paul_h

https://www.satherley.com/ seems to be empty of content :(

Nobody cares perhaps, but F5 as a refresh key came in the mid 80s (I was told it was a $1m of usability consulting). The midrange AS/400 got [CUA](https://www.ibm.com/docs/en/zos/3.1.0?topic=actions-default-...) first. There were many more F1 for help, F12 for go-back-a-page, F3 for big-exit of app. Green screen stuff (3270 and 5250 where two common wire/frame protocols). The same keys made the jump to some windows apps. One by one they got whittled away until F5 for refresh was left. For that one it wasn't just Windows - as browser rollout took it wider. Has it gone yet? I'm on a Chromebook and it doesn't have F keys so I can't see for myself.

ThoughtWorks' in joke about pressuring leadership to get us off Notes and into the Google mail/docs regardless of the loss of easy apps: https://noteshater.blogspot.com

Update the original article :)

That's hard to read. The writing assums familarity too much. Decades, years, manufacturer names, references.

I think all of Anthropic's own staff use --dangerously else they'd work on the usability of the massively imperfect 1-approve-once and 2-approve-forever multi-choice and outcome.

Chris Hinsley started TaOS toward the end of the 80's .. Him sick of being asked by games industry managers to "do it again but for PC". By 1991 there was a start up, and by 1993 there were awards for his endian-independent VpAsm OS. I mention cos the tubling cube demo (two windows) would allow one to be dragged over the other without cubes slowing in their tumble at all, and without the envelope-drag-letGo-repaint effect. No stammering for the window manager. This tech was legendary and the multi-tasking was an order better than WinNT later which itself was an order better than Mac's cooperative-task-thunking. Lots of the videos of that time lost sadly, and all the source code. Chris is doing it all again, but making it harder for himself -> https://github.com/vygr/ChrysaLisp - while also using LLMs for lots of it.

A fun hindsight LLM project :)

Way back lots of https://en.wikipedia.org/wiki/Operation_Ore couln't bring cases against all because the expert witnesses (US law enforcement website trap-setters) wouldn't go to the UK to testify. How come Canada was able to press ahead without the expert winesses being cross examinable?

I recall a couple of JetBrains staff visiting in 2008 to see IntelliJ's struggle with the size of checkout for AdWords. Soon after the reindexing front/center dialog moved from blocking and center of window, to a status line message that was non-bloking of edits at the bottom of the window. It may have been due anyway but was shown as problematic in the same moment.

TL;DR:

  There once was a ledger at Uber
  On DynamoDB it ran like a dream,
  Fast writes at great scale,
  But costs set sail,
  And consistency wasn't what it seemed.

I could not work "find someone to fire" into it, but I did not try that hard to be honest. I would not like to work with someone who'd leave a blameless postmortem with "those as asshats wouldn't apportion blame" tale to tell their smug crew later.

Semantic merge. PlasticSCM had that a feature many years back

> Bzzt. Since all of these people are correct and smart, it is now your job to have great answers to their objections.

Power assymetry and tenure are a factor. So is "culture eats strategy for breakfast" realities in organizations.

e.g. Change of pandemic messaging in early 2020 from "covid transmits by droplets and fomites and 2 meters and handwashing keep you safe" to "covid is airborne and fills a room like smoke" (and multiple mitigations around that) was attempted in groups by multiple expert scientists and ultimately took years when it needed to be done fast. One key moment: https://www.google.com/search?q=%22where+is+your+evidence+li... You can come with evidence for those tough questions to support your position, but one grandee scoffs at your suggestion in the meeting you networked hard to get and there's no recovery.

Awesome that LLM generated and still an engaging account. Automated testing (as a software improvement technique) is an AI blind spot. That tweak of spec is the iterative cycle, with no mention of additional automated tests is telling.

no_unused_seconds_meme.jpg

I was looking at TrueNAS CORE to see if it was a viable way to bsd-jail Linux containers. I'm really only doing this to get some protection from supply chain attacks given I'm fairly promiscuous at git-clone-and-run-a-build. Before that I was aiming for the same with Bastille and had got to the give up stage because it felt too fiddly to set up. This was a year ago. Maybe its better now

Note to self: look for second hand unlocked Pixel 10 pro!

