Man, it would be nice if `R CMD check` would display the filenames of the files involved when showing warnings. You'd think that would be both trivial, and trivially obvious.

#rstats

uncaught exception: Object { stack, message }

Thanks, that's super helpful #javascript debug console.

Holy shit, there's a `dplyr::filter_out`!

This sure could have saved me some awkward conditionals.

#rstats

I do not want the Wellbutrin. Please give me your TopShelfButrin.

Okay, so all my posts lately are starting with "JFC", but:

JFC, just when I thought #Tableau couldn't get any worse:

#AI #fuck_ai #LLM #H8LLMs

Man, a reporter on #NPR this morning (I wish I could remember her name), referred to, "The Iranian government's harsh crackdown on protesters this January, and the war with the United States that followed," which seems to me a _really_ irresponsible way of talking about this. It makes it seem like our operation in Iran had _anything_ to do with Iran's treatment of its citizens, which it obviously didn't.

Y'all should not be okaywashing this shit.

RE: https://mastodon.social/@cmconseils/116645958782664252

OMG. This couldn't be truer.

Even after, what, ten years, maybe, now? #Rust compiler messages continue to surprise me.

"Just FYI, I'll allow this, but you're going to be confused when you look at this two weeks from now."

#Rustlang

* #RaspberryPi 5 w/1 G RAM: 45 USD
* RaspberryPi 5 w/16 G RAM: 305 USD

#ThanksAI

I desperately need #rstats help from someone with package-building experience:

https://codeberg.org/CORE-NC/cncdata/issues/2

I have maintained a couple of internal packages at work for a while, so it's not like I have _none_, but I just don't understand what's wrong here. I have banged my head against StackOverflow and found a few similar questions with some dubious-sounding solutions; I've tried them all.

I humbly shower everyone reading this with gratitude in advance.

@dannotdaniel

BTW. If protestors getting killed in the streets by "security forces" and having nuclear capability controlled by a theocratic infused mad man is grounds for attacking a country.... just saying.

I once again find myself consumed with rage about the #MicrosoftExcel file format.

OMG, #RStats, why is it `read.csv()` and `write.csv()` but `readRDS()` and `saveRDS()` ??!?

WHY, #TABLEAU, WHY?!?!

My data is not large!

Y U TAKE FOREVER WITH EACH CLICK?!?

Even bringing up the RIGHT CLICK MENU takes several seconds. Why? WHY?

Y?Y?Y?Y?Y?Y?Y?

👎

OMG, it finally works.

Writing web servers in R is a huge pain in the ass, in case anyone's wondering. Thank the Old Gods and the New for (rextendr).

#rstats #rustlang

I helped a woman install #Linux for the first time today at the coffee shop.

The #Gleam language tour's usefulness is evergreen.

#Erlang strings are linked lists of (bytes? code points?), and Gleam spit this one out in the terminal accordingly, which is not particularly helpful. Copy-paste into the tour's textarea to the rescue!

#gleamlang

Monopoly wasn't invented by the Parker Brothers, nor the man they gave it credit for. In 1904, Monopoly was originally called The Landlord's Game, and was invented by a radical woman. Elizabeth Magie's original game had not one, but two sets of rules to choose from.
One was called "Prosperity", where every player won money anytime another gained a property. And the game was won by everyone playing only when the person with the least doubled their resources. A game of collaboration and social good.
The second set of rules was called "Monopoly", where players succeeded by taking properties and rent from those with less luck rolling the dice. The winner was the person who used their power to eliminate everyone else.
Magie's mission was to teach us how different we feel when playing Prosperity vs Monopoly, hoping that it would one day change national policies.
When the Parker Bros adopted the game, they erased the "Prosperity" rules and celebrated "Monopoly".

Okay, lazy #webdev ,

Say I am responding to an HTTP request (from the frontend of my app) with some basic biographical information:

last_name (string)
first_name (string)
bio_text (big long string)

The obvious thing here is to have the response be a JSON structure, like

{
"last_name": "Santos-Dumont",
"first_name": "Alberto",
bio_text: "Most famous for his pioneering aeronautical work..."
}

Now, if the last_name and first_name fields are _guaranteed_ to be ASCII (because in my use case they are not actually names, but are something else), is it "bad" (and if so, _how_ bad), to use a couple of custom headers, and have the body just be plain text:

HTTP/1.1 200 OK
...irrelevant headers elided...
content-type: text/plain
x-last-name: Santos-Dumont
x-first-name: Alberto

Most famous for his pioneering aeronautical work...

??

This isn't a general-purpose HTTP API endpoint; it's going to be responding only via HTTPS to my front-end user-agent that has already authenticated. This seems okay to me, even if it is blurring "data" and "metadata" a little bit.

OMG.

Writing #javascript for the browser is truly the most wretched programming experience.