Flundstrom2

There's a lot to comment on that post about strange conclusions.

Yes, C has an ISO standard. It took more than 15 years from the conception of the language. Same with C++. Before that, there didn't even exist a normative compiler. And if you want a certified C or C++ compiler, you have to pay a lot, or use the "it is wildly used with a well-known set of bugs".

Either way, there's a huge load of constructs that are Undefined Behavior which you need to know about and avoid. The discussion about the borrow checker being hard to understand and work around, only highlights a belief that is different from what you need to do in C or C++ anyway - but without help from the compiler.

And cost? Heck, Visual Studio Pro cost twice that of Ferrocene. The Enterprise version starts at 10x the price - and those aren't certified.

Then there's the discussion about proving to the compiler that the code is correct - while at the same time dismissing proof using formal verification, that in turn is an entire language and alphabet of symbols that doesn't even exist on a keyboard.

Yes, the Rust syntax could have been easier to read, to that I agree. And Rust is no silver bullet. But in general, the article could benefit from a review.

Read Daniel Pink: "Whole new mind".

Whatever you do, someone else will be able to do it faster and cheaper.

Also, read up on the "Red ocean" vs "Blue ocean" theories about competition.

Yes, you can compete in the cut-throat red ocean, where all other competitors try to survive by offering roughly the same product or service. Look at basically every AliExpress shop.

But it will only get you so far. Will they survive in the long run? To really succeed, you have to offer something unique. Production efficiency is of course on everyones radar. No company ever said "we dont care about how much it cost to produce our product/service".

Google wasn't unique when they started. There were several other search engines out there, Alta Vista being the largest one. Both offered their services for free - so how can you possibly compete on price in the Red Ocean against all other search engine? No matter how efficient the execution was, they had to attract people. Because that was what they were selling. People are the raw material. So, Google offered a different experience than all other engines. That way, they eventually got more and better raw material (people) to sell.

Apple tried what everyone else did back in the early 2000s; Sell mobile phones. But they did not become the world's second largest handset maker by being the most efficient at developing or producing them. In fact, they deliberately sold their phones at a loss for the first years. They decided that they would provide a completely new experience, and that eventually people would be willing to pay a premium which turned out to be twice as much than an android phone with comparable hardware features and costs.

SpaceX, on the other hand, have gotten the world leader in rocket launcher thanks to outstanding execution ; instead of destroying every rocket once the payload has been deployed, they are reusing the rockets, thus keeping the cost per launch much lower. Noone else have yet been near their reusability and cost per launch - because its litterarily rocket science, not because the competitors are less efficient in their execution.

Thomann and Behringer are some of the world's largest companies in the music instrument business. They have perfected execution to such degree that it is hard to find cheaper alternatives that are comparable even on AliExpress!

Ingvar Kamprad started his first company by buying pens in big-pack, selling them piecewize at a large markup. Later the company decided - similar to Behringer - to copy the design of well-known furnitures, but optimize the logistics, eventually also optimizing the manufacturing process. IKEA is now the world's most famous furniture company. But, even they can't sell the cheapest bookshelf anymore.

The irony that an article about Germany is unavailable to EU citizens because the publisher don't want to comply with GDPR. I think I'll pass on that publisher.

451: Unavailable due to legal reasons We recognize you are attempting to access this website from a country belonging to the European Economic Area (EEA) including the EU which enforces the General Data Protection Regulation (GDPR) and therefore access cannot be granted at this time.

If the US wants to keep its population down and medical costs up by banning safe products, that's up to the US to decide.

At work office; three 1080p screens+laptop. Allows me to keep 6 browser windows+outlook/teams visible,or 4 browsers+vscode+outlook/teams.At my home office; a single 4k and laptop. Same at home.

Neither combos are perfect. Especially whenever I undock for a meeting or move between work and home office, there's always a mismatch between window placements.

I'm using virtual desktops to separate work for different projects, but there's two critical programs that don't allow me to be properly (in accordance to my needs, that is) opened on several desktops at once; Word (where I keep my diary) and VS Code (!).

No way.

