Mark Tomczak @mtomczak@qoto.org

@stargirl I agree. The easier solution is to teach people TypeScript and npm, rewrite those libraries people care about, and leave Python in the history pile next to Perl.

@mcc @mhoemmen It's frustrating because it simultaneously got better and worse when React moved to functional components... On the one hand, `useContext` is a lot easier to use than the Provider component or the hooks they added to class components to declare they use contexts. On the other hand, it broke the symmetry in the original API that now Providers are part of the render tree but Consumers are an algebraic effect.

@mcc @mhoemmen Yes, the thing you're imagining is a "context." A bit of a pain-in-the-ass in most OOPs because it requires every single function in the universe to have a new argument on the front, but they're definitely a concept in other languages (or found in other frameworks... Game development, for example, is quite fond of them since it's common sense that 90% of the code is talking about the same "present state of the game world" so basically a global variable is always available to get that present state).

React offers a neat twist on these in the form of their Contexts. The render tree can include Providers for various Contexts, and anything under that point in the tree that declares itself a Consumer of the Context gets a reference to the Context the Provider maintains. Different parts of the tree could be under different Providers, but you don't see any of this unless one of your React Components explicitly declares it's a Consumer.

I've grown quite fond of this pattern (and its neighbor in React, Algebraic Effects) for taming some of the complexity (although it can add its own, other form of complexity: looking statically at some code that uses a Consumer, you can't know which Producer fed it without tracing the render tree, so it invites spaghetti code if overused).

@tess @mcc Older I get, more convinced I become that the "god object" anti-pattern is indicative of languages where the object-creation / management abstraction is too heavyweight.

To make one new Java class for general consumption, I don't just... Make a class. I create a file, and put that file in the right place, and declare its package membership, and create the class, and declare its members, and-and-and-and... IDEs help but it's still actually a lot of lines of code.

Whereas one new function in an existing class is way, way cheaper, and everything that already knows about that class now has access to it.

God objects are indicative, I think, that what the user wants is easy access to a *context,* either static or runtime, for which a class family and object instantiation are a poor substitute.

@lauren ... which reminds me. Purely as an experiment / finger-in-the-wind, I need to figure out how to carve out some time to make a Chrome Extension to put a yellow box behind every ad result.

@lauren So I'm mentoring an #frc team this year out of a trade-magnet high school.

During a meeting at the school, needed to get on Discord voice chat with a team member outside the building because cell reception is too crap for a phone call. And of course, the school IT had locked down Discord's voice ports.

Put my head in my hands and *may* have muttered something moderately un-gracious-or-professional about sometimes feeling like doing work on school networks is like solving problems post-lobotomy.

One of the students saw me and came over with his phone patched through the voice channel on the Discord server. I asked him how he did it. "Oh, I spend some of my after-school job money on a VPN because the school locks down so many ports I need."

😕

@SamantazFox Oh, logged in accounts carry all sorts of information that advertisers can micro-target. If I have a different search and ads history, I won't see the ads they see.

I wouldn't be surprised if some malicious actors are using that micro-targeting toolkit for "stochastic spear-phishing," if you will. Blech.

@ocdtrekkie exclusive possession or control. Unless you're using the European definition, which is quite different from the American definition.

@ocdtrekkie > claim Google isn't a monopoly.

Uh huh. Get back to me when DuckDuckGo stops resolving on my Android phone. Or my Chromebook.

@ocdtrekkie I think this conversation has run its course. At this point, you're throwing around terms like "monopoly" for an arrangement that is neither legally nor practically one. They are a major player, but at the end of the day, the concern about companies having to buy their own keyword is missing one specific and important concept... There are alternative search engines if it actually bothered users.

That's what matters! And Google keeps consistently offering the most user value. Otherwise an upstart would displace them with better results.

If you think wedding online advertising to search is evil I won't stop you. But that makes Google no more evil than DuckDuckGo or literally every other player in the space.

@ocdtrekkie right, I think you misunderstood me. I was trying to pin down whether your issue was Google or ad-revenue-backed search engines in genera, because you're making industry-standard practices and industry-standard problems as reasons Google is evil.

If anything, my assumption would be that these problems get worse if that company evaporates. They have the most experience in the industry at this point with dealing with attempts to use an ad engine for malware distribution at scale. I don't imagine Bing would fare much better if they became the primary target.

If your actual concern is that ad-funded search is evil, well, I'd love to see the alternative succeed too. But not even DDG is that alternative; they are also ad-funded.

@ocdtrekkie Let's imagine Google evaporates tomorrow. Do you believe that would eliminate or even ameliorate those two concerns in the medium (or even short)-term?

@ocdtrekkie This is off-topic from the previous conversation and I'm not interested in engaging with it. Suffice to say I'm not at all shocked the search engine those guys built operates differently decades later when neither of them are CEO.

... Besides, I thought you already care not for Brin's character; why are you giving air time to that philanderer's opinions?

@ocdtrekkie @SamantazFox Tell me you don't know anything about the ad industry without telling me you don't know anything about the ad industry. :(

@ocdtrekkie @SamantazFox that's a hard number to swallow without concrete sources. Digital ad spend worldwide was only $521 billion in 2021. I would be absolutely shocked if fraudulent ads were anywhere near 0.1% of the market; every online advertiser would consider that an absolute pandemic and a threat to the existence of the industry, and we'd be hearing hundreds to thousands of reports daily, not single incidents blowing up on Twitter.

@ocdtrekkie @SamantazFox That's unfortunate. :(

FWIW, I don't doubt it happens more than zero but the relevant number that none of us have is malware to safe serves, because that's where you derive the probability that a visit to Google is going to get your system compromised.

I suspect it's higher than people want it to be (and, of course, it's some classic Google obscurity that they don't publish it---not that folks would trust a self-published number from Google if they already believe Google is the Antichrist anyway), but the malware in ads problem is a measure counter measure game with millions if not billions of dollars on the table and both sides grasping for the smartest people they can get to solve the problem for them. No search engine that runs ads as immune to it, and every company keeps their methods secret to detect it for obvious reasons.

https://www.google.com/amp/s/www.howtogeek.com/fyi/bing-is-pushing-malware-when-you-search-for-chrome/amp/

@SamantazFox weird. If I do a search for OBS studio (logged in or logged out), I just get OBS studio.

I have to assume that somebody figured out a clever work around for Google's malware protection and they've already patched it.

@lauren Warfare is racist now?

(Okay, so sometimes probably *yes* but given the term's origins referring to "Applying theoretical knowledge in the field of battle" dating back to the 1400s, I'm finding it challenging to grasp where the racism comes in.)

If anything, changing for the reasons cited by the NPR report seems to support the idea that field work is "worse," which is the opposite of pedagogy (not to mention, y'know... Maybe a little racist!).