@creachadair@mastodon.social avatar

creachadair

@creachadair@mastodon.social

A lover of language, a writer of words, a spinner of yarns, and a mangler of bits. A page, torn from a book.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

danderson, to random
@danderson@hachyderm.io avatar

Okay I've now written enough Elixir+Phoenix to ship something modestly spiffy, and I'm reaching the point of framework fatigue, I think. In particular, the breaking point in Phoenix is the second you have to make a form that doesn't match the shape of your data model, things suddenly get pretty unpleasant.

Not unmanageable, just... enough to make me pine for Go again, and less frame/more work.

creachadair,
@creachadair@mastodon.social avatar

@danderson I have been noodling around with https://github.com/a-h/templ a bit, as a potential answer to what seems like an inevitable Gordian knot at the heart of every so-called declarative framework.

So far it seems like a fair compromise between letting me program in a real language rather than some weird library API, while integrating decently with UI helper libraries.

I haven't done anything super crazy with it, but it hasn't made me hate my life yet.

creachadair,
@creachadair@mastodon.social avatar

@danderson stipulated, this is a much lower level than what you've been describing, but I find the Total Institution style always comes back to bite me in the ass, so for me at least that's a feature.

creachadair,
@creachadair@mastodon.social avatar

@danderson yeah. That's what drew me to it as well—the stdlib template engine is ok for very basic stuff, but pretty quickly gets overwhelmed and kind of awful to use. I set about trying to write more or less this, only to find it already written (and better than I'd have managed).

creachadair,
@creachadair@mastodon.social avatar

@danderson Yeah. My approach was going to be basically building a baby Lisp to compile down into Go, it's certainly not a new idea. I feel like the template forms fit pretty nicely into the existing code.

fugueish, to random

Do you suffer anxiety because the world is all messed up? I suggest soothing yourself by writing some nice, clear, well-specified, well-tested computer programs. Mmmm. Yes. Words mean specific things, and the program performs its task. Feels good. Deep in your bones.

Now, as an unrelated aside, one tiny reason among many that the world is messed up is that people keep writing computer programs. Try not to think about this

creachadair,
@creachadair@mastodon.social avatar

@fugueish one might reasonably argue that the true moment things went wrong was much earlier, when we learned to enjoy writing computer programs in the first place.

creachadair,
@creachadair@mastodon.social avatar

@annaleen @fugueish but this prison has such good stories…

creachadair, to random
@creachadair@mastodon.social avatar

It seems like the GitHub analogue of getting ratioed is when your repository has been forked more times than it's been starred.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

Today I learned that if you host your lab web page on the cloud via your own domain name, then when you are hit with vexatious legal threats your university will refuse to provide legal assistance on the grounds it is a "personal professional page" rather a work created in your university capacity.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom oh FFS.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom and no doubt if you'd tried to get the university to provision it, they'd have stuck you on a disused 286 in a forgotten closet sporting a 10 year old build of Wordpress hosted on Windows NT and a 500M spinning disk mostly full of OEM crapware you can't uninstall because they won't let you be administrator.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

I deleted a dark post about the new cybertruck logo. Why? No one needs more bad feelings, and I think we all could use more owls. Much better to leave them at the top of my feed.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom One of the things I like about the fediverse is that there's plenty of time between posting something and anyone noticing it for such reflections to take hold 😂

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar
  1. In five days, I'm headed to Australia: Sydney, Brisbane, and Melbourne. I can't wait. So many colleagues to meet and birds to see!

For those who are interested in such things, I'll be giving a few public talks in addition to a number of academic lectures.

The public talks will mostly focus on something I've been thinking about lately, "The Crisis of Human Collective Decision-Making in a Social Media World".

Details below.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom Do you expect any of the versions of your Crisis talk will be available as recordings? (Or if not, might you give it again sometime at the U so us locals can wander down and listen?)

danderson, to random
@danderson@hachyderm.io avatar

Welp, the timegrapher is here, and the news isn't that great for my old soviet movement. It's quite sick, and visibly unhappy compared to last time I inspected it :/

Timegrapher reports 220deg amplitude (terrible, should be 280-330), 4.5ms beat error. But worse, if I rotate the movement to the vertical, the amplitude drops to barely over 100 degrees, and the time graph turns into snow. It's visibly barely able to run in that orientation.

creachadair,
@creachadair@mastodon.social avatar

@danderson I would have made a joke about Soviet movements running down over time but as a U.S. resident these days I'm pretty sure that would be an own goal.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

So when I was in graduate school in 1993 and you wanted to compile a latex manuscript on an IBM 386 processer, you hit the button and went to get a cup of coffee. Five or ten minutes later, it would be done.

These days it happens in the blink of an eye.

So why is it that in 2023 basically every manuscript submission system in existance takes at least 10 minutes to latex a file?

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom Without a profile I'd have to guess—likely they're running it in some low-cost containerized cloud environment on a feeble multi-tenant VM with nothing cached and all the data in slow storage like EBS or S3.

The cloud faithfully recreates the ceremony and performance of mainframe era batch processing: it's slow, but at least it's expensive!

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

PNAS doing the Lord's work.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom "Also, do not even ask for extra time beyond the submission deadline. We do not, despite what you may have been told, grant PNAS extensions."

