roelschroeven

The numeric conversion functions in the STL are terrible. They will happily accept strings with non-numeric characters in them: they will convert "123abc" to 123 without giving an error. The std::sto* functions will also ignore leading whitespace.

Yes, you can ask the std::sto* functions for the position where they stopped because of invalid characters and see if that position is the end of the string, but that is much more complex than should be needed for something like that.

These functions don't convert a string to a number, they try to extract a number from a string. I would argue that most of the time, that's not what you want. Or at least, most of the time it's not what I need.

atoi has the same problem of course, but even worse.

"The reasonable man adapts himself to the world. The unreasonable man persists in trying to adapt the world to himself. Therefore, all progress depends on the unreasonable man." (George Bernard Shaw)

But are those really different classes of people? Isn't everyone a maximizer up to a point where they think "good enough"? Where that limit changes between people, and for each person probably depending on area of interest, area of expertise, and so on?

That solves part of the problem, but doesn't help when you're in a supermarket or any other event where you're moving around.

My previous cars had keys that I could manually switch off and on, which is also not a full solution because it only works for people who take the effort to always do that, but at least it gives people to opportunity to complete prevent relay attacks.

All in all I'm not a big fan of key-less entry. Having to press a button on a key to gain entry can maybe be a bit of an annoyance, but in my opinion it's not a big deal compared to the advantage of completely preventing relay attacks.

Wouldn't it be easier to use WSL? Microsoft has done al the hard work in getting native Linux code to run on Windows. I don't know how bare-bones you want it to be, but if you install Debian in WSL without extra packages, that seems pretty bare-bone to me. And it seems to me you avoid unnecessary friction from switching from OpenBSD to Linux later on.

WSL requires Windows 10 or 11 though.

Fixed now!

Issue created: https://github.com/invisal/sqlite-internal/issues/3

I tried with a simple database, and it just shows me a blank screen. The console in developer tools shows an exception (both in Firefox and Chromium):

    index-CGfYebmt.js:85 Uncaught RangeError: Offset is outside the bounds of the DataView
        at DataView.prototype.getUint32 (<anonymous>)
        at R0 (index-CGfYebmt.js:85:4683)
        at X0 (index-CGfYebmt.js:85:7966)
        at index-CGfYebmt.js:85:8157
        at Array.map (<anonymous>)
        at Q0 (index-CGfYebmt.js:85:8150)
        at index-CGfYebmt.js:85:8240
        at Object.useMemo (index-CGfYebmt.js:48:46087)
        at My.$.useMemo (index-CGfYebmt.js:17:7276)
        at V0 (index-CGfYebmt.js:85:8228)
    R0 @ index-CGfYebmt.js:85
    X0 @ index-CGfYebmt.js:85
    (anonymous) @ index-CGfYebmt.js:85
    Q0 @ index-CGfYebmt.js:85
    (anonymous) @ index-CGfYebmt.js:85
    useMemo @ index-CGfYebmt.js:48
    My.$.useMemo @ index-CGfYebmt.js:17
    V0 @ index-CGfYebmt.js:85
    tc @ index-CGfYebmt.js:48
    bc @ index-CGfYebmt.js:48
    lo @ index-CGfYebmt.js:48
    No @ index-CGfYebmt.js:48
    wm @ index-CGfYebmt.js:48
    Xc @ index-CGfYebmt.js:48
    Ao @ index-CGfYebmt.js:48
    Qo @ index-CGfYebmt.js:48
    me @ index-CGfYebmt.js:25

"11.4% of respondents had experienced at least one crash that they suspected was due partly to structural weakness".

11.4%?? Are Australian bikes trying to eat their riders?

I live in Flanders, where bikes are used quite a lot for commuting, doing groceries, going to school, all that kind of stuff (less than the Netherlands, probably quite a lot more than many other countries), and also quite a lot for sport. The bikes are a mix between new and old (sometimes very old), and everything in between. Crashes because of structural weaknesses are really exceptional.

Best is no ads. Second best is ads for things that I have no interest in, that I'm never going to buy anyway. Anything else has the potential to influence my buying decisions, which I don't want.

But I do care about privacy, and other people do too. And I can read between the lines, and never take PR for fact. What remains is that Mozilla is looking for new cash, and sees selling user data as a solution for that. They want to call it differently, they want us to think it's all OK, but it's not. It's still better than other browsers though.

