Hmm, it looks like someone took comic sans and tried to make it more like arial. My first thought was that they made a boring version of comic sans.

> It's an annoying behavior of their models and in complete contrast to say Anthropic's Claude which ime will immediately and directly admit to things it had responded incorrectly about when the user mentions it

I don't know whats better here. ChatGPT did have a tendency to reply with things like "Oh, I'm sorry, you are right that x is wrong because of y. Instead of x, you should do x"

To save others a click: the video is a pile of customers packages with addresses ready to send.

“It’s” are the Address lines, which are blurred instead of blacked or whited out, potentially revealing customers private information.

I had plans to try the same, but with Python background and having never touched any lispy languages or any macros I found the Janet for Mortals book surprisingly difficult to follow and gave up.

Funny it doesn’t add comparison to date times in pandas, which is probably used to handle more dates than any of the others.

Pandas uses stdlib or numpy for it seems.

I've been toying with the idea of having arbitrary string types, like

    sql"SELECT FROM ..."

or

    re"\d\d[abc]"

that the development environment could highlight properly, that would ... I don't know. In the end t and f string don't do anything that a t() and f() function couldn't have done, except they are nice. So it would be nice to have more.

IIRC, previous versions of the PEP wanted things to work this way. You might also be interested in the "syntactic macros" PEP (https://peps.python.org/pep-0638/).

It's getting a lot better, but R in production was something companies 10 years ago would say "so we figured out a way".

The problem is pinning dependencies. So while an R analysis written using base R 20 or 30 years ago works fine, something using dplyr is probably really difficult to get up and running.

At my old work we took a copy of CRAN when we started a new project and added dependencies from then.

So instead of asking for dplyr version x.y, as you'd do ... anywhere, we added dplyr as it and its dependencies where stored on CRAN on this specific date.

We also did a lot of systems programming in R, which I thought of as weird, but for the exact same reason as you are saying for Python.

But R is really easy to install, so I don't see why you can't setup a step in your pipeline that does R - or even both R and Python. They can read dataframes from eachothers memory.

renv and rocker have really addressed these issues for using R in production

https://rstudio.github.io/renv/index.html

https://rocker-project.org/images/

I've never played GM Tetris, but have watched the exhibitions on Games Done Quick.

It's my impression that Tetris 99 follows GM more closely than it follows NES.

Tetris 99 is very much official guideline Tetris. Arika, the developer of TGM, (and Tetris 99, as pointed out by an adjacent comment,) has done guideline Tetris before, and even implemented it as an option in TGM games in the past (and presumably in TGM 4 which I haven't played yet) and it is assumed this is because of the Tetris Company enforcing it.

You could certainly never do a T-Spin Triple in ARS rotation.

(Though whether it's closer to GM than NES is an interesting question. It's certainly neither since NES Tetris is quite far from what guideline Tetris would become.)

Tetris 99 was developed by Arika, the developers of the TGM games.

Are you sure parent is the author of that blog post?

Maybe I’m reading the whole thread wrong, but it looks like you are screaming at a maintainer of pg_search that someone else did a poor benchmark

Constantly use search. Using chatgpt exclusively is like those kids that only use tiktok