HNHacker News
TopNewBestAskShowJobs

grumbel

1,782 karma · joined April 16, 2015

submissionscomments
grumbel··on GNU Stow needs a co-maintainer
Nix and Guix do pretty much that, at the core they are just some symlinks and environment variables. But unlike stow, they build all the rest of the package management and build infrastructure as well.

The big issue with stow is that you have to manually get the packages into DESTDIR and quite a lot of packages don't directly support that or do it in their own non-standard way, so there is far too much manual work involved getting anything installed from source. With Nix and Guix you can write a package definition once and than everybody can reuse that instead of reinventing the wheel.

grumbel··on ActivityPub/Nostr/At-Bluesky Compared
What if I want something that isn't an echo-chamber?
grumbel··on ActivityPub/Nostr/At-Bluesky Compared
With ActivityPub you are still putting everything in one basket. All the federation is completely optional and identities are not transferable. But unlike big corps, that generally care very little about what you do, a whole lot of ActivityPub instances are heavily politically motivated and love to block or defederate.
grumbel··on Autodafe: Tools for freeing your project from the clammy grip of autotools
With autotools you get DESTDIR, --prefix and a bunch of other things that work the same across all projects. With Makefiles everybody is rolling their own thing and you never know what to expect, or frequently have to implement those things yourself.

That said, autotools, with its multiple layers of file generation, makes debugging rather annoying. And it's generally much easier to fix a broken Makefile than figuring out why autotools goes wrong.

grumbel··on Nix – A One Pager
Nix allows you to install packages without naming or version conflicts. It allows you to make source repositories act as first-class packages. It can be installed on any Linux distribution and MacOS. Changing and customizing packages is trivial. All the building and installing happens outside of the source tree, so you don't clutter up your $HOME. It's reproducible. It's simple. And it's build with the needs of Free Software in mind (unlike snap or flatpak which seem mostly concerned with packaging proprietary software). Also Nix is 20 years old at this point.

Simply put, with Nix I can do:

   nix run github:project/name
to run a program. Not many other package managers are capable of that.
grumbel··on AI-generated sad girl with piano performs the text of the MIT License
It's suno.ai (has a free trial), works much the same as image generation, you give it a description and it writes a song in a couple of seconds. Lyrics can be customized:

https://app.suno.ai/song/41fde9b6-a722-4c39-92dc-8a8296c018c...

grumbel··on From xz to ibus: more questionable tarballs
> the author of the project controls what is put in the Git repository itself

The author doesn't control the Git repository. The Git repository is a shared data structure, not just bits on a server. Neither the author nor anybody else can just change the Git repository arbitrarily. They can only change it in ways allowed by Git or the alarm bells will be going off. A tarball has no such restrictions.

grumbel··on From xz to ibus: more questionable tarballs
> But both come from the same authors.

The Git repository comes with a history containing each and every commit, along with checksums and sometimes cryptographic signatures for it all. The tarball is just a bunch of random files.

If somebody wants to rewrite the Git history, that will get instantly noticed by anybody that had an older clone of it. If somebody adds a hack to it, it will be at the top of the history and easy to notice as well.

If somebody adds something to a tarball, there is nothing to tell you that something changed, you'll have to audit the whole thing from top to bottom (or compare it to the Git repository it came from).

Also the author isn't in control here, the tarball can get changed after it was uploaded to the server and there would be nothing in the tarball itself to tell you.

People are using PGP signatures and checksums to secure tarballs some more, but that's always an extra step and doesn't happen automatically. And even that still doesn't give you the history.

grumbel··on Nanofont3x4: Smallest readable 3x4 font with lowercase (2015)
DSLinux[1] used a 4x6 font for the terminal, I found that surprisingly usable.

[1] https://www.dslinux.org/dslinux-cpuinfo.jpg

grumbel··on Radicle: Open-Source, Peer-to-Peer, GitHub Alternative
Git evades censorship just fine, since it is properly decentralized and doesn't care about where you got the repository from. Plain HTTP transport however does not and most Git repositories are referred to by HTTP URL.