No, I think people who use ad blockers are a minority. And it's not getting better with Chrome/Chromium switching to Manifest v3 which has significantly worse support for ad blockers.

At my (small) workplace we all use Thunderbird, and I use it for my personal email as well.

A good desktop client, once configured, works a lot better than web-based email clients, especially (but not only) when you have different email accounts that you want to use in the same interface.

Yes. CSV is a data interchange format, it's meant to be written on one computer and read by another. Making the representation of data dependent on the locale in use is stupid af. Locales are for interaction with the user, not for data interchange.

One problem is that Excel uses locale settings for parsing CSV files (and, to be fair, other text files). So if you're in e.g. Europe and you've configured Excel to use commas as decimal separators, Excel imports numbers with decimals (with points as decimal separator) as text. Or it thinks the point is a thousands separator. I forgot exactly which one of those incorrect options it chooses.

I don't know what they were thinking, using a UI setting for parsing an interchange format.

There's a way around, IIRC, with the "From text / csv" command, but that looses a lot of the convenience of double-clicking a CSV file in Explorer or whatever to open it in Excel.

What data is used to train this AI? The article doesn't say anything about that (tough I have to admit I didn't read it super carefully). My first thought would be exactly all this historical data, but then you can't use that same data to test the AI's performance. Are different subsets of the available historical data used for training vs testing?

Are existing whois-clients going to be updated to support RDAP next to Whois, or will we have to use different clients?

As far as I can see, an RDAP request is a simple HTTP request, looking like http://example.com/rdap/ip/192.0.2.0. Web servers still support HTTP/1.1 (or probably even HTTP/1.0 and HTTP/0.9). This is trivial to implement for clients. A simple HTTP request like that is about he simplest thing to do. You'll have to use curl or wget instead of netcat if you want to do it manually. No big deal.

"A JS-derived serialization format" ... You mean JSON, which is about the lowest common denominator in Internet data exchange these days (and has been ever since we found out that XML was overly complex and JSON was much easier to use). You'll have to use something like jq instead of grep to extract information from the data manually. Or rather, you'll be able to use the powers of jq. Again, I don't really see the problem here.

Thanks, I didn't know that.

It is? I've only ever seen dd/mm/yyyy or dd/mm/yy as standard date format, with different separators instead of / depending on the country.

You can donate to MZLA Technologies Corporation for Thunderbird development however. If it works for Thunderbird, why can't it work for Firefox?

> Fracking lowers emmissions?

How? Burning fuel from fracking emits CO₂ into the atmosphere, just like any other fossil fuel. On top of that unburned methane is often released into the atmosphere, which is much more potent than CO₂ is a greenhouse gas.

> On the one hand, they claim Global Warming is going to bring us to the brink of human extinction unless we act rapidly.

Extinction is exaggerated (I'm not even sure who exactly claims that?), but we are indeed making our environment tougher and tougher to live in. There are already places that have become too hot to live in for parts of the year, to name just one thing.

> EVs have ferocious acceleration

Not all of them. It's pretty easy for manufacturers to make ridiculously fast accelerating cars, and it seems those cars get all the attention. But there are others. I drive a Volvo XC40, which comes in two variants: one with two motors, one with a single engine. The two motors variant does 0 - 100 kmh in 4.9 seconds, which is pretty fast; to get that acceleration in an ICE car you'd need a pretty sporty sport car. But the variant with a single engine (which is the one I drive) does it in about 8 seconds, which is plenty fast for me but certainly not in ferocious acceleration territory.

I have no numbers on relative market shares of very fast vs more normal EVs. It wouldn't surprise me that the share of normally accelerating EVs will increase over time as we get cheaper cars meant for a larger public instead of expensive cars for enthusiasts.

BTW that fast acceleration of EVs is mostly at lower speeds. At higher speeds the benefits of an electric drive train diminish and depending on the specifics the EV might very well accelerate slower than an ICE.

According to carsized.com an ICE Volvo XC40 (not sure whether that's a plug-in hybrid, probably not) weighs 1580 kg, while the BEV one weighs 2040 kg. That's a difference of almost 500 kg. The BEV weighs almost 30% more, that seems pretty significant to me.