Java and Rust are the only language that I know of that has proper handling of exceptions; mandatory declaration as part of method declaration, since exceptions ARE an integral part of the contract. (Yes, I consider the Result<T, E> being corresponding to exception declaration, since the return value MUST be checked prior to use of T.)

I don't know; My privat computer is still good enough for what I use it for, but it's too old according to Microsoft.

At work, the most annoying thing was the ads, but after disabling then, WSL works just as fine on W11 as W10.

I simply fail to see any benefit of upgrading - apart from security patches.

Three truths:

Most people at healthy sceptical of Ai quality and accuracy.

You don't want to pay your employees taking courses where they learn falsehoods or pure bad practices. Combine that with the above truth, and you're up for a big uphill in getting sales.

If you can't sell your product or service at a cost less than the €100-200 (usually the highest amount the lowest level manager can approve), you need to add a zero to compensate for the increaseing cost of obtaining a customer due to the overhead of getting the next level manager's approval. Since you did old fashioned courses for the military industry, you never had that level appear as a Blocker, since they are used to pay.

I can't offer advice, especially since the edu market is really saturated with very varying quality.

Who doesn't use a spellchecker in Word? Use the grammar checker (if available)?

Even those small tools have shown to degrade the quality of writing (unfortunately, I didnt save the link) on a broader scale, although there are less spelling and grammatical errors.

And sure, a Reuters notice is just as boring and inpersonal as a piece of code. But even if you offload the grunt work of the writing, putting at least some effort in making personal will raise it from spam/click-bait level to potentially worth reading. It will make the text stand out.

As for me, English is not even my native language.

I'm learning Rust by writing a multi-platform multi-player football manager game. I'm "only" using occasional LLM queries and autocomplete, which helps with the boilerplate code and making the Rust compiler happy.

Then, there's a small MIDI-utility hardware built on a small Adafruit featherwing that I intend to rewrite in embedded Rust.

The repo doesn't contain what the doc says it should contain

Yep. From the viewpoint of an given photon, it touches both its origin and its final destination - even if it is at the edge of the universe - at once. So, from the photons' point of view, we still live in the Big Bang and time does not pass, never has and never will.

This Guide is intended to contribute to a better understanding of EU product rules and to their more uniform and coherent application across different sectors and throughout the single market.

This Guide applies to the EU Member States but also to Iceland, Liechtenstein and Norway as signatories of the Agreement on the European Economic Area (EEA), as well as Switzerland and Turkey in certain cases. References to the Union or the single market are, accordingly, to be understood as referring to the EEA, or to the EEA market.

This Guide discusses non-food and non-agricultural products referred to as industrial products or products whether for use by consumers or professionals.

No contact in formation, ToS and Privacy Policy basically empty, no customer support links, no company information apart from a name.

Regarding jurisdiction of the ToS only states "These terms shall be governed by and construed in accordance with the laws of the jurisdiction in which Nimrobo AI operates"

Where is even the company registered, where is the data hosted, what about GDPR?

No thanks, I'll pass on this one.

Maybe consider using EU-based alternatives such as Mistral Le Chat, and StackIT. That at least gives privacy and transparency mandated by law.

But yeah, migrating the core business systems to cloud surely adds risk of lock-in, as well as juridical and geopolitical risks.

Interesting concept.

Although, when we get above the epic/ticket level and onto product management, we inevitable get into the process, product and project parts as well, and eventually we also get into the division of the R&D department into teams and what not, which makes a standardized format challenging. Not impossible, but challenging.

However, there are a few fields that wound benefit from having a stricter - but customizable - definition through some additional configuration file; and specifically status, priority, type.

There's also inevitable going to be different roles needed for different companies, such as creator, owner, assignee, sponsor, so those distinct fields (except maybe assignee) would rather be an array of key/value for the roles.

Only looking at the proposal as-is, I lacked a closed_at field.

The labels and estimate are also challenging; what to accept? Some prefer t-shirt, others story point and some actual durations. And who is allowed to define those?

