BLU February Meeting
I attended the monthly Boston Linux/Unix Group meeting for February (but BLU talks are usually geared towards Kubertes-like corporate topics & there are old Vietnam War pilots talking about flight school).
However (due to an uninteresting IRL situation), I was sleep deprived and fell asleep for most of it. I genuinely didn't want to, but I was operating on less than 4 hours of sleep for no good reason.
(Sort of like a physics seminar... please see this joke physics research paper for reference. Also, U Illinois of Urbana-Champaign reposted an excerpt without even providing a link to the original??? I can't believe it...)
All I know is that Randal Schwartz had a talk on Flutter/Dart for a cross-platform app language for web, desktop, iOS, and Android.
(I actually dreamt of a more interesting version of the talk, but maybe that was the earbuds in my hear playing the talk while I slept — sort of like those MKULTRA subliminal messaging experiments.)
Below are my thoughts (with no good order... even though there should be some order).
- Some of the attendees have met Linus Torvalds when Git was only 2 months old in the 1990s.
- Schwartz had asked Torvalds if Git worked on Mac or the *BSDs.
- We definitely need more young ppl doing Linux (but no established YouTubers allowed). It felt simultaneously boring/intimidating that there were so many Boomer age people around. Though I don't think they did Boomer things, it didn't make me want to participate, even via chat.
- BLU needs to have a Matrix bridge, or at least advertise its existence if it already exists. It lists an IRC channel, but most IRC channels already migrated to Libera Chat after the Korean crown prince took over IRC in early summer 2021.
- What a wonderful
gifthorror this pandemic has given us...
- What a wonderful
- I'd like to see 1 hour (max) demo talks or talks that are more interactive/engaging with the audience with more opportunities for questions, instead of scheduling more formal 2 hour presentations with questions.
- I'd like to see presentations on newer things like Qubes OS, GrapheneOS, CalyxOS, LineageOS, and more modern things.
- Flutter is way too cutting edge (if not bleeding edge) for established projects to be used in the real world.
- I see major low-level struggles right now for existing projects like Signal merge their iOS/Swift and Android/Java-Kotlin into 1 unified Flutter software project.
- But Flutter may be worth looking into for completely new projects.
The last major non-indented point is of most importance to me. There's nothing really inherently wrong with Flutter/Dart as I can tell (as a non-programmer). However, I am especially concerned with this social/interpersonal dynamic of setting up Google to practically be the de facto central point of failure in everything hardware and software related.
I can see where Google sees benefit of creating entire programming languages. In fact, many of the packages I use from the AUR use Go (which is literally the first two letters of “Google”). However, unless Dart/Flutter can demonstrate that its people can stand freely enough away from Google, like Rust, I won't be using Flutter.
I think in principle (in which lots of ideas can seem fine) most of what Google does is neutral good. Most of Google's software libraries and projects are open source.
I know that “enumerating badness” (wherever it came from — maybe a privacy/security methodology explainer) is frowned upon but let's do it anyways, ok?
First, you cannot look at any commits, issues, or pull requests on Google projects without signing in with a Google account. Even GitHub looks like an angelic Mr. Rogers in comparison and possibly forgivable for working with ICE... but I'm not that naïve.
This has happened because I tried looking at an upstream contribution from GrapheneOS — however, I couldn't even see it because I needed to sign in with a Google account.
Not so small mini-tangent: Google announced the CameraX library in mid-2019. However, now it's early 2022 and the CameraX library is still in beta (or barely out of beta)! You're telling me that a Silicon Valley company that has comparable job acceptance rates with Facebook is more difficult to land a job offer than it is to apply to Harvard for undergraduate studies? Major BS! And Google still can't deliver within 2.5 years! The GrapheneOS devs are doing most of the usability work for free. Honestly, I would start reading GrapheneOS's build guide to genuinely learn how to build AOSP from source and trust it more than Google's own documentation as a Google coding employee. Google's documentation sucks and I can barely read it. It's like you have a 5-year-old with ADHD an adult programmer brain for 24 hours on Adderall and the mess of an end product accurately describes all Google documentation I've seen so far.
Secondly, nobody trust Big Tech companies, especially Google, and I don't think Google should have a large of a role as it already does in open-source. In the very long run, GrapheneOS wants to entirely abandon Android altogether. This is so that Google will not inadvertently make GrapheneOS as a project bow down to every single one of its whims and wants as a for-profit company. Google is already doing weird stuff with web compliance because none of their actual run of the mill employees can hold a candle (let alone an iota of a candle) of Linux knowledge to even a fictional character, such as Elliot Alderson from Mr. Robot. I already know from Brian Lunduke, even as a person I regard as a person who talks too much and doesn't do enough action, that Big Tech companies are already playing Deus Group-tier moves to emasculate the Linux Foundation and other open source organizations/projects.
