RE: https://mastodon.social/@moshboy/117506519206590981

"We place our faith in Blast Hardcheese"

RE: https://infosec.exchange/@munin/117540198151714360

"Panic is a threat surface."

One day the smoke will clear long enough for the Resilience Engineering people to discover the fascinating world of risk perception. It's tough to effectively fight fires unless you can differentiate fire from non-fire and tell which fires are most important when it seems like everything is on fire.

The only person I need to amuse is me.

I am going to get kicked out of this ceramics class for sure... #SoftwareArchaeology

What I would like to do: Grant minimum privileges between car and phone and Toyota and their 4.7E+19 partners such that I can view maps and play the static audio files on my phone through the car's speakers AND NOTHING ELSE. No optional services, no trial offers, nothing but music and maps, full stop.

My options seem to be full telemetry to everybody + dog or the surveillance infotainment console is a blue-screened brick with an always-on nag screen that can't be dimmed or turned off.

I mean THIS IS COMPLETELY WHAT I EXPECTED WHEN I BOUGHT THIS CAR. I AM NOT SHOCKED OR SURPRISED, just gravely disappointed that Toyota won't do better.

I'm just so tired of zero-consent coercive phone-oriented computing. In the analog hardware world there would be a breaker I could pull or pins I could jumper to coerce the system into the desired operating mode. This is worse than a brick. Bricks don't actively nag you.

I doubt Toyota is the best or worst in this regard and that there's no modern vehicle that doesn't pull this shit. All roads lead back to Google and Apple so what do you expect?

The missing features aren't going to kill me but I'm stuck with this massive dead-to-me display in the middle of the dash that would be better replaced with a double-DIN bay and third-party non-integrated electronics.

I need a sysadmin for my car and I find this situation terribly depressing.

Venn diagrams overlap in the most unexpected places.

Very very occasionally I find independent and eerily related threads by people I initially followed for very different reasons Today's topic is electric power safety and security, protecting against different kinds of bad at different scales but requiring engineering rigor, understanding, and attention to detail. There are parallel questions about ratcheting vendors into delivering better (safer, more secure) products by standards, regulation, and legislation.

Lindsay works on ebikes and electric vehicles and is one of the most multifariously talented non-degreed engineers I have met. Here are her current projects and woes (CANBUS, battery protection, and cheap/dumb/ubiquitous power connectors that IMO should be standardized off the market): https://fed.brid.gy/r/https://bsky.app/profile/did:plc:7e6rye7jke744sqhw4pqsyec/post/3mxe3es4pak2p

Bert is reporting from a higher level European energy system security forum where similar problems are being surfaced and addressed about grid scale devices and systems - thread starts here: https://eupolicy.social/@bert_hubert/117403890256929658

Some takeaways:
* How do we motivate a security or resilience culture around grid-scale systems?
* Budgets don't make it to the people doing the actual security and protection work
* Legislation might motivate boards especially if there were teeth to the laws
* Regulation lags; it can be effective but needs attention in design and implementation. There are also challenges in regulating the parties who need regulation
* Too many suppliers aren't responsive to security and safety issues. Some are better than others at the engineering work of designing problems out of products
* How do you build an industrial culture that values safety and security and how do you do that in less than 50 years?
* How easy is it to brick an EV? Or a million electrical customers?

Thank you both for today's glimpse inside your respective universes. There are some important parallels at different scales. More often than not I feel lucky to meet the people I do on the socials and you two are shining examples of that. #EveryDayIsTellPeopleYouAppreciateThemDay

Leaving accuracy on the table and just pulling numbers out of your butt is fine for cocktail napkin engineering. If you are writing production code (safety significant or not) don't throw away accuracy. It's just responsible practice; it's not about precision loss, it's about being thorough, paying attention to detail, and being a responsible and competent engineer not ChatGPT or a Magic 8 Ball.

I'm more than a little cheesed at scientists and engineers write code and suddenly forget they need a reference for

every
single
number

in their code. You didn't get your Ph.D playing fast and loose with your technical work, why is it that you completely forget your scientific/engineering fundamentals when you open an IDE?

Maybe software development actually makes you lazy and stupid. It would explain a lot.

It's not a difficult calculation but it's a ton of tedious arithmetic that references a metric buttload of physical property correlations and constants, most of which have no citation or pedigree. None of them are "interesting" enough to discuss in technical reports or academic papers but they are absolutely vital for defending and reproducing the work.

A big part of my code QA effort goes into finding these undocumented and hidden nuggets of joy and trying to reconstruct their pedigree so we have a good reason to believe our code output. It's hella tedious but it's fundamental science & engineering practice and absolutely necessary when dealing with safety significant code.

