Put together some notes on the Gemini terms-of-service: it looks like their paid API tier doesn't train on your inputs to the model, but the free API tier does:
https://simonwillison.net/2024/Oct/17/gemini-terms-of-service/
Data used to create fine-tuned models, even on the free tier (they have a free tier!) won't be used for training. I hadn't realized you could fine-tune a model there for free, that's pretty wild: https://ai.google.dev/gemini-api/docs/model-tuning
On this day in 1893, silent superstar Lillian Gish was born in Springfield, Ohio. In 1988, I wrote to her asking if silent films had actual dialogue written out in the scripts or if there was just a general description of what was being said. This was her remarkable reply.
@StoliarSteve #Silentmovies #LillianGish #letters
the end of the .io domain: https://every.to/p/the-disappearance-of-an-internet-domain
unexpected consequences from territory changes…
This is the UK Government "Digital and Data Profession Capability Framework" for Software Developer.
It is an attempt to define expectations for staff in different IT jobs and compares the same role at Junior, Senior, Lead levels.
https://ddat-capability-framework.service.gov.uk/role/software-developer
They have Data Engineer, Data Scientist and others, too.
Following up on #RSECon24. #SoftwareEngineering #DataScience #dataengineering
48 years ago, on Sept. 17, 1976, NASA rolled out its first space shuttle, named Enterprise, from its manufacturing plant in Palmdale, CA.
The shuttle was planned to be named the Constitution, but thanks to a massive write-in campaign by Star Trek fans, President Ford relented and advised NASA to change the name to Enterprise.
Star Trek creator Roddenberry and many of the show’s cast members attended the Shuttle rollout ceremony.
https://www.nasa.gov/history/45-years-ago-space-shuttle-enterprise-makes-its-public-debut/
1/n
This is the sort of madness I absolutely love to see in the world. Marvellous.
Man builds £30,000 canal in his garden to house his barge
https://metro.co.uk/2024/08/20/man-builds-30-000-canal-garden-house-barge-21458913/
"Whenever I, or someone else, posts a link to this blog on Mastodon, it DDoS's me and brings the site down for a couple minutes."
The frontend community is in crisis. I know, because I could spend every waking hour helping e-commerce and productivity apps fix the *unbelievably* bad performance that is now the hallmark of contemporary, JavaScript-first web development.
But it's worse than that. This stuff has infected public services; the sorts of sites that have to serve *everyone*, iPhone or no.
Part 2 of this series is the hardest to watch, but essential to understand how far we've fallen:
@slightlyoff I’ve wanted to get back to you for a while because you were a big reason for our e-commerce website being the way it was, technically speaking.
We took your earlier publications to heart, and we applied every kind of optimization we could think of, just to try and stay a little bit ahead of the competition. I think we did all right in that respect.
When the company was shut down at the end of 2018, I finally put together a document outlining our techniques. It’s not a tutorial and lots of things aren’t perfect, but I think a lot of this can still be used.
I was weirdly proud of the server-side cached pre-renders for several cohorts (at FRONTEND > web server > httpd in the link below), which we were continuously building for the 5 most viewed pages. These on disk static pages didn’t have a long TTL but this allowed us to reevaluate the need for a pricy CDN.
These Apache2 rules allow for high throughput of the pages that are cached in this way. Because of this, we have been able to withstand traffic generated through TV mentions and other publicity without the help of a CDN or any autoscaling.
Of course we eventually placed nginx in front of httpd and these rules became somewhat redundant, but not if you include brotli and other CPU-heavy compression.
Anyway I’m geeking out on a dead website. Here’s the doc:
@splorp and froze some of it for good measure...
The whole AI thing has me endlessly confused. Half the market is crashing because investors didn't see any signs of payoff in the quarterly earnings report, but I'm so lost as to what exactly they were expecting to see. Did they just not pay any attention at all to what these companies were actually doing with AI?
Were they expecting exponential Instagram usage growth as a result of Meta making it so you can have a conversation with the search bar? Or maybe everyone was going to buy 10 new Windows licenses in celebration of Microsoft announcing they want to install AI powered spyware on everyone's computer? Or was Google going to sell more ads by replacing all the search results with reddit shitposts. Either I'm missing something or everyone's 2 remaining brain cells are just really busy fighting to death for 3rd place.
Mac people: if you miss the older style menus bars, when they still looked cool, give Lickable Menu Bar a try. mtm set up with the shiny Tiger menu bar again, and really like it.
https://apps.apple.com/ca/app/lickable-menu-bar/id6444217677?mt=12
Facts, not wishful thinking.