If you simply host Git on IPFS you have it properly decentralized without the limits of HTTP. IPNS (DNS of IPFS), which you need to point people to the latest version of your repository, however wasn't working all that reliably last time I tried.

grumbel··on Marilyn vos Savant and the Monty Hall Problem (2015)
> Suppose you were new to the contest and were allowed to put money down on the choice at the point that there were two doors left.

That would be a completely different game. The goat doesn't get reshuffled. The doors that remain are not random and most importantly, you know which door you picked in the first round. It doesn't become a 50:50 chance just because there are two doors, as the goat isn't distributed over those two doors, but across all three.

Imagine 1000 doors. You pick one. In round two you are asked if you want to stay with that one door or pick all the other 999 doors at once. What do you do?

grumbel··on Marilyn vos Savant and the Monty Hall Problem (2015)
The remaining doors are not random. The one you picked still has a 1/3 chance of winning, that doesn't change with opening another door. However switching doors is essentially the same as picking all the other doors at once, since the door with a goat is not opened, thus the 2/3.

Imagine the thing with 100 doors and imagine the host isn't opening any doors, but just telling you what's behind them and it becomes pretty obvious.

grumbel··on Just bought a new PC, it won't let me use it unless I create a Microsoft account
When you view a Twitter post without being logged in it will only show you the post itself, not the rest of thread. You only get to see that when logged in. And of course the Twitter.com start page itself gives you no way to even find any posts to begin with.
grumbel··on RFC 9512: YAML Media Type
The handling of multi-line strings is needlessly complex and can easily end up in white-space getting eaten unintentionally.
grumbel··on Coding in Vision Pro
The original Oculus Rift had 1080x1200, VisionPro has 3660x3200, field of view is similar. In virtual monitor terms that means a Rift can display a 640x480 monitor and the VisionPro a 1920x1080 one. Still quite far away from Retina, but should be pretty usable, especially since it allows making the virtual screens bigger than a real monitor can be.
grumbel··on How to copy a file between devices? (2022)
SSH requires to have shell access on the receiving end, that renders it unusable in most situations. Rsync can in theory do anonymous file offerings, but it's not a shell one-liner to get it up and running either.
grumbel··on How to copy a file between devices? (2022)
There is also:

    python -m http.server
Which tends to be my go-to solution for file transfer between devices, it shares the current directory over HTTP. Not perfect, since HTTP makes downloading directories rather cumbersome, but at least every device that can open a website can download stuff.
grumbel··on “Wherever you get your podcasts” is a radical statement
Meanwhile if put a podcast URL into my Webbrowser I just get a bunch of raw XML. I never liked that RSS, while having some nice properties, exists as this awkward thing next to the Web, but isn't really part of the Web.

What RSS is doing feels like something that should be done with just a few extra HTML tags, not a completely new format. But the idea of HTML as markup language feels kind of lost these days.

grumbel··on Vision 1993
Oculus DK1 sold 56,334

Oculus DK2 sold 118,930

Rift and Vive together ~1 million

PSVR1 ~6 million

Quest123 20+ million

grumbel··on Apple Vision Pro review
The major problem VR has isn't the games, but all the boring and basic stuff, like using 2D apps in VR or running multiple VR apps at the same time.

The discontinued WMR Portal, essentially the Window's desktop in VR, was so far the only software that tried to be a full workspace in VR. But even that was missing a lot of important features and Microsoft gave up on it years ago and never made it accessible to non-Microsoft headsets. It's currently scheduled for removal from Windows.

VisionPro seems very similar to WMRPortal so far, with a few key improvements like allowing apps do add 3D objects into a shared space.

grumbel··on Apple Vision Pro review
> 4K monitor

It's more like 1080p monitor. The virtual monitor only covers a small part of the VisionPro's display. You can compensate a bit for a lack of resolution by making the virtual screen bigger or by leaning in, but none of that gives you a 4k display.