This is a reason I study software-related incidents and disasters - I'm paranoid of being the main character in a USCSB video, the voiceover part when Sheldon Smith says "... but analysts never verified their safety software was using the correct data." I never want Sheldon Smith to narrate any part of my life https://www.imdb.com/name/nm14185312/

My favorite episode of the The Twilight Zone is "What You Need" which aired on Christmas day, 1959 https://en.wikipedia.org/wiki/What_You_Need_(The_Twilight_Zone)

I had sort of a What You Need moment at work. I wrote a small program and a large amount of documentation because the process for calculating one of the crucial parameter we need in safety is not well documented and it only exists in one of the codes we use, a code I nominally own. I'm the code custodian not the code author. When I get questions about the code I have to dig through the source and sparse documentation to get an answer. A few years ago I predicted that if we got questions about this parameter we'd have a hard time answering them and it would all fall in my lap. I saw a grim future so I took some steps to head it off. The predicted issue arose about a week ago and I'm being dragged in to help solve it.

I'm no psychic but sometimes I see a risk on the horizon and Present Me lays some groundwork to keep Future Me's ass out of the fire. I'm feeling just a little vindicated today - thanks a bunch, Past Me! :)

Have programmers been this awful and clueless forever or did I just miss that section in "The Psychology of Computer Programming"?

https://geraldmweinberg.com/Site/Programming_Psychology.html

How much of this stems from driving women and minorities out of programming?

I mean, intro improv classes are _right there_. If you need a psychologically safe space to work through personal interaction and rejection, it's under $300 for 8 weeks. And it's fun! It's no substitute for therapy but it can be really effective at boosting your social awareness and comfort and self-confidence.

Dunno, I just can't accept putting in the work for allegedly smart people who are told and don't want to hear. Grow up, take an improv class, whatever - just don't expect everyone else to do your work for you.

In case the original thread gets deleted, I want to preserve this because it's really a mind-bogglingly tall ask. I've left off the author and the thread; this isn't about dragging someone being wrong on the internet, it's about poorly socialized devs and how we're being asked to cater to people who don't accept no for an answer.
----

In response to the following:

"Yeah. I'm looking at individual contributors who might well get quite upset over having LLM-generated contributions bounced at the door and who might well sincerely want the ease LLMs provide for their own priorities, but who have little power, and trying to find empathy."
--
Let me get this straight: Someone wants to participate in a community but does not want to work within the community's standards and practices, is turned away as a result, and somehow wants empathy? Because they're the Little Guy and are very sincere and are now very upset because someone in the community pointed at the sign and told them no?

Do software developers just not understand consent and how society works or do they just actively reject it? While I have some empathy for people so poorly socialized they don't understand social interaction and can't handle rejection well (looks in mirror), these are chronological adults who have a lot of growing up to do.

I've made submissions to projects, changes that served both a personal and a community need and had them rejected. I thanked one project for clarifying their goals, I understood and agreed with their decision, and happily forked the project with zero animosity. Another project had lost control of their Windows packaging process leaving the Windows release stagnant and lagging the main unix-oriented release. I offered to fix their packaging process with the explicit goal of minimizing changes to existing architecture and workflow, to avoid disrupting their way of doing things any more than necessary. The project was seeking volunteers and assistance but apparently Not Like That. I had actively communicated with the devs and put in a fair bit of work only to find that they had decided to take a different approach without telling me. In that case I just walked away - I clearly wasn't going to be able to help in any way they recognized. It's aggravating as hell to put in the work and try to stay within the community practices and goals and find out later that their stated goals and actual goals were substantially different and not clearly communicated. I wasn't happy but I didn't need empathy; this crap happens all the time, I wish I hadn't wasted the effort but at the end of the day this was simple miscommunication not malice. Their Project, Their Rules.

Neither of these cases involved AI.

The difference between my experiences and the hypothetical is that I actively tried to understand project goals and to work within their rules, standards, and expectations ahead of time and to cause as little disruption as possible. In the hypothetical, community rules mean little to the contributor. There's a sense of entitlement that the rules and standards and expectations shouldn't apply to them because they are Very Sincere and their changes make their life easier. And they deserve some empathy because now they're very upset because they were told no before they started and still told no after they did all that work.

I'm not going to infantilize developers; they are adult enough to build complex technical systems and they should be adult enough to navigate a simple social system. It's not a project's job to socialize developers and it's mind-boggling that I should be asked to empathize with people who through intent or ignorance do not understand consent and do not accept that rules and standards apply to them. Again, this is an area I struggled with - I've been there. You don't always get your way when you follow the rules so what should you expect when you ignore them?

RE: https://eupolicy.social/@hpod16/117286020053320946

"Teenagers are spending on AVERAGE 4-6 hours online per day, instead of going outside and interacting with humans."

