Loqi[preview] [Stargirl] So You Want to Solve Python Packaging: A Practical GuideFirst, the technical: Python is used by vastly different groups of people, some that don't identify as "developers". Those groups often have disparate expectations about how packaging should wor...
[Murray]I find it interesting how people always state that opt-in cannot work for certain types of projects or that people simply wouldn't ever opt-in in numbers large enough to make them work. It'd be slow, sure, but I think that's more of a "I want to work on this now, and I need data to do that" rather than a reality. I see a *lot* of people in certain communities on Mastodon asking for opt-in search to be a thing, for instance.
Loqi[preview] [[tantek]] the framing as a messaging service is kinda funny, because you've been able to search every email you've received since like forever
[tantek]While the context that is documented in that post (interest rates, pandemic) is accurate (these are matters of history), I completely disagree with the conclusions about the "causes" of said discontent, frustration, and disillusionment
[tantek]1 Twitter. By this I mean everything Derek Powazek said in Argument Machine and the logical extension of the existence of Twitter altering people’s incentives (social incentives are insanely strong) to be toward superficial ego inflation, rather than "stuff that works"
[tantek]2 Finbros reskinning themselves as Techbros post 2008 crash, mass migrating westward, bringing their "culture" with them, and finding out they had a receptive audience in VCs who shared much of a common language, and thus being able to "sell" themselves as knowing far beyond what they actually did. "Visions", "Thought Leadership" etc. just because they said so and knew what words to in a Powerpoint deck
[tantek]3 Prioritizing "DX" (there's the ego thing again) over "UX", and cargo-culting amplified by, yes, Twitter. From "toolchains" to frameworks, there was a massive shift from simple tools that make "stuff that works" FOR USERS to complex tools (that sound good, that BigCos are using/shipping, that companies are hiring for) that most of all marginally made things "easier" for developers to make crap that sorta worked, slowly, unreliably,
[tantek]4 Growing disconnect between making and using (told you there was an IndieWeb connection). All of that contributed to massive shift in software engineers (and the management that ordered them) making crappier & crappier stuff that "worked" worse & worse, which either they never bothered to use in the first place, or gave up using because that wasn't their job / what they were incentivized for, by their managers, CEOs, or peers on
[tantek]^ [KevinMarks] there's my response to IMO the poor reasoning in Kellan's essay where he completely fails to identify the IMO much stronger causes of "why software engineering has gotten less successful" (has become much crappier, made crappier things)
[tantek]tl;dr: amplifying ego is bad, amplifying empathy is good. one way to amplify empathy more than ego is to use what you make, then iterate and fix pain points, and make what you/we all need, and "go to" use what you make.
@LumenPinkRT @mauve@mastodon.mauve.moe
It'd be cool if more sites supported WebMentions (or similar).
For example, it'd be cool if I could have the handy features Github has where they inline references with backlinks, but across arbitrary domains. This'd make stuff like project (1/2) (twitter.com/_/status/1615104800240148483)