Right click, view source?

Firefox is really in a sad state. The politics, ideology is getting worse and the tech is still bad. Kind of crazy that they waste everyones battery and cpu https://b.43z.one/2025-02-12/

Am I the one who is crazy and nitpicking here?

Firefox works fine. It's my daily browser on a fairly old Macbook. No issues whatsoever, other than some websites that weren't tested on it.

People seem to think I'm crazy in my other comments when I said I often run into websites that don't test Firefox and thus do not work properly.

there are some observations of mine:

- when the browser is not in the visible workspace, chromium's cpu usage actually goes to 0 whilst firefox's usually oscilates around 5%

- every new window costs 30mb of ram for chromium and 100mb for firefox

- this one is weird, but when lightly using firefox while overall ram usage spilled well into swap, often after some time terminal emulator process will get swapped-out as if firefox eventually touched memory used by every tab sitting in the background

That's a fascinating bit of sleuthing! 1.5 watts is a substantial amount of energy when you're in the 4 to 5 watt overall usage range. You're not crazy and nitpicking.

Did I understand that correctly that I will be able to get a certificate for an IP?

Yep! Should be available to the general public (as long as you're using an ACME client that can be configured to request a specific profile) later this week.

I recently wanted to find out which company crawls the deepest. The openAI bot was the most thorough one, it followed 405 links [1].

[1] https://deep.43z.one

Such a cool idea. I have https://chess.maxmcd.com/ and (before blocking most of them) many bots played thousands of moves deep. I remember bingbot was very active.

Wondering if after this comment, you'll get more visit from those bots.

Can someone explain why we are still using a umask of 022 in ubuntu and debian?

Would it really be so hard to make that switch to a more privacy focused umask?

Because in June 2005 the simple response to the Debian bug filed in September 2004 was to comment the global setting out of /etc/login.defs rather than change it to 0027. And after some back and forth there's now the explanation in /etc/login.defs that you can read today (q.v.).

* https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=269583

Doesn't feel like much of an explanation to me.

  # UMASK is the default umask value for pam_umask and is used by
  # useradd and newusers to set the mode of the new home directories.
  # 022 is the "historical" value in Debian for UMASK
  # 027, or even 077, could be considered better for privacy
  # There is no One True Answer here : each sysadmin must make up his/her
  # mind.

That comment was in Bullseye. In Trixie's /etc/login.defs the comment is gone.

With Trixie, PAM's "User Private Groups" are by default enabled and default umask thus is 002 instead of 022.

(Personally, I'm irritated by the rather silent way this invasive change got introduced -- it is mentioned in /usr/share/doc/libpam-modules/NEWS.Debian.gz together with instructions to restore the old behavior.)

Ah the classic "There is no One True Answer so it's ok to default to a bad answer".

And also, some tools still break when using the non-default umask.

Yes, yes, we all run Postgres in containers, but if you don't, and you upgrade to a new Postgres major version, gladly using the Debian scripts that make it all more comfortable, while using umask 027, you will enjoy your day. Though I don't remember if those upgrade-scripts where from Debian proper or from Postgres.

Since that experience I always wondered what other tools may have such bugs lurking around.

Is this really a big deal on effectively single user systems with in-person hardware? On the other hand, why is this such a hard decision for Debian to make?

The span one is the important one though as most of the modern web does not use <button> anymore.

I used to use vimium (and tried similar extensions) but it always scared me how big the codebase was for most of the popular extensions. In the end I came up with a tiny extension for just the things I need https://github.com/h43z/jkscroll

This works on links and buttons in chrome. In firefox just on links. In both it doesn't work for other html elements.

I was going to disagree with you, but… huh. It doesn’t work on buttons. Thought it did. Selection is set, but focus isn’t. Further workaround: Shift+Tab.

This feels like it may be a bug, but at the very least it’s not a recent regression—I tested Firefox 44 and it shows the same behaviour. (44.0 is the oldest version I can run now, apparently. I tried 4.0 first, the first version with linux-x86_64 builds, which I have run successfully in the past, I think even in the last year, but now all versions before 44.0 are crashing on startup.)

The links yes, but what about buttons or other html elements?

Buttons work too.

The issue is with any JavaScript driven on click events tags. Some sites even have their <a> tags not responding to keyboards events, because they have a hash href, and a JavaScript handler to redirect.

The web is beautiful place.

I don't see it working on firefox for buttons but on chrome. Both don't work on other html elements (div,span,..).

Do I understand this correctly that the server here is only needed to make sure the secret it's only read once?

Yes, you are understanding it correctly, the server (Cloudflare Worker + Durable Object + KV) in Stasher is only needed to enforce the burn-after-read behaviour