I did that in the early 80s with a book, TV, videogame, or computer. If there's a problem here, it's that parents and institutions do a really poor job of socializing teens. Antidepressants, anti-anxiety meds, a therapist, and improv class would've done miracles at that age but those weren't available to me at that time and place.

Had I been asked in a way that I could articulate an answer about _why_ I didn't want to go outside and interact with other humans, I'd have said primarily anxiety and self-conciousness. Basically there was no place to go that was interesting that didn't involve spending money or dealing with "cops" (those that police a space, not necessarily literal police).

Of the people who registered with me, most were dull or assholes. Some of that was perception but a lot was that I got less enjoyment being around most people than I did buried in a book, game, or programming project.

I doubt my experience is universal but I did experience it.

I'm no sociologist or psychologist but I'm going to guess a lot of the core issues around spaces and people and guidance and behaviors haven't changed much in the last 4 decades. I seriously doubt there was ever a halcyon era for being a teen; parents and institutions and communities do a crap job of socializing teens and always have. It's difficult; it takes time, effort, space, willingness, ability, and resources. Plus there's the keys-locked-in-car problem that you need well-socialized adults to socialize the next generation and nobody has figured out how that first generation of well-socialized adults was socialized. Well-adjusted chicken and the egg.

I'm not going to address the remainder of the original post; I don't have a useful opinion on EU regulation. The "instead of going outside and interacting with humans" part struck a nerve; solitude is often a coping mechanism, a symptom rather than a disorder, and as with many dysfunctions it's systemic and organizational rather being purely isolated to the subject.

Finally, much of that "screen time" is peers interacting with peers. It's just not recognized as social for reasons of ignorance and agenda.

As I recall, @Garwboy has written extensively on this subject (go buy his books or at least read them). He's a trained professional and can speak better about the issue.

Power cycled. Back to base level of dysfunction but at least I have a watch face again.

And remember, you can't power-cycle the watch using its buttons. Long press brings up an AI assistant because of course it does.

Power cycling the watch involves going to the Emegency Call/Dial 911 screen and paging down a ways to get to power icon. Because those are two functions that are closely related: summon paramedics and turn off watch.

Now maybe I can defuse my fury at garbage technology and go back to sleep.

I just broke my Galaxy watch.

It no longer shows a watch face which is literally it's primary function. There is a small circled minus sign at the top of the display. I can select and customize the watch face, I just can't make it display instead of thid baffling black screen with the minus sign.

I have no idea how to search for a resolution - what search terms do you use for "please make my Galaxy watch show a watch face instead of this weird minus sign"?

I'm so angry I feel dead inside. I have never - NEVER - been able to configure this watch in any meaningful manner. I want it to tell time and perform time-related functions (alarm, timer). Once I have it configured, I want it to stay that way until I intentionally and explicitly change it.

This is easily the most unreasonably complex, difficult, malleable, snd unrecoverable piece of compuying hardware I have ever owned. Nothing about it makes any sense and there is no meaningful documentation.

This watch is an abomination, the physical condensate of absolutely everything wrong with computing and user interface design today. It is a hateful thing.

This is not a matter of being old - this device is basically nonfunctional and never has been. The hardware is fine but everything else about it is nightmarishly incomprehensible. I cannot understand who this device is intended for or what the designer's intent for this device to be. It looks like a watch but absolutely refuses to act like one.

And the killer is there is nobody I can ask for help, for an explanation, for any sort of assistance with the thing. I can search the web, I can rummage through Samsung's various web sites and forums, and none of it will be of any relevance. None of it is navigable.

How have we gotten to this point in computing, in product design, that is so alien and nonfunctional as the status quo? How does something like this end up on the market? I cannot understand the deranged malignancy of its user interface designers. You have to work very very hard to make a system as fragile, broken, and unrecoverable as this. One does not make something this bad by accident or incompetence - it takes a concerted effort to create a user interface this frustrating and just plain broken.

I just want my watch to tell time. That should not be a tall ask but here we are.

Anyway, I've been flipping through "String and List Processing in SNOBOL4" and it's been very interesting, especially the example code.

Thanks to whoever pointed me toward the Icon language. That led to the Unicon extension which may give me a better understanding of FP than Lisp/Scheme. I think the big legibility issue is in the lack of intermediate variables. In Lisp people seem to be encouraged to accrete code into a single nested expression rather than defining intermediate quantities. Even though the intermediate variables aren't required at runtime, they document and isolate parts of the code into legible chunks. In this respect even Forth makes more sense to me. I just can't visually digest an s-expression longer than half a line.

RE: https://mastodon.social/@blogdiva/117154164438276531

"oh please, oh please..." https://store.gocomics.com/product/the-far-side-comic-art-print-cat-fud/

When the Coyote is on the cliff's edge, toss him an anvil...