Very, very cool. Besides the top-performing models, it's interesting (if I'm reading this correctly) that gpt-5.2 did ~2x better than gpt-5.2-codex.. why?
> gpt-5.2 did ~2x better than gpt-5.2-codex.. why?
Optimising a model for a certain task, via fine-tuning (aka post-training), can lead to loss of performance on other tasks. People want codex to "generate code" and "drive agents" and so on. So oAI fine-tuned for that.
A tragedy, yes. I can't be the only old fart around here with fond memories of John Siracusa's macOS ("OS X") reviews & Jon "Hannibal" Stokes' deep dives in CPU microarchitectures...
Dr Dobbs was pretty good until almost the end, no? If memory serves me well, I recall the magazine got thinner and more sparse towards the end, but still high signal-to-noise ratio. Quite the opposite of Ars T.
Huge debt of gratitude to DDJ. I remember taking the bus to the capital every month just to buy the magazine on the newsstand.
I would go to the library on my bicycle to scour for a new copy of DDJ as a 10 year old.
I had dreams of someday meeting “Dr.
Dobbs.” Of course, that was back in the day when Microsoft mailed me a free Windows SDK with printed manuals when I sent them a letter asking them how to write Windows programs, complete with a note from somebody important (maybe Ballmer) wishing me luck programming for Windows. Wish I’d kept it.
While we are here, may I ask what are some blogs you guys read regularly? (Regularly as in: going back to read new articles as opposed to a one-off link shared on some other platform.)
Neat. Personally, I wasn't aware of mathjax, it's a bit of a revelation to learn you can do this. I like very much the fact that the ASCII side is highly readable - compared to say, LaTeX (to me, anyway!) - as something I could use in code comments.
I've wanted a music player like the early versions of iTunes for a while, and this looks like it might fit the bill.
Those who've only known Music.app and later iTunes versions might be surprised to learn that there was a time when iTunes actually had a clean, intuitive UI: https://www.versionmuseum.com/history-of/itunes-app
Swinsian was the only Mac music player I could find that could come close to replicating my old MusicBee setup. The license fee was annoying but I paid it anyway and have no regrets.
and plus one here!
I don't know, I like my mac workflow but irritation and aggravation have crept in more frequently of late. Last week I was told a binary that clang++ had just produced from my own code could not be run because Apple couldn't check whether it was safe..
And what to make of power users complaining bitterly about Tahoe & liquid glass etc?
I'm hanging on to Ventura for now.
I thought it's interesting that GPT5's comments (on prompting it for feedback on the article) seem to overlap with some of the points you guys made:
My [GPT5's -poster's note] take / Reflections
I find the article a useful provocation:
it asks us to reflect on what we value in being programmers.
It’s not anti-AI per se, but it is anti-losing-the-core craft.
For someone in your position (in *redacted* / Europe)
it raises questions about what kind of programming work you want:
deep, challenging, craft-oriented, or more tool/AI mediated.
It might also suggest you think about building skills
that are robust to automation: e.g., architecture,
critical thinking, complex problem solving, domain knowledge.
The identity crisis is less about “will we have programmers” and
more “what shapes will programming roles take”.
