mtjm @mtjm@mtjm.eu

One of the things I like about Rust's memory safety is that rustc randomly crashing motivated me to run memtest and find a faulty DIMM. I too often assume that random segfaults and hangs are a normal consequence of usual software quality and not hardware issues or problems that I can fix.

@cwebber are NaNs equal for this canonicalization or do they not occur in your domain? (these casts seem to preserve their bit patterns, but a NaN is not equal to itself, so it's not lossy unless using floating point equality checks)

@cwebber @aradinfinity This looks very similar to what http://www.alwaysinnovating.com/products/smartbook.htm and some other cancelled or otherwise non-available devices could have been. (I imagine all my software issues with a phone and external keyboard or a laptop and external screen would get much more complex with such use cases.)

What I most like in the #SmackJeeves "renewal" is that most comics that I used to read there are also hosted elsewhere, so I can change the links in my feed reader to sites that still provide RSS. I hope future technologies will automate finding which platform hosts a work.

@selfisekai you can also make a CSS file not reload due to a magic cookie like ETag or an URL parameter; your post reminds me how often my attempts at debugging "succeed" without understanding how the whole system works

@selfisekai What was the original error? I almost never had to reinstall after package configuration errors, while I have seen some especially on Unstable.

Migrated my laptop running Debian from btrfs to ext4. What I did wrong: (0) continuing to run btrfs after it got uncorrectable errors and remount read-only several times (it might work better in some more usual use cases than mine); (1) not rebuilding initramfs after the change, making mount fail; (2) passing a different label to cryptsetup luksOpen than in /etc/crypttab, causing an infinite loop in initramfs.

@forteller Grafana can be used for that and it's what I'd choose unless already having a metrics database with an integrated UI

I dislike exactly two things about the #Firefox reader mode: (1) it replaces the URL with an about:reader one, but keeps displaying the original one in the address bar (so I don't see why bookmarks are separate with and without it) (2) it's too common in Web design to make pages unreadable when it's disabled.

https://xkcd.com/1205/ might be a good rule when working, but not when using a computer at home: automating the task might be more fun, I might learn more, while spending much more time than just doing the task. (I'm not sure this would convince me to learn the Emacs style of regular expressions; I always have to guess how many backslashes to add to turn a trivial POSIX ERE into an Emacs regexp.)

Received the #fsfbulletin. I was surprised seeing the envelope with proper diacritics in my name and address when Polish companies still use replacement character boxes in invoices.

Seen a "disable adblock or register" banner (disabling adblock doesn't hide it); it should have been "practice #adblock element hiding or disable #PrivacyBadger and get tracked more".

@cwebber @natecull It looks easier with writers only, not parsers, in client libraries. Cases like PostgreSQL getting a new feature, or implementing such a library in a new programming language, are very good arguments for a consistent protocol syntax. (Or being able to hack tools for e.g. generating a smaller query reproducing a bug.)

@cwebber @natecull Isn't this solved by libraries like SQLAlchemy Core? Any text protocol might have attacks like SQL injections; I have seen code not escaping ' in JSON, there are attacks on apps accepting newlines in HTTP headers. Or is the problem in libraries (not protocols and languages) treating SQL/HTML/... as text?

Set up a Mastodon instance; finally had an opportunity to use Docker with a real-world application.