> Scotland has a lot of planted pine forests that have drowned out native species

Team Land management for Grouse hunting enters the chat.

https://www.mossy.earth/rewilding-knowledge/rewilding-scotla... ... "In the 1700s, large scale sport shooting and sheep grazing began to leave its mark on the landscape. Overabundant herbivores and over grazing, alongside regular burning, prevented woodlands from naturally regenerating, causing soil erosion, soil acidification, flooding, biodiversity loss and more"

And - https://www.youtube.com/watch?v=MuTV1SiYeu4 - is a longer story that bubbled up in social media a few months ago. Maybe not Hacker News.

For a while, I've been thinking that open source package portals will at some point take over making of binaries that get released. Dev teams will run their own CI with whatever automated test pipelines they think is appropriate. For a tests-pass situation and will pass the git hash to the portal system for release, which just runs compile and making the binary. Well, not all CI runs would result in a release, of course. Then the package portal's own software kicks in to calculate an independent since-last-release report that's attached alongside the maintainer release notes.

All such portals upgrade their hash/sig noting of binaries, and keep those in a history retaining merkle tree of sorts. Of nothing, else a git repo. Something like this https://github.com/hboutemy/mcmm-yaml/blob/master/aws/sdk/ko... but with SHA256s, and maybe not the entire world on one repo.

Exciting .. I'm typing on that HP Dragonfly now :) Google - put me in the testing group pls - Paul H

Agree. But the first build you do after that clone/checkout is risky too. Maybe not as wide open, as the build-tool makers are a line of defence if they're acting on classes of vuln.

I have not met a ThinkPad I didn't want to chisel the PgUp and PgDn keys off

How bold to start with "Listen to me" then jump into something that doesn't make much economic sense and has not been properly considered

He he, I might now be retiring my Ubuntu25 passwordless-sudoer NUC that's only for yolo mode projects. Or giving it more duties. Also - hello from Edinburgh!

Brit here. UK Government's position "we will hold ourselves to equivalent standards via the Cyber Action Plan, just without legal obligations" -is institutionally equivalent to "trust the PDF." Fast forward to the non-repudiable era, please.

isolated-vm (https://www.npmjs.com/package/isolated-vm) here we come for increased sandboxing of node bits and pieces? And we are a year after Java took out the security manager that could sandbox jars in separate classloaders - a standout feature since 1995.

Diverged as well as standardized. I did some research into "out of pocket" and how it differs in meaning in UK-English (paying from one's own funds) and American-English (uncontactable) and I recall 1908 being the current thought as to when the divergence happened: 1908 short story by O. Henry titled "Buried Treasure."

My sister in law has a stem cell transplant under this team https://sheffield.ac.uk/neuroscience-institute/research/brea... for Multiple Sclerosis. It was a small multi-country trial of about 100 people, I recall. This was nine years ago and she counts herself as cured, though to has a follow up once a year with the same team. The treatment is expensive, and though this was all UK National-Health-Service linked, the NHS is seeking cheaper MS treatments now: https://www.mssociety.org.uk/research/latest-research/latest... (Metformin, ALA, ?) There was a trial before this trial - some 20 people in Canada (?) and a couple died. In the 100 person trial they changed the selection criteria for patients, and I'd like to think nobody died cos of the treatment, but I've not heard anything to be able to add a figure. Its Mexico only now perhaps, people did self-fund to go to Moscow for the same treatment: https://www.bbc.co.uk/news/uk-scotland-54560236. These two ladies are not updating how it went for them, years later, but one is on Instagram living a normal life, and the other Facebook

Agree. I'm having lots of fun with Gen-AI and I still have insights it does not. Test-at-the-same-time-as-prod-code is also doable with gen-AI. All the ones I tried are shit at testability by default in my experience. And every now and again they forget about tests being important.

Ask AI number 2 to summarize the intention (look at the .patch) to markdown. Reset. Ask you AI to read the intention as of the orig author had written it, and say you've grave doubts about the contrib's functionally and non-functionally and for it to help you put that into words to feel back to the contributor. Basically the playbook from https://paulhammant.com/images/SimonSinghsFermatExcerpt.jpg