Lastly, mostly everyone cannot have enough nuance to discriminate the good parts of Google from the bad parts. Most of the neutral to good parts are the open source stuff that nobody who just wants to watch PewDiePie on YouTube proper will ever understand. Honestly, there's a lot more stuff that's open source from Google than one would expect. However, it's not “exciting” stuff. Mostly software libraries, which are the backbones of apps. It's mostly the “end products” of the software stuff that Google does that is where its very small yet potent “evil” Siamese twin come out to try to kill everyone... kind of like the spoiler in the film Malignant.
Chromium is open-source, but Chrome is closed-source. I'd say that Chromium is more than 95% of the “assignment” to make a browser, but Chrome is the final part less than 5% that is the figurative fly in the archetypal ointment. Chrome is the “end product” in which Google makes evil via telemetry, other tracking, and signing into your Google account — as part of a non-exhaustive list.
I'd estimate that more than 95% of stock out of the box Android is AOSP — it's the heavily integrated Google Play Services and preinstalled proprietary Google apps that can't be install which ruin the standard Google Pixel experience for me.
Obviously, some products/services are more of a black box than others. I'm sure there is a curtain darker than a black hole cast upon the source code for Google Search. I'm also sure that the Titan M chip (or the Tensor SoC on the Pixel 6/6 Pro) is a black box. It's good that the security implementation of the Titan M chip is open and has an open bounty offer for security researchers to try to test its actual security IRL, but we still have to trust Google that it's not making the Tensor actually evil — so, it doesn't pass the Andrew “bunnie” Huang test of electrical engineering. And of course, YouTube is closed source, but it has always been that way, even before Google acquired YouTube.
Conclusion
Basically, what I'm saying is that Randal Schwartz was being a little too upbeat and positive on Google because he is a GDE (Google Developer Expert). However, somehow he couldn't recite that Swift is the primary language for iOS app development and that Kotlin is currently the de facto language for Android app development very quickly.
Also, why do people use the telemetry-filled VSCode and not the truly free VSCodium???
Lastly, when is that first non-Pixel phone that works with GrapheneOS coming out? This needs to come out sooner than later so that uninitiated people who don't understand or even haven't even tried installing GrapheneOS keep saying stuff like this:
GrapheneOS is weird. They ship a privacy-first OS which runs only on hardware sold by a privacy-last company.
You can get a privacy-respecting phone, but you're financing a company pushing for the exact opposite. — Hugo Barrera (@WhyNotHugo) on Feb 26, 2022 · 5:52 PM UTC
Yes, I do hate the situation we currently have in 2022. My family member would ordinarily retort with, “So, what have you done for a secure smartphone?” However, I don't think that sort of confrontational rhetorical question will be received well on the already clickbait hate-filled Twitter. So, instead I'll refer any readers(reader) to this video from Side of Burritos on YouTube for an explainer on why the currently supported Google Pixels are pragmatically the best smartphones to use for local security — sort of like how good dark and moody Batman villains are fundamentally and inherently correct, even after they are defeated for their errant ways.
And to show that I'm not a GrapheneOS Matrix room NEET simp, I think GrapheneOS should stop tweeting at people like Sylvia van OS.
Also, the move for GrapheneOS to stop using Reddit and Twitter is both good and bad: hopefully I'll only get actual updates on GrapheneOS's Twitter so that I can finally add GrapheneOS's tweets to Newsboat from Nitter's RSS feed. It's bad because now the Daniel Micay puppy followers will now enforce even worse acts of groupthink that chased Micay out of Rust and CopperheadOS into his own universe that he personally controls.
Snowden should really set Micay straight on being a scumbag...
You do realize that Linus Torvalds, Richard Stallman (and by extension the entire FSF), and Eric S. Raymond are no longer relevant?
Though I doubt Micay will ever “revert” back into a beginner like Torvards and Stallman, I expect what comes around goes around and that Micay's continued toxicity from his Rust days will cause others to not promote GrapheneOS as much as it deserves on a technical level.
Even Moxie Marlinspike isn't exactly relevant anymore, but Marlinspike is a good exception. I think it's good that Marlinspike can relax and catch a breath, even if Signal has shortcomings with centralization and offering a service being still tied to (cell) phone numbers.
I hope someone (non-corporate) takes the lead to succeed both the best security parts of GrapheneOS with the communication skills of CalyxOS/the Calyx Institute, but right now I can only hope that the BLU doesn't die out soon...