mattbee

And the advice is not (necessarily) "work extra, unpaid hours".

Ah OK - personally that would move it from an "annoying app with adverts" charge (that I'd resist) to a "leasehold bullshit" charge (that I'd pay).

I am stuck similarly with the ClassDojo app that my kids school uses to communicate with parents. The notifications are just "You have a new notification!" which leads to a slow app load, an upsell splash, before finally having to scroll to find the important message from a teacher. In this case though, paying would not make it any less slow to use.

I just check once a week instead, and the parents WhatsApp group fills in the gaps for me.

I've been happy with the solution of switching off notifications from apps that interrupt me with promotions - one strike and they're out.

The remaining notifications are _still_ frequent enough that no single app can expect to get my attention with a single buzz.

It's not like apps don't upsell to when I _open_ them and have to swipe away ads before I can use them. So why give them another channel?

25-years ago me is going to roll his eyes so hard, but you know where I don't mind slightly-targeted ads? My email & my doormat. Send me a catalogue, I love a catalogue.

Go on then, what answer do you get? Was it right?

And users would flee not just because they're seeing the ads but because Firefox is obviously the slowest browser again. Stripping the ads is a big performance boost, so right now Firefox feels snappier than Chrome on ad-laden pages.

I'm not sure I've read any of Marcus' previous writing, but there's no way that essay could have been written by an AI. It's personal and has a structure that follows human thought rather than a prompt.

For sure he describes an education in English that seems misguided and showy. And I get the context - if you don't show off in your English, you'll never aspire to the status of an Englishman. But doggedly sticking to anyone's "rules of good writing" never results in good writing. And I don't think that's what the author is doing, if only because he is writing about the limitations of what he was taught!

So idk maybe he does write like ChatGPT in other contexts? But not on this evidence.

I have seen people use "you're using AI" as a lazy dismissal of someone else's writing, for whatever reasons. That usually tells you more about the person saying it than the writing though.

My initiation into shaders was porting some graphics code from OpenGL on Windows to PS5 and Xbox, and (for your NDA and devkit fees) they give you some very nice debuggers on both platforms.

But yes, when you're stumbling around a black screen, tooling is everything. Porting bits of shader code between syntaxes is the easy bit.

Can you get better tooling on Windows if you stick to DirectX rather than OpenGL?

I agree with you! But the goal for people who want to turn money into power isn't to drive a single narrative, Big Brother style, to the whole world. Not even to a whole country! It's to drive a narrative to the subset of people who can influence political outcomes.

With enough data, a wonky-enough voting system, and poor enforcement of any kind of laws protecting the democratic process - this might be a very very small number of people.

Then the discord really is a problem, because you've ended up with government by a resented minority.

I think you're saying that mass broadcasting is going away? If so, I believe that's true in a technological sense - we don't watch TV or read newspapers as much as before.

And that certainly means niches can flourish, the dream of the 90s.

But I think mass broadcasting is still available, if you can pay for it - troll armies, bots, ads etc. It's just much much harder to recognize and regulate.

(Why that matters to me I guess) Here in the UK with a first past the post electoral system, ideological coherence isn't necessary to turn niche opinion into state power - we're now looking at 25 percent being a winning vote share for a far-right party.

And a really dreary blog describing something simple at great length :/

Yep if you ask Google what Matthew Garrett's reputation is, its AI description includes the (libellous!) accusation of "professional troll". Incredibly, the original articles are still up.

The author's benchmarks suggest 10× would be a pathological case!

But even so - what price correct & secure software? We all lost a tonne of performance overnight when we applied the first Meltdown and Spectre workarounds. This doesn't seem much different.

The Fil-C project ( https://fil-c.org/ ) seems like a more pragmatic way to deal with C security holes in old, well-loved userspace code. It effectively turns C into a managed language rather than a bare metal one, seems to remove a lot of the impetus to rewrite.

First I'd heard of this project; here's an introduction from the author: https://fwsgonzo.medium.com/tinykvm-the-fastest-sandbox-564a...

I reckon they gave away too much, and are clumsily rowing it back.

Gitlab seemed to do a good job of navigating a community edition as an on-ramp for sales. But it's obviously a lot of work to maintain that edition, and VC must be feeling less geenerous than 10-15 years ago.

e.g. maybe if it were my project I'd have kept back the S3-compatible ACL support and put in something super-basic. Or even cluster support. Right now it feels like they're cutting off everything they can while still being able to call it "open source".

They abandoned documentation (edit: for the open source codebase) a couple of weeks ago - that seems more significant.

From their Slack on Oct 10:

"The documentation sites at docs.min.io/community have been pulled of this morning and will redirect to the equivalent AIStor documentation where possible". [emphasis mine]

The minio/docs repository hasn't been updated in 2 weeks now, and the implication is that isn't going to be.

Even when I set up a minio cluster this February, it was both impressively easy and hard in a few small aspects. The most crucial installation tips - around 100Gb networking, Linux kernel tunables and fault-finding - were hung off comments on their github, talking about files that were deleted from the repository years ago.

I've built a cluster for a client that's being expanded to ≈100PB this year. The price of support comes in at at slightly less than the equivalent amount of S3 storage (not including the actual hosting costs!). The value of it just isn't that high to my client - so I guess we're just coasting on what we can get now, and will have to see what real community might form around the source.

