#Fediverse I know this is a bit late, but I recently found out that mastodon.technology is shutting down. If you haven't found your new home (which you probably have in fosstodon.org or other great Mastodon instances), I'd like to welcome you to move to technodon.social (which I started today / few hours ago). Looking forward to use this instance to bring tech folks together and share some good tech content among professionals and enthusiasts #twittermigration #introduction 🙏

Part of the reason I advocate for Guile and Guix is that, especially because of Guix, there's been a lot of effort spent on trying to be nice and friendly towards people. The crustier end of things tends to come from some of the crustier parts of GNU, but in many ways Guix also is a redemption arc for GNU, a place where GNU's better ideas have been revived and made interesting.

Show thread

Amazon has bought:
• Blink cameras
• Ring home security
• Eero Wi-Fi
• iRobot Roomba

They tried to coerce Ecobee into handing over private data and when they refused, Amazon retaliated by creating a competing smart thermostat.

ok #lazyweb I have found a few myself, but:
does anyone have any favorite guides for best practices/etc. for a home server? one of the things I want to do post-diss is to deplatform all my shit, so I think that means having at least two servers: one personal one for photos, videos, calendar, personal website, wikis, etc. and one public one for code, hosting external web stuff, etc.
thinking my hardware for personal and vps for public???

"Innovation" has become a curseword, thanks to...innovation. Some of the world's most imaginative, best-funded sociopaths have spent decades innovating ways to fuck you over. While the whole tech sector likes to get in on this game, no one "innovates" like inkjet printer companies.

Printer companies are true fuckery pioneers: the tactical innovations they've developed in the war on their customers would make Otto von Bismarck blush.


I like that my Anker bot has no wifi and no map, just random walk

Oh sure, but regardless of if I am on tumblr, twitter or fedi, 98.5% of my output is reblogs, so… 🙃

Not that I am for the reddit homogenization, but … anything but PHPBB! Discourse is the main one I know with a nice, modern and usable interface.

Unlock Lisp / Scheme's magic: beginner to Scheme-written-in-Scheme in one hour!

YouTube: youtube.com/watch?v=DDROSL-gGO
Peertube: share.tube/w/gdtnuipKbbVdR2u1m

This is the live talk version of A Scheme Primer, as published by @spritelyinst: spritely.institute/static/pape

Never has human body language said “your problems are not my problems” as clearly as a cat’s. #cats



Literal any other uncompressed or lossless compressed format would be better than PSD


We should be using ASCII delimited values. We have perfectly good control characters *right there* and we didn't use them…

> It is even part of the design of the file encoding system. The ASCII standard calls these fields

31 Unit Separator
30 Record Separator

And ASCII has two more levels with Group and File Separators

29 Group Separator
28 File Separator


@oreolek @ewankeep
yeah QC will always be at most highly specialized side computer (maybe add in card or co-processor if you really want to stretch). A QC will never run a general purpose OS with "apps" and UI and such

OOXML (docx, etc) is waaaaay uglier than ODF. It might outlast ODF, but if so it will simply be due to MS power.

I suppose in theory we could see epub take over as a production format for word processing (not just end user)… but wedging spreadsheets and other "office" stuff into epub seems like a bad idea. And even for just WP, epub would at a minimum need to grow stuff like revision support…

@ewankeep@mspsocial.net hi! I'm a fantasy writer but also a tech person (sadly), so you might find my opinion an interesting intersection.
File formats have a couple of interesting intersections in what they can require and achieve.

The simpler a file format is, the easier it is to (re) implement from scratch (something that may end up being needed in case of a catastrophe), and ensures it being commonplace in some regard (such as by just... being really widespread). This is the case with .txt

These "simple" formats are fairly limited, though. For instance, the audio one like this is the .wav, and the image one is .bmp
You might notice they are huge and have significant limitations.

One way to go "beyond" those is by having non-software (or software-optional) approaches to extension - this is why markdown (.md) is so commonplace - it works as a plain .txt but you can also do additional things with it.
This includes stuff like spreadsheets - via org mode, for instance, where there's some automatic processing that can take place.

Some formats, meanwhile, are defined by the data model, which in turn are influenced by the underlying material conditions. Sqlite works great when B-trees work on your storage medium, and not so much when they don't.
These formats aren't quite as "simple" as the "simple" formats, but they are less restricted, more featureful, and maintain a "simplicity" due to their ties to what is underlying them.

The thing is, these data model formats really do strongly depend on the "underlying" tech. Oh, your civilization found a way to encode data in crystals in a wide variety of applications? Out go all of your featureful formats!

So there is a constant balance to be had between "simple" formats and "optimized" formats - both of which are "forever", but with specific conditions attached (not needing additional functionality and the underlying data model being applicable respectively). It's only the "complex" formats that are unlikely to last - ones that optimize for the specific use-case without being logically tied to the underlying data format (most of the examples you've given).

So to me, it would depend on the setting and the growth of the civilization - what kinds of technologies did they discover? This would
directly determine what "old tech" they still use, and what would feel alien to them.
One example of this already happening is tar - tar archives made sense because of the tape data model (Tape ARchive), but we still use them for historic reasons ... except outside of unix, where zips are far more commonplace (being a simpler archival format, though not by that much).

So it's very much something that I would personally reason through for any given civilization, setting, etc.
Though I also recognize it'd take quite a lot of technical knowledge to be able to do this ^^;;
Good thing I can just write about cursed books of knowledge instaed

I mean wouldn't want to encode a jpg as XML, but for text like data, XML is probably one of the best options. XML got a bad rep due to misuse (config files etc) but in the right place it is great. ODF uses it, even MS (ab)uses it for docx etc.

Or HTML+CSS is pretty good for display optimized text, which is what epub uses.

Increasingly not, huge numbers of things that play mp3s now are full computers, often android devices (smart TVs/sound bars/whatever, some auto entertainment systems, etc). In the US certainly I basically never see iPods/other "mp3" players anymore. And ton of mp3 players could/can play ogg vorbis, flac, aac, etc

re: file formats 

@scruss @malte @clarfonthey
32 bit machines should be perfectly possible to make if we are making computers at all. Plenty of old mainframes built with macro transistors (not chips) had large word sizes like 32 bit and 36 bit. Making them powerful and fast might be harder…

Show older
Mastodon for Tech Folks

mastodon.technology is shutting down by the end of 2022. Please migrate your data immediately. This Mastodon instance is for people interested in technology. Discussions aren't limited to technology, because tech folks shouldn't be limited to technology either!