That's me. But I must admit that I prefer Cholula Hot Sauce.
In addition to the imperative `foreach` construct, Flix has two constructs for applicative[1] and monadic[2] comprehensions: `forA` and `forM`. Since applicatives and monads are related, it is useful that their syntax is similar, since it makes it easy to switch between the two. While having camelCase keywords may seem strange, in this case there is a feeling that it works out well. Certainly, `form` or `fora` would be much worse.
Applicative and monadic programming is not a big part of Flix, but it is something we want to support and make ergonomic. Also, these features may have scary names, but the concepts are not too difficult. See [1] and [2] for simple examples of how these features can be used for e.g. error handling.
Gotta be honest, this isn't very inspiring : "Wait, division by zero is zero, really?
Yes. But focusing on this is a bit like focusing on the color of the seats in a spacecraft."
Curious if anyone can weigh in on why Flix requires a developer to explicitly mark a function as pure. I'd imagine in almost all cases this can be derived through static analysis.
I could be wrong, but the sentence "Flix precisely tracks the purity of every expression in a program." together with some examples of function definitions without the purity/impurity annotation, gave me the impression it's optional, because the compiler can infer it on its own most of the time.
Being able to buy groceries with euros is not a strong, or even related, argument to the point you're making. Even ignoring the real economic cost of having two currencies, there is no serious economist that would argue Poland joining the EUR zone is negative for Poland (as opposed to for stronger EU economies). Every single historic metric points in the other direction. Poland's economy is maturing and strengthening and median household income is slowly reaching parity with neighbouring countries so this particular argument may only hold for maybe another decade but until then it's a bit misguided. Also, not for nothing but the Polish economy is mostly doing as well as it does because of the metric ton of EU subsidies injected into it the past decades. Poland is one of the largest net receivers of EU money since 2004 so arguing it or its single currency was somehow a net negative to Poland is, and by extension an odd argument to make for someone benefiting from the above as a Polish citizen.
You have to be careful to make a distinction being in the EU and using the Euro as your currency. You can benefit from the EU common market, with its uniform rules/standards, easy capital flows, subsidies, and industrial policy. All without using the Euro as your currency, and being subject to the monetary policy of the ECB.
Being in the EU without using the Euro has been pretty good for Poland.
They are free to be. When joining the EU, Poland accepted to join the monetary union at some point, but there is no date set, so in theory it can be infinite. And with its current strategy to drastically modernize and increase its military force, there is no doubt it will not met the financial criteria anytime soon so nobody is worried about that.
> Even ignoring the real economic cost of having two currencies, there is no serious economist that would argue Poland joining the EUR zone is negative for Poland (as opposed to for stronger EU economies). Every single historic metric points in the other direction.
The inability to set monetary policy is a strong argument. Just ask Greece (and Spain):
Also, not for nothing but the Polish economy is mostly doing as well as it does because of the metric ton of EU subsidies injected into it the past decades.
That's one part of it. The other part is that the country had just as much human capital and economic potential as Western European states, but was held back artificially by the Partitions, WW2, the Soviet Union, and the lack of Marshall Plan investment that Western Europe received.
Sorry, but the narrative of "Poland is only doing well because the EU is helping" (of which German companies are benefitting from tremendously) is a historically narrow way to analyze the situation.
Well, it's a fact, not a narrative. I think the actual debate is whether or not boosting economies of countries in the EU that for one reason or another were or are behind on the curve is a net win. I think it is. Just as it is perfectly justified for other countries to feel a bit hard done by if that same country after an estimates 280 billion euro injection is still rejecting the shared currency of that body.
The point is that "feeling a bit hard done by" rings a little hollow if the country feeling this way is the reason the recipient country needs investment in the first place. 280 billion euro is still less than a third of the estimated amount of damage from WW2 alone.
So far there are zero data points on the success side of the fence when it comes to visual programming. The exceedingly rare times I come across it professionally it's a toy project or something that's basically in the state of "We wish it was just code but we don't dare touch it".
In the era of vibe coders, visual programming seems like an interesting way to build consensus with an LLM when making a decision. LLM shows you a visual for its execution and the vibe coders feel like its approachable enough to tinker with the visual to improve execution.
I own the meta glasses. I enjoy them. Last night, I asked meta to look at the menu and calculate the best meal by price. It calculated that the protein in the chicken Marsala was the best deal for the price. I then asked it to look at the wine menu and tell me a good dry wine that wouldn’t be too expensive. The wine list was over 100 wines and the answer retuned in seconds. I enjoyed the meal and the wine. I will continue to use the glasses in 2025 as I’m finding creative ways to use them.
How does it know the affordability of a menu item? If it’s only looking at a menu the odds are that it doesn’t have a lot of information such as serving size and calorie cost. You just asked an AI to give the statically correct item for affordability.
