Mark Tomczak @mtomczak@qoto.org

@lauren So an interesting wrinkle in the story of this documentary being taken down from YouTube... Is it being taken down at the behest of the Indian government, or at the behest of the original copyright owner?

Because it looks like archive.org killed their links at the request of the BBC.

https://blog.archive.org/2023/01/27/bbc-modi-documentary-removal/

@codinghorror "Yeah right. What good will a Dyson Sphere do you when the sun sets?" ;)

@lauren This leaves the imagination to speculate wildly on the nature of the technical advice.

... I'm going to go with data OpSec. ;) "No, see, if you don't require two-factor here, your talent's gonna get fucked in ways they *weren't* paid for..."

@SwiftOnSecurity Can we develop an algorithm for making fursuits?

I'm thinking start with Stable Diffusion and work out from there. As long as fursuiters aren't too picky on how many fingers their foxtaur has, it'll work great.

Unwritten office etiquette point number whatever:

If you schedule an hour-long meeting, it is most reasonable to provide an agenda; you're asking for an eighth of a regular work day from people; you'd better be telling them why.

... Really, this rule applies to half-hour meetings as well, but half-hour meetings can *sometimes* be summarized by the name of the meeting alone.

The #KSP2 tutorial video on achieving orbit is called "Missing the Ground" and I am already in love. :)

https://www.youtube.com/watch?v=U3DgZsrA-xQ

I think time has shown that resolving namespace conflicts by remapping local names (#Python or #TypeScript style) is a better solution than resolving namespace conflicts by requiring names to be globally unique in the universe of possible names (#Java style). This is a subset of the larger problem "Solutions that require global knowledge are doomed."

There are some downsides. Increases the complexity of finding all uses with a simple text search... But simple text search is the wrong tool to use for comprehending code anyway.

Relative to other kinds of software I use, databases I picture as steam engines, or reactors... The seemingly-constant tuning, indexing, manual analysis and (in the case of PostgreSQL) "vacuuming" remind me of the picture of the engineer crawling endlessly over their machines, turning that knob, tweaking that setting, clearing up that spill when there's time.

It's easy to see how it becomes a full-time job just to maintain these machines.

@lauren I'm concerned with my perception that they cut from unsuccessful projects.

That's not Google's MO and when you couple it with a workforce encouraged to be self-managing, it's harmful to Google's previously-successful strategy of letting engineers try wild things to see if some of them are successful. Engineers who see "everyone who put time into that project got fired" learn quickly to keep their heads down and only select conservative, obvious-revenue-positive bids (like buggy whips before cars).

Good way to build a calcified organization that gets outflanked by the next generation of startup though.

@SwiftOnSecurity If you want to know XBox game release dates, it's a Google NDA you want to sign.

@moddedbear I'm not exactly sure, but I'd guess it backs up your history to your Google account so that it can be restored if your phone breaks.

@lauren Is YouTube pulling them down globally or blocking them for Indian users?

The former is SOP to make it possible to do what they do at international scale; the latter is extremely disquieting.

Tired: "AI is coming for artists."

Wired: "AI is coming for data analysts."

https://www.patterns.app/blog/2023/01/18/crunchbot-sql-analyst-gpt/

@moliver @nova and technically, the backstop against this has been the startup.

The average on software engineering cost was lifted by the possibility that instead of working for you, an engineer could take a couple business courses, incorporate, and compete with you.

@lance @GossiTheDog Exactly the opposite. Malware prevention is one of their major spend sectors in the ad space.

The scale of difficulty of the problem is astronomical. Consider that you can tune an ad to spearphish a target demographic, then you can tune the server response so it only vends the malware payload if the requester sends a request that is fingerprinting correctly to look like the target. Damn near undetectable, even with the techniques Google uses to try and fool malware delivery machines.

The solutions to this problem would be significantly more invasive (such as Google proxying request traffic through their servers to confirm that sites look like they looked when the Google validation checkers requested the page).

They say everyone experiences two deaths.

The first is when your body dies.

The second is when someone reads your browser history.

I'm not sure people understand how much #StackOverflow revolutionized software development.

There were entire *languages* that were unapproachable in the Olde Tymes because you'd get a compiler error like "tycon mismatch" and trying to Google what that meant would come up empty (and yes, you'd read the documentation---if you had it---and it *still* wouldn't explain it in language you could understand without sharing the mental model of the language designer).

StackOverflow had a massive broadening and flattening effect on software dev, and most of the complaints that it changes programming into more of a game of lookup and less of a game of learning the depths of the tools are complaining about the *virtues* of the service in an ecosystem where by the time you've attained deep knowledge of your toolset, the industry has moved on to another toolset or your problem domain has changed.