foresterre

How do you use it with i18n and theming and templates? Do you define all of these outside of typst and (e.g.) add a shadow file to the typst compiler?

Asking because I've recenly been doing something similar, but since my typst files are templates, which with the same inputs can differ and may need different translations as well, I've written a small i18n library in typst (but I guess you could require a set of translations to be present with a given typst file as well).

While I agree with the sentiment, any answer would likely be vague (unless it is only interests based) since the models used to choose specific ads to be displayed can be rather complex.

I'm not trying to suggest that reviewing (code) can't feel (more) mentally demanding. The thing with at least reviews is that if you have worked through the problem yourself, you will also hit certain walls where stuff didn't work or wasn't as beautiful as you wanted (ie trial and error), and you have to activate your brain to think of other solutions, ie really think a problem through multiple times with active feedback. As a reviewer you don't get the active feedback usually.

Now this i think applies in lesser degree to the comment I was replying to. With maths, and some comp sci problems, you often can check an answer much faster than actually solving it, eg by using simple substitution for simple math problems. That is a useful skill, but a different one from solving the problem yourself. It employs different types of skills.

I've been in uni classes which used this concept (about 10 years ago though). It also requires that content is given in the right size and complexity each week, otherwise there will also be people coming to class who did do their homework but didn't understand it properly and just like the people who didn't read it need full explanations.

I also felt having multiple explanations (in a regular class where you could also read the book in advance) from regular lecturers helped me in gaining understanding because having something explained in multiple ways made sure there would be at least one way, or a combination, which made it 'click'.

I always liked the regular lectures where you would be required to do some pre class reading and a smaller set of exercises the most, as you would be engaged in the class itself and would be getting repetition with different explanations.

This sounds like it would come at the cost of the honest students who don't want to use LLM's and train their own understanding.

Reviewing is just not the same as working through a problem yourself. You often can take shortcuts when checking answers for correctness, which generation (with mind or LLM) of the solution can't take. This isn't limited to Math but equally true for programming and many other skills.

I find I still use the paid IDE's (RustRover and WebStorm), even when using LLM's for reviewing code and applying corrections. It's often just faster to do that, especially with an IDE which allows for quick non linear navigation through the files (following functions etc.). I think this is where JetBrain's IDE's excel. I can live with the high startup cost as long as it is quick afterwards.

I do feel however they have been neglecting anything non-AI for a while now. I in particular loved their auto complete, and with the new guesswork AI auto complete it's just not the same. I want deterministic behaviour from my IDE.

On the LLM side, I couldn't use the Anthropic subscription with their offerings, so I just use it from a harness outside the IDE, and switch to the IDE for review. This may not be JB's fault, but it still makes their offering useless to me.

I may be one of the few remaining but I also still program myself sometimes (I weigh it based on the task). The IDE has been pushing so much AI it has been getting in my way sometimes, and it is definitely using more resources for stuff I don't use.

I too am on the fence on whether I'll resume my subscription. Luckily for JB I dearly dislike VSCode's code navigation and interface, and would severely miss their debugging tools integration.

Just because something is publicly accessible doesn't mean you can use it for free, or that it gives you rights to do whatever you want with it.

I can access a public park, but that doesn't necessarily give me the right to also bike on its sidewalks, or walk on the grass, or take some of the plants home with me.

> ... or has been obtained illegally

Similarly, content that is _accessible_ publicly may be illegal to _obtain_, these aren't mutually exclusive.

On the internet, you'll find there are terms of services and licenses. These restrict how you can use even publicly accessible material. Public availability doesn't give you a license to use it however you want.

Far more likely Ireland, like it is for AWS. Could also be Paris, London, Amsterdam, etc.

> - Under a playlist you get a list of suggested songs, the UI element looks like a "regular song", but unlike every other "song" element the user has been trained to recognize, the "swipe right to queue" does not apply to these elements. For some reason.

There's a ton of these inconsistencies. For example, the new releases albums where songs become available over time also do not have it. It's infuriating how inconsistent the UI is sometimes.