The wine it chose was how far down the list? Was it chosen at random or did it just match “chicken wine” with the first result?
This seems like you’re assuming the glasses help you decide but really it probably did what anyone would do and pick a random option or just give you the popular choice. Why even use glasses for that and instead just point somewhere on the menu?
It reasoned its choices. Told me the vineyard, what most people say about it. It gave me 3 food options originally and I said go by best value. It would be the same as if a friend recommended it to you or a server. How do I know the AI is giving me good advice? I don’t, but I’ve lived on earth long enough to know it’s not bad advice. End of day, the human makes the decision. We don’t lose agency unless we just say ok to everything. In this case, it was reasonable other cases maybe not. We still make decisions even if we use ai.
So in the spirit of not losing agency, out of the 3 items you told it best value, which appears to mean “lowest cost”. Best value can also mean portion size relative to cost. Further even relative to your diet restrictions and relative to the area of economic activity. All of those things I garuntee weren’t involved in the LLM “reasoning”. It’s okay to be wowed by sleight of hand LLM tricks, but expect everyone to be skeptical when you describe the situation.
The above criteria is how I would evaluate best value of menu items. If an LLM just gave me the cheapest of 3 dishes I wouldn’t use it because it is not adding any agency I already had.
Here’s another possible scenario:
The LLM retrieved “best value” or some relative synonym from the online review of the restaurant. The review says “I think the chicken marsala is the best value.” The review was posted 5 years ago and the portion sizes have changed when the new owner wanted to shrinkflate the food. Is chicken marsala still the best value? Was it ever the best value?
I think it’s clear you don’t understand llm’s and how they operate. The data is trained on orthotic patterns. Your description would be the contrary of real world llm.
I'm trying very hard not to be judgemental, but this comment fascinated (or horrified) me.
I grew up with some relatively minor food insecurity, and so outsourcing my food and beverage choice to a soulless machine sounds like my own personal hell.
All that fancy technology just to address an optimization problem that's entirely solvable with a can of Vienna sausages and a bottle of Two-Buck Chuck.
Unless I'm really watching my budget (though, why would I be eating out at a place with a huge wine list and buying Meta glasses if that's the case?), I cannot possibly imagine the joy behind having a nice meal based on what AI told me is "the best deal" rather than trusting my own palate/gut and ordering what I know will bring me the greatest pleasure (or working with the sommelier to discuss my tastes and find something best-suited to both myself and my choice[s] of food).
You do you, by all means, if it brings you the most joy. But removing my own agency while allowing a company to track my entire dining experience to better serve me ads sounds so... hollow and dystopic.
And now Meta knows your location at that time, can use that lovely menu to derive ads better suited for its competition, has data on when you like to eat out, who else was there (let's not forget about DeepFace!), etc. While you ate your meal, you fed some juicy tidbits to their data repo on you and those around you.
Do you really need a pair of glasses to tell you what to eat and drink? Wow, that must be the latest techie way to order food. A friendly reminder: you don't own those glasses, Meta does.
Never thought that once we had Skype we'll still not solve video calls cross-platform on every bloody device by 2025, including, but not limited to, linux, TVs, Android boxes, Chromecast with Google TV, and so on.
I really want to get off Whatsapp, but everyone simply refuses to do so out of simplicity, lazyness, and ignorance.
Signal is the easy and obvious alternative.
XMPP/Matrix are different paths that also work (and both rely on TURN/STUN, like everything), but that requires someone running those servers in your vicinity of friends.
There are plenty of people who are aware of the "trade-off" of using their products but are still happy to do so. Just because they've come to a different conclusion to you doesn't mean they are in some way "wrong" to do that.
I know people are going on about privacy and that's fair but easily sorted with the appropriate data privacy policies and actually adhering to your commitments in them.
If you are approaching this from a potential business perspective that will not be the major issue. As you can tell from the comments most people do want their screen recordings on a remote service, now matter the policies involved.
That leaves you with some options :
1) Encrypt the outgoing WebRTC stream before it leaves the client device/browser (e.g. through WebRTC Insertable Streams or similar), encoded with a client local key that only your customer knows/manages. This ensures any screen recordings stored on your end cannot be accessed in any way. They can then download the recordings as needed and decrypt on demand. Note that this still has the implication of you storing and transmitting massive amounts of data which is likely to become cost prohibitive as a service.
2) As above but do the decrypting for your customer on your end so they get to download playable media. This obviously introduces the security risk of your platform becoming temporarily able to 'see' the content.
3) Save your recording directly through MediaRecorder WebM or similar (has various browser compatibility challenges as standards materialize) and uses less obvious encodings (we're still living with the bullshit of H.264 requiring licensing).
