evenwicht

Dumbing down of technology and competency over the past decade has led to a number of situations where sophisticated users are actually the ones being marginalised.

• PDFs are being wrapped with some exclusive JavaScript garbage that only works for GUI users. Terminal users are losing the ability to simply download PDFs. Links to files with a “.pdf” extension are often not actually PDFs anymore - they are HTML w/js embedded masquerading as PDFs. Mozilla is on board with this deception.
• Ethernet: public libraries have disabled ethernet ports, mostly. Some librarians even go apeshit when someone plugs into them (not understanding that it’s another way for wifi-less people to connect). You either subscribe to mobile phone service & disclose your number to pass the captive portal’s verfication, or you can fuck off, as far as the library is concerned. And yes, people are generally okay with /public/ libraries excluding people this way.
• The value of compatibility is totally lost. Young network admins just assume everyone runs the same latest browser as them, and that everyone has a recent model smartphone. If you don’t buy a new phone every couple years, they believe it’s your fault you’re excluded. The concept of design and engineering for compatibility is a lost competency. The word “compatibility” is becoming history despite the decline of interoperability. Soon dictionaries will tag the entry for “compatibility” with “(rarely used)”.
• We can no longer access public services like court system search tools, business registries, and public libraries book catalogs from a text terminal. The drive to dumb everything down has led to fancy UIs that work with fewer clients.
• Access restrictions block access to resources unless you have a non-Tor IP address. Sophisticated users know better than to expose their personal IP addresses while also exposing to their ISP where they go. Sophisticated users are in such a small minority that it’s trivial to oppress them.
• Using asymetric encryption to protect email payloads was a thing in the 90s. Who predicted that we would /devolve/ to 100% in-the-clear email payloads ~25 years later?

There are a lot more examples but to cut to the chase: How did we fuck this up?

Instead of teaching users to become sophisticated, as a society we just threw in the towel and decided we cannot teach people.. that they cannot even learn the speed and utility of terminals and keyboards. So we said “fuck it, give everyone a GUI and a mouse”. And so now we are at a point where even the technicians themselves seem to be helpless without a GUI and mouse, so they are oblivious to the demographic of users who are slowed down by their UIs.

Then we decided: since everyone has a GUI and a mouse, throw graphical CAPTCHAs their way. Surely no one uses terminals anymore, right? And why stop there.. get rid of documents (simple HTML).. make every webpage an /application/ instead, because surely everyone can run any random JavaScript we shoot their way.

This is not to say low tech users should be left behind. Indeed some people are truly incapable of terminals, scripting, Tor, PGP, etc. The problem is catoring for the tech illiterates exclusively results in disempowering sophisticated users.

It parallels the situation where classroom instruction moves so slow for some of the faster learners at the top of the class that they get bored and drop out of school, and waste their potential. I’m at a point where I’m fighting to retain an analog life because the digital workflows being pushed on us are so dumbed down that I just cannot accept being forced to click through shitty oppressive technology that forces interaction with tech giants and walled-gardens.

If I could choose between broadband with today’s garbage (ads, CAPTCHAs, Cloudflare, anti-bot, anti-tor, …) and 9600 baud dial-up to garbage-free text services that just work, I would seriously choose the latter. I am serious about that.

Anyone know of a template or sample doc that prints markers around the edge of an A4 paper?

Or even just a good centralised reference?

I can’t believe what shit results my searches are getting. Surely this must be a common need for millions of people. I am not going to go to the printshop, write down their printer model numbers, try to locate online manuals in an ocean of shitty manual sites, to try to dig up the printable area specs, which are likely untrustworthy anyway. I’ve done that before, and IIRC Canon specs were a lie.

Canons seem to have a quite large unprintable area. I know Ricoh does better. It would be useful to see a centralised table with the printable area specs of (at least) all the large industrial printers.

\documentclass[DIV=66, draft=true]{scrartcl} % The draft switch produces a ruler along the boundary of the printed space (which is controlled by the DIV value)

Update1: CUPS test print reveals unprintable area dimensions

It’s worth noting that the test page for CUPS gives “media limits” info. Which is vague but seems to correspond with the printer’s edge of printable area. It’s unclear if that comes from the printer driver or if the printer is somehow queried for that info.

This is of course only useful if you’re not using a print shop.

Update2: came up with code to generate a test print:

% Purpose:
%
% 1) Test whether the unprintable region documented in the printer specs is accurate.
% 2) If not, find the real dimensions.
% 3) Find the maximum DIV setting for the KOMAscript package that does not encroach into the unprintable area.
%
% Procedure:
%
% 1) Lookup the expected unprintable area dimensions for the printer under test.
% 2) Edit SetBgContents below to match the dimensions, which are added to (current page.*)
% 3) Trial and error/tuning: Set DIV=99 and compile. Then set DIV=9 and compile. Notice how the rectangle ruler gets smaller as DIV gets smaller. Find the max value for which the rectangle does not go outside of the violet rectangle.
% 4) With DIV at the max, fiddle with the size and position parameters of the large circle (in DeclareNewLayer). The goal is for the circle to touch the top and bottom edges of the paper.