I'm not a free software die-hard so I'm grateful for the work minio have put into the world, and the business it's enabling. But it seems super-clear they're stopping those contributions, and I'd bet the final open source release will happen in the next year.

If anyone else is hosting with minio & can't afford the support either :) please drop me a line and maybe we can get something going.

If you have an external monitor and don't mind killing your hinge within an hour it's perfect for Trombone Champ

https://store.steampowered.com/app/1059990/Trombone_Champ/

LLMs can't count letters, their writing is boring, and you can trick them into talking gibberish. That is a long way off the Turing test, even if we were fooled for a couple of weeks in 2022.

IMO when people declare that LLMs "pass" at a particular skill, it's a sign that they don't have the taste or experience to judge that skill themselves. Or - when it's CEOs - they have an interest in devaluing it.

So yes if you're trying to fool an experienced open source maintainer with unrefined LLM-generated code, good luck (especially one who's said he doesn't want that).

I'd extrapolate that the OP's view would be: you've still put in less effort, so your PR is less worthy of his attention than someone who'd done the same without using LLMs.

That's a pretty nice offer from one of the most famous and accomplished free software maintainers in the world. He's promising not to take a short-cut reviewing your PR, in exchange for you not taking a short-cut writing it in the first place.

The observation that inspired this policy is that if you used AI, it is likely you don't know if the code, the documentation or tests are good or appropriate.

Absolutely right!

A week ago I downloaded a couple of movies and shows from Netflix for my 6yo daughter, to watch on a 3hr flight. Worked nicely!

Today we made the return flight. She opens Netflix, and ⅔ of the films have now "expired" with no notice and she can't watch the one she wanted.

For the next flight I'll remember to pirate!

This speaks to me!

I've worked on 3 different projects / workplaces in the last year where I've been the only talker on a Slack with 10s of people on it. Sometimes I'll write 5-6 detailed messages in a day, talking about particular problems or updates on things I'm working on, and almost nothing from anyone else.

These are in subject-specific channels (#network-engineering or #programmers or the like) where colleagues are subscribed.

But the real business of each company seemed to happen in private group chats or video meetings, with no long-term records.

I'm like to state a problem before I've solved it, for the rubber ducking. Very occasionally someone would reply to help, and occasionally I reply with a :facepalm: if I realise I've just been hasty or sloppy. But even if nobody replies, I am very happy to have a public log of my work, the problems I've solved (or not), and the people I might tag for particular input.

If someone DMs me for help with something that is possibly of interest to >1 person, I tend to re-state the issue (without identifying who asked me), then answer in public, and thank them for asking.

If I have a question for a colleague, I will tend to ask it in a channel, as it becomes something searchable.

I ask work questions on Stack Overflow for the same reason, and often self-answer because the place is dead as a community, but the search works well. After a few years I find my own answers as a complete surprise.

But I have colleagues who've not said a line in public basically forever. I've not been a manager for years, always an IC lately. I've not had any objections, but it seems like nobody wants to join me to make "work in public" the default.

Apparently I'm happy to be the exhibitionist little freak?

> No - and in fact in my view this is the core problem with these kinds of situations - there isn't a long established process validating a set of accusations, that if true, fall short of criminality but should result in your exclusion from a community. > Individual communities have to establish ground rules for these sorts of things to protect the vulnerable.

You can never sue anyone for ostracizing you from an open community, or for the consequences of that ostracism. There's no limit on who global communities might choose to ostracize. It's so fundamental to how we group together; you always have to know the norms.

British law is famously friendly to wealthy litigants, and the High Court for awarding ruinous damages. The OP took an opportunity to sue four signatories who (from my understanding of the court order) put their name to harmful allegations that they didn't know the truth of. The four defendants paid £20,000 in costs and damages.

Unfortunately for the OP, the ostracism clearly still stands, and despite going to the High Court to sue for libel, the first-hand reports of his conduct are still online.

I don't see this as a lesson in the terrifyingly and unpredictable consequences of Cancellation - seems like more "don't shit where you eat".

Yep in the same way it would allow `var u userID = 1` it allows `Work(1)` rather than insisting on `var u userID = userID(1)` and `Work(userID(1))`.

I teach Go a few times a year, and this comes up a few times a year. I've not got a good answer why this is consistent with such an otherwise-explicit language.

You can definitely test 999 for your own VoIP without getting into trouble. I did it a couple of times in the 2010s after moving offices. You just say you're testing, there's no emergency, thank you, and hang up.

10 years ago I'd have settled for this if it only worked on Game Of Thrones characters.

I paid for Memrise to polish up French. The scripted lessons alwere great but it dropped me into an AI conversation assistant that did exactly the same. It forgot the vocab and grammar level that the scripted lessons had taught, and often broke into idiom. I haven't picked it up since.

And you can get it on Windows with just "winget install zyedidia.micro". Reminds me of 8 & 16-bit editors of a similar era.

The GP was very precise in the experience they shared, and I thought it was interesting.

They were obviously not trying to make a sweeping comment about the entire future of the field.

Are you using ChatGPT to write your loquacious replies?

I've still got one! I can't believe anyone loved them though, they miss key presses if you type too fast.