Not so long ago they also had two versions of the "swipe right to queue" UI, one with long swipe (the current one), and one with a much shorter one. I preferred the shorter (with the hard break) one because it is much clearer when a song actually gets added to the queue. What often happens when you're adding multiple songs to the queue, if you only swipe a little bit, it is unclear if it was enough to add it to the queue. With one song this isn't a problem because a message with "added to queue" is shown, but for multiple songs, the messages are shown on top of each other.

Not saying that LLM development isn't fast, but doing a re-implementation of something you know in-and-out is almost always going to be faster, as you skip the probably most time intensive step of product research and requirement engineering.

I had this almost exact discussion today. Adyen (the payment provider) provides merchants with webhooks so they can update a local modal of payment and payment modification data (checkout).

But there is no way (for a merchant) to get the latest 'true' state as held by Adyen. So you better hope your data is exactly in sync with the notifications you got from the webhook (which it never exactly is, because there are so so many points of failures, and unlike what this author says, the docs aren't thát well presented to hold the same model as the PSP does. It is often close enough though, but you are constantly gardening your implementation, because the model also changes on their end with little information in the changelogs).

The "latest state" data exists though! If you open the customer portal it is presented to you without problem.

I don't think the article tries to argue against the good sides.

In my opinion it asks us to not just look at individual gain of using LLM's (which can be powerful), but also place it in context how we as humans built long-term sustainable societies. Most societies are currently ignoring that part for the most part, because they fear missing out.

Probably the hardest problem of LLM's is their ethics. As the author put it:

> Modern multi-billion parameter AI models are scaffolded on and made possible by the largest heist in human history: theft of everything that could be scraped from every corner of the digital spaces we share. Without prevention of and justice for this damage caused by current models, their use is highly fraught, ethically. We, as human beings, have developed complex social forms of intelligence when it comes to dealing with things like credit and provenance, two things that modern models are incapable of.

No one currently seems to have serious solution for this. Solving it the right way would be choosing the hard path. While some artists or writers are fighting for their rights, us programmers are collectively ignoring the issue. It is funny how most companies (and some individuals) only want to use LLM's of they don't train on their code. But is just fine to use some 40 years of collective open-source without any compensation or even acknowledgement.

This side of the story doesn't make LLM's less technically impressive, nor does it make the tools themselves less powerful. But unlike the industrial revolution, which some like to compare the AI boom to, during the industrial revolution humans built stronger productivity solely on innovation without strip mining all intellectual property in human existence.

I can see the benefits of LLM's, enjoy the technical capabilities and be impressed by the technology, and at the same time find these tools wholly unethical in their current form.

In addition, there's about 2x year a 25% sale and you have a year to activate the passes.

The in southern Europe (e.g. France, Spain, Italy), required seat reservation is most common and most expensive.

I don't mind requiring seat reservations, but that it is separate from the ticket price and significant (eg 15€/seat reservation in Italy), feels like price gouging. It also feels different from say the optional (and way lower priced seat reservations in German ICE's (high speed rail)). I rather pay for a "high speed rail supplement" instead of seat reservation haha :).

I interrailed last year through Germany, Austria, Slovenia, Italy and Switzerland.

In Germany I was lucky, I only had a small delay on the way back.

Austria was exceptional in everything. On time, modern trains and facilities. I guess the food on the train was expensive and bland, but I've never seen a train where that's different.

Slovenia was the weirdest and had the most delays. Train cars for which I had seat reservations consistently didn't exist or arrive. They use old stock, but that also made it kind of fun and there were great views. I couldn't rely on the time table though.

Italy has lots of high speed rail, but required (paid) seat reservations. The problem is that for almost any medium-long distance there's no slower speed alternative. The normal speed stock is fine (can be taken to go to smaller cities) and was generally on time.

Trains in Switzerland are exceptional too. Funnily enough, I did have fairly significant delays 2/5 times.