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

At a fascinating Santa Fe Institute meeting on Accelerating Science. Chiara Franzoni gave a fascinating talk about what it takes to fund high risk science, and it left me thinking.

When we evaluate grant proposals for an NSF or NIH panel, several panel members evaluate each proposal — and this sets up the classic sort of preference aggregation problem that is the bread and butter of public choice theory.

Yet we typically ignore the preference aggregation issue.

creachadair,
@creachadair@mastodon.social avatar

@chiasm @ct_bergstrom Even when we don't do the comparisons explicitly, a score on a shared scale is a proxy for the pairwise evaluation; assuming the scores are a good proxy (which is arguable, but it's an assumption we often make) ranking by score should produce the same partial orders.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom I feel like there's a "STOP DOING AVERAGES" meme in here just waiting to get out.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

Google Scholar is a flaming piece of shit. Why do we use it for scholarly evaluation?

Here, it misattributes 4873 citations — a decent count for entire research career — from Keeling and Rohani's landmark book, giving them instead to the authors who wrote a book review of the book.

Yet, there's no correction or appeal process.

Book review to which the citations are ascribed.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom A key thing to remember about Google is that, while they're really good (or were once) at solving problems you can handle by throwing machines at it, any problem that requires human intervention they are not even going to try to solve. Hence their AI bullshit.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

Must suck to realize that there are some forms of epistemic grounding that you and your pals can't simply purchase and torch to the ground.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom Which is especially poignant, given that Wikipedia is one of the very few things left preventing search engines from complete uselessness.

creachadair, to random
@creachadair@mastodon.social avatar

Even after all these years I'm still bitter that Python3 made such small changes and yet managed to break so much working code without realizing any meaningful improvement to performance, usability, or reliability. It's still slow, still wasteful of memory, still bad at concurrency, still has terrible string representation, still requires tests to detect typographical errors. It's all the costs of a second system with none of the benefits.

danderson, to random
@danderson@hachyderm.io avatar

While mucking about in a DOM inspector, turns out youtube calls one of the various places that shows ads on the site an "engagement panel" 🤮

creachadair,
@creachadair@mastodon.social avatar

@danderson I suppose that is an appropriate name when viewed in the context of engaging the teeth of a very large metal gear.

danderson, to random
@danderson@hachyderm.io avatar

I was just looking at SponsorBlock's homepage, and it says there are 13 million users who are actively crowdsourcing the data that powers it.

That feels like a good litmus test for whether you're doing a good thing in tech: are thirteen million people donating their free time to make your stuff go away because it's that hostile?

creachadair,
@creachadair@mastodon.social avatar

@phire @fugueish @danderson Grammarly is the world's most perfect troll, it gets people who can't be bothered to learn grammar to pay money to have their incorrect grammar made differently incorrect in a way they can't spot for the same reason they bought it in the first place.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

If I were to set up a wordpress blog to write about bullshit, science, big tech, large language models, and all that, what would you think I should title it?

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom "All the News that's Shit to Print"

Though given the way the NYT has been lately, that might be a little too on the nose.

danderson, to random
@danderson@hachyderm.io avatar

Watching the Foundation series made me wonder about a "reverse 4X" kind of game, a strategy game where you start with a galaxy-spanning empire and have to manage the fall. Feels like it could be an interesting twist. Surely this has been explored already... But I don't know what games might have done this.

creachadair,
@creachadair@mastodon.social avatar

@danderson I would play the hell out of that.

ct_bergstrom, to random
@ct_bergstrom@fediscience.org avatar

I wanted to consolidate a few thoughts on google, misinformation, large language models, enshittification, and the fate of the web as we know it.

It started when Carl Zimmer shared this remarkable example of Google being fooled by machine-generated bullshit online.

creachadair,
@creachadair@mastodon.social avatar

@ct_bergstrom It is tempting to imagine that because Google employs a lot of skilled and rational people, their product decisions will be smart and sensible. But Google's leadership is uninspired and clannish, shot through with FOMO, groupthink, obsession with fads, impracticality, and arrogance. Google of today is no longer capable of the kinds of technical accomplishments that wowed the world in 1998. It's a prisoner of its devil's bargain with advertising, chasing taillights.

danderson, to random
@danderson@hachyderm.io avatar

Oh look, the thing holding up this giant rebuild on a 32-core machine is, once again, some rust library that somehow cannot parallelize compilation at all, and because rust takes 15min to get anything done.

I could tell when it started and stopped holding up the build, because my system's fan speed audibly changed as it just killed all concurrency.

creachadair,
@creachadair@mastodon.social avatar

@danderson That said, Rust is paying the predictable costs of a type system that requires whole-program analysis. With C++ you could (sorta) compile separately (since there is no encapsulation) and then only the linker is terrible. Rust elevated that to the front end. (Though at least it gives you back something for that cost)

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • megavids
  • thenastyranch
  • rosin
  • GTA5RPClips
  • osvaldo12
  • love
  • Youngstown
  • slotface
  • khanakhh
  • everett
  • kavyap
  • mdbf
  • DreamBathrooms
  • ngwrru68w68
  • provamag3
  • magazineikmin
  • InstantRegret
  • normalnudes
  • tacticalgear
  • cubers
  • ethstaker
  • modclub
  • cisconetworking
  • Durango
  • anitta
  • Leos
  • tester
  • JUstTest
  • All magazines
    •