lblume

13 ms·

151.
by lblume 1y ago
It's very much worth it, but absolutely not necessary. You can always generate images/figures beforehand and include them in Typst. There are also packages for various styles of diagrams, e.g. diagraph ships with a Graphviz render
152.
by lblume 1y ago
I found it much easier to use Typst to create any of these documents than any other software I used before. Just loading data from a JSON and incrementally regenerating the document on each change is game-changing for me.
153.
by lblume 1y ago
They do now, experimentally. The support is getting improved a lot currently.
154.
by lblume 1y ago
https://rickiheicklen.com/unparalleled-misalignments.html
155.
by lblume 1y ago
People still use the web without adblockers?
156.
by lblume 1y ago
Sure, but nonetheless whether the model is called "small" at some time t should depend on the parameter count and t, not some arbitrarily specified metric of deployability.
157.
by lblume 1y ago
> the simplest and most friendly effects library Highly unlikely. JavaScript simply isn't a language built with this kind of evaluation model in mind, so an external library introducing completely orthogonal concepts surely should n
158.
by lblume 1y ago
It might, if the effect is handled using these functions. You could also handle the effect using another handler, doing something entirely different. It is up to the caller how to handle effects, or propagate them up.
159.
by lblume 1y ago
I agree that this may sometimes be the case, but in this case the author does seem to have very strong points in favor of the proposal. I also think that the accusation of dynomight not understanding NumPy enough is unjustified based on the
160.
by lblume 1y ago
I do not — parameter count is objective, practical impact depends on such a multitude of factors that any comparison becomes virtually meaningless.
161.
by lblume 1y ago
> Better security here doesn't seem hard to implement. You seem to assume it would be very simple to implement this — how do you come to this conclusion? My priors would suggest that the vast amount of effort that went into the Sign
162.
by lblume 1y ago
I tried giving this to ChatGPT. (Just by uploading the image to the base OpenAI interface.) I expected to either (a) have the model already know the question and give the right answer, (b) hallucinate an answer or (c) refuse to engage with
163.
by lblume 1y ago
One of the strongest example for me is newtypes. I strongly believe in Parse, don't validate ( https://news.ycombinator.com/item?id=21476261 ) and think Rust's newtype pattern is able to really enforce this strongly
164.
by lblume 1y ago
I can see why people don't like to write Rust. It is more tedious, harder because of the ownership model, and has a very steep learning curve. What I don't get is what people have against programs written in Rust from the bottom u
165.
Beating Floating Point at Its Own Game: Posit Arithmetic [pdf](johngustafson.net)
3 points by lblume 1y ago|0 comments
166.
by lblume 1y ago
They absolutely can if you prompt them to. You can even add it to your system prompt for it to happen in every new conversation!
167.
by lblume 1y ago
Since LLMs are basically linear algebra all the way down, this is vaguely reminiscent of how human brains also have a very hard time understanding neural circuitry despite literally being made from it.
168.
by lblume 1y ago
What percentage of people are actually experts at their jobs though?
169.
by lblume 1y ago
If a human reviewer rejects a proof, what do they consider apart from local errors? Formal proofs have formal local errors (which can be checked mechanically), while informal proofs have informal local errors (which require humans and some
170.
by lblume 1y ago
It defaults to it, but the preferred language can be changed in all major browsers, though currently not on a per-website basis.
171.
by lblume 1y ago
One example is qutebrowser: https://qutebrowser.org/doc/help/settings.html#content.heade... I assume other developer-centric browsers have similar options.
172.
by lblume 1y ago
What incentives does Google have to improve UX in this way? I absolutely agree that it should be the case, but the people for whom it matters are (1) completely insignificant wrt to the whole user base and (2) mostly care about tracking and
173.
by lblume 1y ago
That should not be the business of the site, but of the browser. It is precisely what Accept-Language is for. The site should respect the header, allowing circumvention should be a very low accessibility priority.
174.
by lblume 1y ago
Then create some status bar (or even a popup if necessary) that informs the user about their Accept-Language header-IP address mismatch (in terms understandable to a non-technical user), offering the option to choose either page.
175.
by lblume 1y ago
Bold of you to assume Google would actually show your their most precise and accurate guess.
176.
by lblume 1y ago
Modern browsers with auto-translation features always offer the option to "Never translate X" — never translate a language the user speaks. If websites really need translation (which they very likely do not!), this should be the n
177.
by lblume 1y ago
> requires knowledge of the proof itself (in general) Why? If a proof is wrong it has to be locally invalid, i.e. draw some inference which is invalid according to rules of logic. Of course the antecedent could have been defined pages ea
178.
by lblume 1y ago
The information is also more relevant. How many lines a function or block has might make a difference, but the absolute file position is, more often than not, irrelevant.
179.
by lblume 1y ago
Could you not compile them on the voting machines itself? But yes, there is always going to be some level of trust involved, and the bar for manipulation seems to be lower than re: manual counting.
180.
by lblume 1y ago
Yes. It might sound counter-intuitive or even ironical but the next logical step is to make the type system more, not less expressive and to add completely dependent types. Every language already has them to the extent of Array<T, N>

More ›