@helge I visualize ActivityPub as a very loose constraint in the space of all web based communication networks. Which might be an advantage if you aim for universal adoption (you avoid herding cats) but clearly this means that any real project must impose many additional (undocumented) constraints. But in turn those have implications on the level of interoperability with other projects, network performance and ultimately the utility of the whole vision.
first you create a system with two parties. Then you wonder why you get polarisation. Its by *definition*, stupid.
I mean even the tired left-right one-dimensional axis is not sufficient to reflect meaningful political views. E.g. why would "conservatives" not care about sustainability. What do they want to "conserve"?
Democracy is difficult, imperfect, vulnerable at the best of times. Structure it in archaic ways and watch it transmogrify into bizarre states
@helge i am trying to understand the protocol and all its gory implications but i have difficulty piecing together a logical picture. Disconnected comments/complaints from implementers help point out what are important/ill defined topics but further fragment the overall coherence. Is there like a concise mathematical distillation of what is supposed to happen (and what actually does happen)? Or rather, what is the closest approximation to such a terse but complete depiction of information flow.
@jayvii using firefox on ubuntu via snap (after a system upgrade) and not happy at all. Horrendous start up time (to actually load first pages). Willing to try alternative channels.
Major shocks, like a pandemic, or minor ones, like an evening without electricity are constant reminders to people used to the blessings of functioning health and energy systems not to take those things for granted - because they are not.
Fragility, unsustainability, conflict, false accounting, delusional founding myths and narratives create a toxic mismatch between expectations and realisations, a bottomless downward spiral.
@steve from my very incomplete understanding of the spec it seems to be occasionally mixing practical, implementation specific concerns (pagination, handling the firehose problem) with logical structure.
"Meaningful work strives to make a meaningful difference in some aspect of the world we live in. Making a meaningful difference on a global scale will take the best use of the minds and machines that our society can muster. Machines need direction from human minds and human minds need direction from human leaders."
These are the end sentences from "Ideas and Information", 1989 by physics Nobel prize winner Arno #Penzias, who passed away yesterday.
@researchfairy its not a pipe dream. Old school corporate sectors that work with people's data, whether its the medical industry, banking, insurance etc dont get such a free pass (They do capture regulators, mind you, but in different, less abominable ways).
#bigtech is an aberration in terms of what they were allowed to do to us. A moral bankruptcy without much precedent which was normalized only thanks to the incredible and improbable profitability that goes with such a major theft and abuse.
@sirber@amin scratch the mighty #python and invariably you find some C/C++ lib that is as high performance as can get. Case in point the #lxml lib for parsing whatever you fetch from the web... https://pypi.org/project/lxml/
Where python speed will let you down performance wise is if you write "naive" code for a computationally intense task and dont use something like #numpy, #cython etc.
@vid if you click through you get links to source code, API, forums etc. The issues you mention seem to be on the radar of developers. An important site in this respect (to see where things are heading technically) is the collection of Fediverse Enhancement Proposals. Quote: "The goal of a FEP is to improve interoperability and well-being of diverse services, applications and communities that form the Fediverse."
@jgoerzen I think we should be acting with all our energy towards positive outcomes even if it seemed "hopeless": We simply don't know the future with any degree of certainty. Assuming too much on the negative is unwise too. And even if we were more certain, what else is there to do anyway?
But the #doomerism is understandable. In this day and age *all* our problems are due to social malfunctions of one form or another. What is missing is in-your-face evidence that we are learning any lessons.
Seems to me we are well on our way towards that inspiring vision. Kinda hard to be objective but it feels that the #fediverse mindshare is reaching critical mass https://fedidb.org/software
My sympathy for the "not another twitter" feeling is moderated by the fact we wouldn't be here if #mastodon was not around to offer a tangible alternative at the right moment (the federation concepts behind #activitypub are not particularly new and were for a long time in limbo)
@kta@scottjenson did not know about #sway, after a quick look it feels indeed like an interesting take in a fertile direction (For most of the time fighting windows with a mouse feels like a waste of time and energy without obvious benefit).
Is it a mode that could be reasonably emulated in #kde?
Thinking that the concept might also give a usability boost to small android screens running #termux (some of us still dream of physical keyboards on mobile)
I don't think there is a substantial difference in #UX between #MacOS and #Win11. There are lots of visual 'skinning' tweaks but the basic structural model of files/folders/2d windowing/invisible clipboard feels pretty baked at this point.
Are there any #Linux distros that break the mold? Shake things up a bit?
My point is that I feel *all* desktop #UXs have pretty much stagnated and no one is really trying anything different. I'd LOVE to hear of any crazy experiments.
@missmythreyi you cant create something out of nothing. But making optimal use of scarce information may be facilitated by validating ideas using datasets that are not limited by sample size.
To put it otherwise: if your algorithm, tool or whatever doesnt perform with plenty of synthetic data, it wont perform with scarce real data.