because Ubuntu could also be configured to the fullest extent in the classic Linux way of reaching into the guts of the system and rearranging things to taste. Not that I would ever recommend tweaking Ubuntu too much, but it can be done.

What about the other end? Most people who like fiddling with Linux by reaching into its internals have settled on distributions such as Arch, where this way of managing the system is expected and thus the distribution works to ensure this experience is as easy and predictable as it can be, by providing a good happy path experience for common scenarios, and providing top-notch documentation for common and uncommon customization options, or minority hardware platforms and devices.

From the FAQ:

> Each time the Leta application is restarted (due to an upgrade, or new version) server side, a new secret hash is generated, meaning that all previous search queries are no longer visible to Leta.

If I read this correctly, the cached data is per-instance, there would be no way to share cached data among instances if each one has its own secret hash and they are cycled on each start.

This is literally the case. Pirated streams keep working, while a good chunk of the internet is rendered inoperative during weekends.

No problems here, both the normal view and reader mode seem to work well.

For anyone not familiar with previous designs, each component in https://m3.material.io/components has a "comparison with Material v2" section.

For everyday use, wireless headphones offer a superior experience simply due to the lack of a cable, and for the cases where an audio output is desired, it should be easy to connect the phone to an audio interface. Is any of this a problem in the Android ecosystem?

Bisecting with --first-parent takes care of this.

The "squash everything" mantra turns git commit history into a series of snapshots devoid of any logical notion about how code evolves.

Squashed commits are strictly worse than plain, non-fast-forwarded merges from rebased branches.

Congratulations on the project! This is an impressive feat at any age. Personally, I love how you found your own balance between extreme minimalism and a useful set of features, arriving at a solution that is indeed below 2k lines of C code. That kind of thinking is not a quality every software developer has, and you could make a career out of it if you decide to work professionally in this field.