\usepackage{scraddr}
\usepackage{scrlayer-scrpage} % needed for \cofoot
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc} % suggested to avoid ``OT1 encoding''
\usepackage{pict2e}
\usepackage{scrlayer}

\usepackage[firstpage=true, color=violet]{background}
\usepackage{tikz}
\usetikzlibrary{calc}

% from another suggestion below:
\SetBgPosition{current page.north west}% Select location
\SetBgOpacity{1.0}
\SetBgAngle{0.0}
\SetBgScale{1.0}
% \SetBgColor{black}

% The line width setting below specifies 1pt but it really looks thicker compared to other lines. Nonetheless, it gives a good thickness for the job.
\SetBgContents{%
  \begin{tikzpicture}[overlay,remember picture]
    \draw [line width=1pt]%,rounded corners=4pt,]
    ($ (current page.north west) + (4.2mm,-4.2mm) + (1pt,-1pt) $)
    rectangle
    ($ (current page.south east) + (-4.2mm,4.2mm) + (-1pt,1pt) $);
  \end{tikzpicture}}

% The following gives circles and must /follow/ the tikz stuff above.
\DeclareNewLayer[%
textarea,background,mode=picture,
contents={%
  \putC{\circle{\LenToUnit{\paperwidth}}}%
  \put(0.5\layerwidth,0.5\layerheight-3pt){\circle{\LenToUnit{\paperheight}-0pt}}%
}
]{showtextarea}
\DeclareNewPageStyleByLayers{test}{showtextarea}
\pagestyle{test}


\begin{document}
\phantom0 % There must be /something/ here or else 0 pages are generated. So we put an invisible phantom object.
\end{document}

I was reading EU Directive 2019/882 which mandates accessibility requirements for products and services -- not because I am disabled or impaired but I’m always looking for legal angles to use against enshitified products/services or to liberate data. Not much interesting law except this:

Section IV
Additional accessibility requirements related to specific services
…
(b) Services providing access to audiovisual media services:
(i) providing electronic programme guides (EPGs) which are perceivable, operable, understandable and robust and provide information about the availability of accessibility;

IIUC, the EPG tech is already baked into DAB radio standards. But many broadcasters do not exploit the option, and even fewer receivers make use of it. In fact I have never seen a DAB radio that exploits EPG info (only album art and metadata for what’s playing at the moment).

It falls a bit short of being complete. Broadcast services may have a legal obligation to send EPG info, but I see no requirement for hardware to exploit it.

(hope no one is bothered by the post-brexit irrelevance of this.. it’s the only free-world DAB forum in the threadiverse)

Europe’s block on boycotting banks

No law directly prohibits boycotting banks AFAIK, but it’s effectively illegal to boycott banks because:

• It is illegal to be paid wages in cash in some (all?) countries.
• Some EU countries governments insist on tax payments by bank transfer. This is for all kinds of tax (income tax, property tax, and other forms of tax).
• EU level: all cash transactions above €10k are illegal in the whole of Europe. Most of western Europe reduces that limit to €1—3k.

Belgium’s ban on boycotting energy suppliers

Offgrid energy is illegal.

Denmark: you cannot boycott email, as of this year

Denmark eliminates the postal service this year. This essentially means you cannot boycott email because the snail mail option is generally gone. Exceptionally, you can perhaps send letters using UPS or FedEx, but that’s not really affordable if you are boycotting email. Not sure if hand-delivery is an option. Consider Germany, where postal boxes are not always public access and couriers are given a key to the lobby. If that happens in Denmark, then hand-delivery cannot be relied on.

US ban on boycotting Israel

You can boycott the US in the US if you want, but you cannot boycott Israel if your job is from the US government. This tyranny was showcased in Texas when a Palestinian school contractor who taught kids how to speak Arabic had to renew her contract. The new contract required her to agree to not boycott Israel. She could not in good conscious sign such a bizarrely oppressive contract, so she was let go.

cross-posted from: https://lemmy.sdf.org/post/41050620

Spy chips are:

• Intel CPUs after ~2008
• AMD CPUs after ~2013
• Arm CPUs (not sure when they started the trustzone stuff but likely around 2013 since AMD uses trustzone)

I believe IBM Power9 chips are spy chip free, but not sure about the successors.

Anyway, the question is about Apple chips. Web searches are lousy these days. I find nothing to confirm or deny the presence of management engines in Apple (Motorola?) CPUs.