One thing I've had issues with in all kind of system, are the separation between mandatory, optional and useless fields. In many cases the useless fields are way too many compared to the fields that are actually used for a specific team. Generally, fields tend to lack clear guidance in the tools on who owns them, what to populate them with, and why they exist. Elon Musk once stated that every single requirement must have an owner, and IMHO, so must every single field in the ticket tracking system.

Those things are both company/project/product and team-specific, just as the transitions between the states.

And speaking of transitions, the interesting thing to track against, is rarely the current snapshot of the currently know the requirements, but the changes to the scope, which in turn reflects (from) the company's launch plan.

To generalize, the scope is either increased, decreased or clarified, and that changes all the way from the initial vision statement to the "final" delivery before the EOL and decommsion project has completed.

But, as I mentioned; an interesting concept!

C is 50 years old, and deliberately kept smalm as a KISS language (unlike the almost-as-old nowadays-bloated C++).

There's simply not a lot happening to it - apart from being part of the root causes to 50% of all security vulnerabilities.

Although Gcc and the Linux kernel do have some 50 MLoC, noone would nowadays embark on writing such a large system using C from scratch.

Today, Java, C# or Rust would be much better choices for any large system. Heck, even C++ (with proper use of a selected sub-set of its features) would be a step up.

Dont get me wrong; I've been doing embedded development using C and C++ for a living for the last 25+ years. But, I'm learning Rust, and I love it, hoping it will take C's place and being a serious contender to Java, C# and even python for the use-cases where it would make sense.

Surely, getting a program through the compiler is much harder, but afterwards, it just works. I would suggest all seasoned C developers to give Rust a serious shot.

A couple of days ago, the new EU regulation on disclosing energy efficiency for mobiles and tablets using an extended version of the A-F rating familiar to all European consumers.

https://single-market-economy.ec.europa.eu/news/new-eu-rules...

All manufacturers are required to disclose the time to depleat as well as number of charge cycles before the battery is down to 80% in the searchable database

https://eprel.ec.europa.eu/screen/home

Not surprising, Samsung has phones running circles around xiaomi's best performing phones, with battery times almost 3 times that of X's, and number of charge cycles twice that of X's.

Kind of reminds me of the self-driving cars of the early 80`s. Yes, it worked! I a van, crammed with racks and racks of then-state-of-the-art electronics. But, as someone back then mentioned; While the car crawls at 30 km/h on an empty road with clear painted lines, it was more like driving 70 km/h with closed eyes for 5 seconds, before opening them briefly and then closing them again.

The I would say, we're doing the same now, with loads of electronics, even more power consumption and hallucinating models. Working, but don't bet your life on it.

I asked it to generate an image with the same prompt I gave ChatGPT yesterday. And got the exact same image as yesterday.

What you are writing about are several different things; first of all, you're writing about open information - because truth is always a simplification of information. Information on the other hand, is man-made statements which some people believe are true, some believe is wrong, and some believe are deliberate lies, with or without malicious intent.

And any piece of information can be all at once, depending on from which side you look at them.

But, that detail aside; Yes. We need information to be free. We need to emphasize information that - as a whole - gives growth and prosperity to humankind as a whole. We need to suppress information that - as a whole - delay or reverse human development.

But I don't mean "hide" or "stop" such information.

Let's say dictator X have stolen 10% of her country's wealth for personal gain. But during her reign she managed to lift the standard of living, increase health of the common and the poorest tenfold. Both pieces of information are "true", but depending on the conclusion drawn from each of them in isolation, that information can be used for both good or bad.

That's the philosophical part.

Then comes the ethical part; Who gets to decide what information is good for mankind to emphasize, and what is good for mankind to suppress?

Not every single individual on earth have the empathical or intellectual ability to deal with all kind of information with an open mind in a way that is - for mankind as a whole - positive.

And once again; what is good vs bad for humanity also depends on how long time you're going to consider; What is good for humanity for the next 100 years might be bad if we continue on the same path for 1000 years. But that might be a hypothetical judgment because we don't know what changes it might bring in 1500 years. Let alone 15000 years.

Then there's the technological part. That's the easy one.

