wenc

39 ms·

541.
by wenc 2y ago
How is Quip different from Google Docs? Neither are discussion tools. (I use Quip a lot) Quip has almost exactly the same feature set as google docs — which part of quip were you referring to?
542.
by wenc 2y ago
On one hand, high compile times make you think about your code more deeply. No one will dispute that this is a good discipline. On the other hand, computers can check so many more cases more quickly and more correctly. I think there's
543.
by wenc 2y ago
I read In Praise of Idleness when I was 13-14 because an older person recommended it to me. Although the essay was written in the 1935, he convinced me this was the future, and that it foretold what European life was going to be. And indeed
544.
by wenc 2y ago
VS Code is electron and it has been optimized to the hilt. I have used it on Mac, Windows and Linux and the experience is uniformly great. It feels native on every platform (responsiveness is not as snappy as vim but it’s good enough and I’
545.
by wenc 3y ago
I know Shopify uses GCP. I'm curious, folks who use GCP, how is it?
546.
by wenc 3y ago
I have used brew trouble free for many years.
547.
by wenc 3y ago
I do read for entertainment, but I’m talking about reading to learn (very common use case) rather than reading fiction. I don’t take notes when I’m reading fiction like “A Man Called Ove”. But I do take notes and flip back and forth when I’
548.
by wenc 3y ago
I think Moby Dick is a romance. I guess the term is confusing in English because people think “love story” but an older use of the word just means a novel with adventure, chivalric elements. There is also a giant implicit “for me” in your p
549.
by wenc 3y ago
I read paper books because I can flip back and forth without friction, and I can underline and scribble. I don't know why eReader interfaces make these two functions -- critical for deep reading -- so difficult. Kindle is optimized for
550.
by wenc 3y ago
> There is this false idea that setting up an AWS or Google or Facebook competitor is in some way difficult. I personally think it is difficult due to scale and economics. That is not to say Europeans can't grow a competitor in clou
551.
by wenc 3y ago
Would it be analogous to say instead of having a single Von Neumann who is a polymath, we’re posing the question to a pool of people who are good at their own thing, and one of them gets picked to answer?
552.
by wenc 3y ago
1. This should be tagged (2011) 2. Agree with the title and that cleverness should be avoided. 3. Do not agree with the last sentence: “Instead, my lesson has been: write Python code like it was C.” Python should be written like Python.
553.
by wenc 3y ago
I use Firefox daily (have done so since 2016, switched over from Chrome). I have encountered exactly 1 issue. This language learning plugin only has a Chrome release: https://www.languagereactor.com/ Other than that, I have
554.
by wenc 3y ago
From the article: > The total cost for installing full-height gates in just those 128 stations would be over $7 billion, the agency found, and half-height gates would cost $6.53 billion. Once installed, the agency would have to spend mor
555.
by wenc 3y ago
Apparently python meat is very tough, even after pressure cooking. Guga ran an experiment: https://www.youtube.com/watch?v=bGCYcotfUN4 This makes sense, considering how strong they are. If someone figures how to cook Python
556.
by wenc 3y ago
This is not my lane, but I was looking to fast text search some time ago. I looked into Elastic but it seemed to me it was designed for a scale that I didn't need. Then I found projects like Meilisearch and TypeSense (which are in the
557.
by wenc 3y ago
And compete with CUDA? It's not the hardware that's the lockin but CUDA.
558.
by wenc 3y ago
The Seminary Cooop on the UChicago campus is excellent — I spent many afternoons there. Also there are many book events by famous authors. Down the street is the 57th St bookstore which is less academic, and also a great place to find more
559.
by wenc 3y ago
Sisu?
560.
by wenc 3y ago
* Logitech (Swiss) devices. * Ericsson (Swedish) communications, including 5G equipment. * ARM (UK) processors. OK these are used by non-so-average people: * Python (Dutch) programming language. DuckDB (Dutch). Polars (Dutch).
561.
by wenc 3y ago
Me too. Folks these are physical clocks with retro computer faces. $33 suddenly doesn’t seem too bad.
562.
by wenc 3y ago
Schiphol AMS is my favorite European airport. It’s not as great as SIN, ICN or the once dominant HKG though.
563.
by wenc 3y ago
A recent thread on Amazon’s new Chronos forecasting model showed that an ensemble of simple models outperformed it (a highly parametrized transformed model) on the M competition datasets. https://github.com/Nixtla/nixtl
564.
by wenc 3y ago
> Any convex solver can be a non-convex solver if you don't care about global optimality. Aside: Structurally I’m not sure how this would be true. Convex solvers have very specific forms. For instance a QP solver requires a very par
565.
by wenc 3y ago
I don’t know. I’m curious is all. MA57 seems generic but I’ve tried different linear solvers with IPOPT even ones that were supposed to be better, but MA57 still performed the best. Nonlinear numerical problems involve a bit of art. Theory
566.
by wenc 3y ago
Yes tribal knowledge says NLP solvers aren’t tailored to NLE solvers as NLE solvers are. There have been papers in this past showing this but in my experience, it’s not obvious in practice. Nonlinear systems can often surprise us. The init
567.
by wenc 3y ago
I was talking in terms of convex optimization. The criteria for convex optimization is convex objective, convex inequality constraints and convex feasible region, and linear (not just convex) equality constraints. I’m aware that ipopt is a
568.
by wenc 3y ago
> This article is about “solving” differential equations and not convex optimization. This article is about solving nonlinear equations (not differential equations, not sure where you got that from). All NLP optimizers can solve nonlinea
569.
by wenc 3y ago
They should sign up for the next Makridakis forecasting competition. https://en.wikipedia.org/wiki/Makridakis_Competitions Makridakis and Hibon reached the sad conclusion that "statistically sophisticated and comp
570.
by wenc 3y ago
Would be interested to see this compared to IPOPT. https://coin-or.github.io/Ipopt/ IPOPT isn't tailored to NLEs specifically, but it does solve NLPs well. It also uses some amazing Fortran linear algebra routines

More ›