Intuitively, I don’t think it would make business sense for Apple to do that because a majority of their customers are non-corporate individuals (unlike intel). OTOH, if that were sound logic then it would seem to contradict Arm chips which are also largely bought by non-corporate individuals.

Anyway, if anyone knows plz mention it here, ideally with a source.

Thanks!

Spy chips are:

• Intel CPUs after ~2008
• AMD CPUs after ~2013
• Arm CPUs (not sure when they started the trustzone stuff but likely around 2013 since AMD uses trustzone)

I believe IBM Power9 chips are spy chip free, but not sure about the successors.

Anyway, the question is about Apple chips. Web searches are lousy these days. I find nothing to confirm or deny the presence of management engines in Apple (Motorola?) CPUs.

Intuitively, I don’t think it would make business sense for Apple to do that because a majority of their customers are non-corporate individuals (unlike intel). OTOH, if that were sound logic then it would seem to contradict Arm chips which are also largely bought by non-corporate individuals.

If anyone knows plz mention it here, ideally with a source.

Thanks!

I often acquire quite old hardware either cheap 2nd-hand or rescue stuff dumped on curbs typically w/out drivers or s/w. Ultimately all h/w will eventually be used on linux. But linux is often not ideal for testing to quickly assess whether something functions well -- obviously because very little hardware is designed for linux.

So before investing time researching linux drivers and hacks for whatever obscure thing I am dealing with, I need to quickly test whether the thing works without searching forums for what complex installation procedure worked in Bob’s basement lab.

Apparently Windows is very dicey with both forwards and backwards compatibility. I thought win7 would be good for testing because it’s historically close enough to XP that things designed for XP might run on it, yet just barely new enough that hardware ~2—10+ yrs old might likely run on it.

But it seems to be more of a shit show than I expected. Some drivers demand a specific version of Winblows. Matching OS version is often not good enough either because they demand a particular service pack, or specific DirectX or “.Net” version (what a shitty name, btw), which cannot be too old OR too new (e.g. old TomToms are extremely fussy about .Net version IIRC). So even though some form of Windows has the best official support for any given piece of hardware which underwent the most rigorous of its testing on Windows, using Windows for testing hardware is a shitshow nonetheless. Plus I make it worse because I insist on Windows boxes being airgapped, which limits me to drivers I can get off the web and usb-side-load.

A virtualbox with a few different Windows VMs is not good either because virtualisation brings its own baggage of issues that blow the idea of quickly testing arbitrary hardware to confirm that it works.

Is hackintosh a better solution?

I will not be buying any recent Apple hardware. Fuck that.. the cost defeats the purpose. I can (reluctantly) get really old Apple machines cheaply, but I suspect those tend to be incapable of any somewhat reasonably recent MacOS version. So I am tempted to try the hackintosh route on an old PC. Is it safe to say that MacOS drivers are more flexible across various MacOS versions than windows?

It has been decades since I tinkered with hackintoshes.. is it still practical these days? I get the impression that it might still be good for my purposes (but perhaps not in ~5+ years from now considering this).

cross-posted from: https://lemmy.sdf.org/post/40711081

Most Lemmy users are vulnerable to data loss arising out of an admin spontaneously pulling the plug on their Lemmy instance. I have lost data several times by this cause (both on Lemmy and on Mastodon). Infosec includes availability (thus backup copies), but this has been neglected by developers of clients for fedi platforms.

Mastdon has /something/, at least

We have a crutch for Mastodon: mastodon-archive by Kensenada. It only works on some¹ Mastodon instances, but when it works it’s a quite useful tool. It uses the API to grab all posts you author as well as posts by others who mention you. It would be even more useful if it would grab whole threads for which you participate or bookmark regardless of mentions, but last time I checked there is no plan to implement that. You don’t even have a copy of the parent messages you reply to. And (IIRC) you also don’t get a copy of mentions in situations where fedi barriers prevent responses from reaching the instance you are on.

¹ Some instances are simply incompatible for unkown reasons

What Lemmy needs

A gnu linux tool to fetch whole threads that the user starts as well as whole threads for which they comment. Ideal features:

• produce a searchable local SQL database.
• optionally, grab threads or posts the user upvotes.
• optionally, detect cross-posts and grab those threads too.
• periodically revisit the thread to record new activity, including moderator actions. The period between re-visits should get increasingly longer as the thread ages.
• when an author deletes their post, it should be marked as deleted in the local DB. And users should have the option to have those records purged automatically or selectively purged upon review.
• (science fiction?) get the current host to digitally sign something certifying that the user’s profile/content is the genuine original artifict for the purpose of migrating to another host. The current Mastodon migration mechanism is dysfunctional for cases of a host going down before migrating, and I assume Lemmy might have the same issue.
• fedi politics circumvention: give users the option to grab copies of the same thread from other instances so a browsing tool can compare the various thread versions, suppress dupes, and show the most complete aggregated version.
• for extra credit: integrate the DB with @theblawsybogsy@lemmy.ml’s emacs “Lem” app as a front-end for offline browsing

It’s important for user retention

When a user puts a lot of effort into producing content only to lose it all on the whim of an admin deciding out of the blue to kill the server, it’s demoralising. The user might opt to abandon the fedi or start over from a giant centralised walled-garden like LW. In both cases the decentralised free world shrinks.