The technology to verify the origin of information, and to verify how it has been altered, to ensure it is resilient against accidental and intentional removal exist and is well-established. Block chains, torrents, onion etc are just tools. The most difficult part is to provide the means to give access; If a state decides to simply deny access to internet, you will be back to the 80's way of sharing information, and no matter how perfect the protocol is, it doesn't matter if there's no infrastructure on which it can run.

I would suggest putting it out as open source with a permissive license that don't require upstream commitment.

Because you don't want to become a maintainer. Just make it clear that it is provided as-is, without support.

It does after all represent a lot of value having been poured into it, worthy of a better ending than rm -rf, even if it didn't reach break-even.

Is it legal to teach kids and authors to write, by having them read lots of books?

Reading a lot is a well-known strategy to become better at writing. Thus, I argue that any reasonably skilled LLM that use the scraped text as input, but not produce verbatim copies, should not be considered as violating IP law.

Time is a mess. Always. The author only scratched the surface on all the issues. Even if we exclude the time dilation of relativity which affects GPS/GNSS satellites - independent of if it is due to difference in gravitational pull or their relative speed over ground, it's still a mess.

Timezones; sure. But what about before timezones got into use? Or even halfway through - which timezone, considering Königsberg used CET when it was part of Germany, but switched to EET after it became Russian. There's even countries that have timezones differenting by 15 minutes.

And dont get me started on daylight savings time. There's been at least one instance where DST was - and was not - in use in Lebanon - at the same time! Good luck booking an appointment...

Not to mention the transition from Julian calendar to Gregorian, which took place over many, many years - different by different countries - as defined by the country borders at that time...

We've even had countries that forgot to insert a leap day in certain years, causing March 1 to occur on different days altogether for a couple of years.

Time is a mess. Is, and aways have been, and always will be.

As a developer in the EU, I don't find the EU regulations burdensome. They mostly represent current best practices such as "don't electrocute or (for PCBs, plastics and metals) poison the customer", "make sure your products don't break within a couple of years", "don't hardcode 'admin' and 'password'", "provide security updates as needed" etc.

Sure, it do put a burden on solo developers to know and be able to prove they follow the applicable CE regulations, but that is equally hard independent of where in the world you're located.

Sourcing components from an EU country, or at least requiring the components to fulfill the RoHS directive, using led-free soldering gets you a very long way towards CE compliance, as well as designing for USB-C and using external AC/DC adapters to power the product.

From today, new EU energy labelling rules apply to smartphones, cordless phones and tablets placed on the EU market.

In addition to the well-known A-F marking, the number of charge cycles for the battery and its initial life, as well as scratch-, drop and water resistance will be prominent in displayed on all products.

This is quite interesting, and might come as a surprise to some non-European marketers:

Consent frameworks can’t just check a box — it’s a wake-up call for anyone still treating compliance as a technicality rather than a structural shift.

Interestingly enough, there's always a lot of trackers - sometimes hundreds on a page - and that aks the consent for "legitimate reasons", in addition to all those who don't claim legitimate reasons.

If they indeed have legitimate reasons, they shouldn't need to ask for consent!

The EU ought to enforce DNT and some new X-Cookie-Consent: Essential-Only header.

This post distills research paper Asterinas: A Linux ABI-Compatible, Rust-Based Framekernel OS with a Small and Sound TCB, which is to be published at USENIX ATC 2025. The preprint can be found on arXiv.

There's nothing worse than UI trying to do data validation...

The number of form validators that still barfs at anything that is not ACSCII7 isalpha() amazes me. Heck, the entire US is built by non-indigenous people! Internet domain names, HTTP 1.0 and HTML/2.0 since the RFC 2070 extension have mandated the support for unicode for almost 30 years - that's longer than the age of many professional developers.

Swedish social security numbers? Why do forms even allow the obsolete 2-digit year format after y2k?

Swedish postal codes have since 1968 by standard been written as 3 digits, space, 2 digits. I have rarely seen any system allowing a space - even UK postal codes requires space (and non-digits, too).

STOP VALIDATING USER INPUT unless there are strict INTERNATIONAL standarads explicitly EXCLUDING a specific format!