I think you have a point, but as others have commented "allowing functions" is not the problem, as fundamental math operations are functions. But if we limit ourselves to only functions that map (tuples of) integers to integers ((Z, Z, ...) -> Z), the spirit of the original game is retained. This disallows sqrt and log, but retains addition, subtraction and multiplication (but not division). Factorial (n!) is allowed, as is exponentiation to a non-negative power.
Wonder if someone could come up with general solution within these constraints.
I think that, if you are restricted to a finite list of n-ary functions, n>=2, each returning a single value, then you can’t do it, as you will only have finitely many valid expressions.
This may be easier to see in a stack machine / RPN model. An expression is a list of operations, drawn from a finite set, each of is either “push the number 2” or something that decreases the stack size by at least 1. And you need exactly 4 pushes. So a valid expression has four pushes and at most 3 other operations, because otherwise the stack would underflow. This gives a finite number of possible expressions, but there are an infinite number of integers, so it can’t work.
Are you sure the number of possible expressions is finite? Or is it countably infinite? I could in theory make an arbitrarily long expression out of the four fundamental operators, and because I can derive -1 from a finite set of operations it is possible to derive any integer from any even integer. You can also derive zero from the rules.
I didn't mean to say that n would need to be 2 or more, unary functions (I even mentioned factorial) should be fine. 1-tuple is valid tuple :)
And yes I think your analysis that only allowing n-ary funcs with n>=2 would make general solution very much impossible since you can only have a limited number of inputs.
Slack can't just willy-nilly hand over US government data to the public. There's a process that needs to be followed for FOIA requests as some classes of information are just not public. In fact, most of the data on a government Slack server would probably fall under those FOIA exemptions.
The purpose of this is to reduce manual work. In Finland, we've had mandatory electronic invoicing in all B2B transactions since 2020 (based on domestic standard called Finvoice which is very similar to PEPPOL.) It is really great. My company is 100% paper-free. When I get an invoice, it pops up to my accountants system and I just go and accept it with one click in their web UI. After that the invoice will be paid from my bank account on the due date and entered into the books automatically.
Gone are the error-prone days of manually copying account numbers, invoice reference codes and amounts from PDFs and paper mail, and scanning those invoices for book keeping.
The standards for e-invoicing are open but it's true that you need to hire a trusted intermediary to process your messages. Anyone can become a processor so it's not a closed system but I bet there's some auditting required before you get a license - which makes sense.
Overall changes like these initially mean some expenses to businesses but once the system is up and working as intended it reduces lots of mandatory pencil pushing type work, bringing savings throughout the economy to all companies.
You're right that the main basis is the automation part. That being said they could have gone with a system that does not require a yearly 2K fee to just be a part of the peppol group.
That's my only negative point about this whole system is that money is being pulled away from freelancers and small SME's for every invoice they send.
The big guys just setup an access point and pay the yearly fee as it is nothing compared to their revenue.
Most companies don't need to become processors, only those who specialise in providing financial services to others.
Unless you have a very small company you already probably have an accountant or bookkeeper. In most situations, the e-invoicing is provided by your accountant as part of their comprehensive financial management solution. I pay about 120€/mo. for a solution that includes bookkeeping, electronic and traditional invoicing, electronic and traditional expenses management, salary/payroll and all tax declarations for my 1-person company. The effort saved/cost ratio is bonkers, and in the grand scheme of things the monthly fee is marginal.
There might be indeed a usecase for SMEs that don't want to buy a comprehensive financial management solution and instead want to send their e-invoices manually. It's probably a pretty niche usecase and someone could probably provide a "proxy" processor for something like 100€/year. It's still a very marginal cost for almost any imaginable going concern. And to receive any money you anyway need a bank account, which is not free for companies.
> That's my only negative point about this whole system is that money is being pulled away from freelancers and small SME's for every invoice they send.
I don't pay anything to send or receive e-invoices on top of my monthly plan. But I do pay a very small fee to my bank for each transaction they process.
> I'm baffled that anyone even thought about using that, given there's such good mt76 support from mainline kernels with hostapd.
Not sure if you noticed but the OpenWRT 21.02.x series (based on mainline kernel 5.4 series) is affected, and these guys generally know their game when it comes to wireless on Linux. So much so that I think the mainline kernel mt76 driver is actually maintained by an OpenWRT developer.
Maybe MediaTek has shipped some modified versions of OpenWRT using this "wappd" thing to their B2B customers (as part of the SDK perhaps?) and are now advertising those as vulnerable.
Yes, I'm assuming that's exactly why OpenWrt is mentioned but it's very misleading.
The OpenWrt folks generally have good enough taste not to ship any drivers or userspace junk from vendor SDKs, though they do have a fair-sized set of backport patches on top of the (somewhat elderly) mainline kernels they do ship.
I'm running up-to-date mainline on my routers, not OpenWrt kernels. The mt76 support in 6.11 (and previously in 6.9 and 6.10) is complete enough that I don't need to carry any patches at all over what's in Linus' tree.
I live next to a highway on-ramp. The motorcycles are the worst, some of them are so loud while accelerating that it is impossible to have a conversation on my yard, over a hundred meters away.
The city is now investing tens of millions of euros to retrofit a noise barrier to protect our neighborhood. I think our house surfers the most from the noise. Yet - if it only wasn't for those loud motorcycles I would think the noise barrier is unnecessary. It would be much more cost effective to set and enforce noise pollution rules than to build expensive noise protection infrastructure everywhere where people live close to roads.