It’s important for digital sovereignty and fedi-balance

There are already users who conciously decide to pile onto the biggest instances for the perception of stability. Nervous Bob might have a specific passion for a small mission-focused instance like lemmy.radio, lemmybefree.net, mander.xyz or linkage.ds8.zone, but is risk-averse. He cannot stomach the thought of losing all content and believes that if an instance is large, the admins will be more careful.

Having an archive settles the nerves of Nervous Bob enough to be able to follow his passion. It disables the cognitive dissonance of licking the boots of an oppressor (such as a Cloudflare instance).

Why this is (or will be) posted in !spreadfediverse@flamewar.social

Some would say information security is essential -- a precondition to transitioning into the fedi. Reguardless, such an app would serve to encourage people to contribute to the threadiverse and ultimately the proportionate growth and spread of it.

cross-posted from: https://lemmy.sdf.org/post/40711081

Most Lemmy users are vulnerable to data loss arising out of an admin spontaneously pulling the plug on their Lemmy instance. I have lost data several times by this cause (both on Lemmy and on Mastodon). Infosec includes availability (thus backup copies), but this has been neglected by developers of clients for fedi platforms.

Mastdon has /something/, at least

We have a crutch for Mastodon: mastodon-archive by Kensenada. It only works on some¹ Mastodon instances, but when it works it’s a quite useful tool. It uses the API to grab all posts you author as well as posts by others who mention you. It would be even more useful if it would grab whole threads for which you participate or bookmark regardless of mentions, but last time I checked there is no plan to implement that. You don’t even have a copy of the parent messages you reply to. And (IIRC) you also don’t get a copy of mentions in situations where fedi barriers prevent responses from reaching the instance you are on.

¹ Some instances are simply incompatible for unkown reasons

What Lemmy needs

A gnu linux tool to fetch whole threads that the user starts as well as whole threads for which they comment. Ideal features:

• produce a searchable local SQL database.
• optionally, grab threads or posts the user upvotes.
• optionally, detect cross-posts and grab those threads too.
• periodically revisit the thread to record new activity, including moderator actions. The period between re-visits should get increasingly longer as the thread ages.
• when an author deletes their post, it should be marked as deleted in the local DB. And users should have the option to have those records purged automatically or selectively purged upon review.
• (science fiction?) get the current host to digitally sign something certifying that the user’s profile/content is the genuine original artifict for the purpose of migrating to another host. The current Mastodon migration mechanism is dysfunctional for cases of a host going down before migrating, and I assume Lemmy might have the same issue.
• fedi politics circumvention: give users the option to grab copies of the same thread from other instances so a browsing tool can compare the various thread versions, suppress dupes, and show the most complete aggregated version.
• for extra credit: integrate the DB with @theblawsybogsy@lemmy.ml’s emacs “Lem” app as a front-end for offline browsing

It’s important for user retention

When a user puts a lot of effort into producing content only to lose it all on the whim of an admin deciding out of the blue to kill the server, it’s demoralising. The user might opt to abandon the fedi or start over from a giant centralised walled-garden like LW. In both cases the decentralised free world shrinks.

It’s important for digital sovereignty and fedi-balance

There are already users who conciously decide to pile onto the biggest instances for the perception of stability. Nervous Bob might have a specific passion for a small mission-focused instance like lemmy.radio, lemmybefree.net, mander.xyz or linkage.ds8.zone, but is risk-averse. He cannot stomach the thought of losing all content and believes that if an instance is large, the admins will be more careful.

Having an archive settles the nerves of Nervous Bob enough to be able to follow his passion. It disables the cognitive dissonance of licking the boots of an oppressor (such as a Cloudflare instance).

Why this is (or will be) posted in !spreadfediverse@flamewar.social

Some would say information security is essential -- a precondition to transitioning into the fedi. Reguardless, such an app would serve to encourage people to contribute to the threadiverse and ultimately the proportionate growth and spread of it.

I posted the linked article and now I need to edit it.. some of the formatting looks terrible. But I am blocked. The editor lets me edit, but no matter what I do I cannot save the edits. There is a pop-up about forbidden language.

It’s an old Lemmy bug to not tell authors what word is the problem. But very bizarre that it accepted the post in the first place if it had a problem with it. My edits are certainly not causing that.

The post is not on sdf.org, so I suppose the problem could be with the hosting server.