Batteries are heavy!

https://www.carsized.com/en/cars/compare/volvo-xc40-2017-suv...

Fully agree, I don't understand why airshows need to have this level of improvisation, or actually any improvisation at all.

Even fuller context:

> Mozilla doesn’t sell data about you (in the way that most people think about “selling data“), and we don’t buy data about you. Since we strive for transparency, and the LEGAL definition of “sale of data“ is extremely broad in some places, we’ve had to step back from making the definitive statements you know and love. We still put a lot of work into making sure that the data that we share with our partners (which we need to do to make Firefox commercially viable) is stripped of any identifying information, or shared only in the aggregate, or is put through our privacy preserving technologies (like OHTTP)."

How is sharing data with partners in order to make Firefox commercially viable (i.e. getting money in exchange) not "selling data"? Anonymized or aggregated data is still data, and it's quite disingenuous of them to try to weasel it in by changing the definition.

> We are as close to not selling data as it is legally possible to be.

Normally when you say "as close to X as legally possible", that means you want to do X fully, but you can't because the law forbids you to. X in this case is "not selling data". But "not selling data" is not illegal at all. What are they even trying to say here?

(Also I don't find that sentence on their FAQ page)

The problem is more that they actually do want to sell user data (albeit anonymized and/or aggregated), and they want to present themselves as a privacy-favoring alternative.

They literally say:

> Mozilla doesn’t sell data about you (in the way that most people think about “selling data“), and we don’t buy data about you. Since we strive for transparency, and the LEGAL definition of “sale of data“ is extremely broad in some places, we’ve had to step back from making the definitive statements you know and love. We still put a lot of work into making sure that the data that we share with our partners (which we need to do to make Firefox commercially viable) is stripped of any identifying information, or shared only in the aggregate, or is put through our privacy preserving technologies (like OHTTP).

Sharing data with partners and getting money in return (that's what "to make Firefox commercially viable" means) is selling user data. They want to change the definition of selling data to exclude what they want to do, but they don't get to decide what the meaning of words is.

So they're not as privacy-friendly as they want to appear, and that's a difficult position, and that's why this policy allows them more access to data than a naive reading would indicate.

That text continues with this:

> We still put a lot of work into making sure that the data that we share with our partners (which we need to do to make Firefox commercially viable) is stripped of any identifying information, or shared only in the aggregate, or is put through our privacy preserving technologies (like OHTTP).

So they share data with partners, which helps to make Firefox commercially viable, meaning those partners pay them for that data. Or in other words, Mozilla sells user data. Even when anonymized or aggregated, it's still selling data.

Burns at perigee can only raise or lower part of the orbit though, right? A burn at perigee lowers or raises the apogee. To lower or raise the perigee, you need to burn at apogee.

Agree completely, I quite like the look of the font.

I sometimes feel who are deep in a subject sometimes are too entrenched in their world with rules and guidelines, that to me don't seem all that important to judge quality in the real world; or at least how I perceive things.

Another example also comes from the world of typography: text figures (non-lining numerals). To me they're ugly and difficult to read. Typographers like them because they fit better into the appearance of text, and that's true, but of low importance to me. Numbers are not words, and I feel they don't have to look the same. I like them to look different. I want them to be easy to read, which text figures aren't. (I actually went to the effort of creating custom CSS using Firefox's Stylus add-on to force lining numerals on all websites I visit.)

> The car has cameras to read speed signs and if it detects you are speeding beep at you. Sounds like a great feature! What's the problem if you aren't speeding?

If it's anything like the speed limit warning system in my Volvo, it's utterly useless. It gets the speed limit right about 80% of the time I would say. The other 20% it sees a sign "max 5 ton allowed" and thinks it means "max 5 kph", or it sees a sign in a side street and thinks it applies on the main street. Or a speed limit ends at a crossroad, which the system doesn't understand. The system often thinks the speed limit is 30 kph where it's actually 50 kph which is annoying; sometimes it thinks the limit is 90 or even 120 where it's only 50 or 70 which is outright dangerous.

The system is not to be trusted at all. There suggestions to make laws that would require cars to enforce these limits, instead of just showing the limits and warnings. I'm not against that in principle, but those systems need to get a lot better than the one in my Volvo first.