To really take proper advantage of the VR environment you really need the ability to pull out apps into their own windows, as than you can move lesser used apps into your peripheral vision and leave only the important stuff right in front of you. You also miss out on the verticality that VR offers when you are stuck with a virtual 16:9 screen.

grumbel··on In loving memory of square checkbox
My "favorite" toggle button is the one where they add on/off labels[1] to it, but only one, and the slider hides the label. So the label actually tells you the state, not the state it would be when you move the slider over the label. Or at least that's how I think how they work, it breaks my brain a little whenever I look at them.

[1] https://itsfoss.com/content/images/size/w1000/wordpress/2022...

grumbel··on Launch HN: Diversion (YC S22) – Cloud-Native Git Alternative
> doesn't destroy the reflog or runs the GC server-side.

Git doesn't give you access to the server side reflog either. So it's of not much use if you don't control the server.

As for losing data with Git, the easiest way to accomplish that is with data that hasn't been committed yet, a simple `git checkout` or `git reset --hard` can wipe out all your changes and even reflog won't keep record of that.

grumbel··on Why isn't Bluesky a peer-to-peer network?
> but that also means they're hard to remember

Nobody remembers domain names either, that's what we have search engines for. The point of public keys is that they give the user something they own completely themselves and that their peers can verify. All the human-readable stuff can be implemented on top of that, e.g. if you wanna use domain names, put the public key into your DNS record. But you aren't forced to use domain names, you can send your keys around by mail if you prefer or put them in your Instgram bio.

The most important part here is that this works towards a "Web as data structure". Everything that is based around domain names will fall apart in a few months or years when those domain names expire, links go dead constantly. Everything that is build around crypto can last forever, as the storage location becomes irrelevant.

We already saw the benefits of this approach back with Usenet, where large parts of it could be reconstructed from random backup tapes, since everything was held together by globally unique message ids. Usenet didn't use crypto for this, but just depended on users behaving nicely, so that approach wouldn't work today, but it could be replicated with public key crypto in a very similar and more robust way.

grumbel··on Why isn't Bluesky a peer-to-peer network?
Domains are a rare and expensive resource, that also happens to get hijacked by spammers all the time when somebody forgets to pay their bills. It's not even possible to buy a domain permanently. Meanwhile if you stick to just a hostname portion of a domain, you at the whim of whoever owns the domain, which is exactly the problem that has plagued the Internet for all its existence.

We should just stick to public keys and forget about the domain name system if we ever want to actually improve the situations instead of just reinventing Twitter and Facebook by another name. It's endlessly frustrating how long public keys have been around and how little use they still get. Finger crossed that Nostr gets some traction, since that so far sounds like the most sensible approach towards a social network that is under the controls of the users themselves and nobody else.

grumbel··on Filecoin Foundation Successfully Deploys IPFS in Space
The person who added it, is not tracked. A file can be added by anybody at anytime and anywhere and it will get the same content address. It's up to the person doing the redistribution to ensure it's legal, but IPFS gives them no tools to do so.
grumbel··on Filecoin Foundation Successfully Deploys IPFS in Space
Everything in your local cache is redistributed, meaning every single website you view through IPFS ends up getting potentially redistributed. Can you guarantee that each and every item you view on the Web allows redistribution?
grumbel··on Filecoin Foundation Successfully Deploys IPFS in Space
Has it? I have never seen anybody using IPFS in the wild, even projects for which it should be well suited (archive.org, Linux package distribution, Git, Lemmy, Imgur, CivitAI), don't use it. Worse yet, IPFS still provides no real way to deal with private or local data, which drastically limits its use.

I love the idea about data being addressable by hash, but I don't feel IPFS has actually delivered anything meaningful in that area yet.

And with ipfs-search.com shutdown, there is not even any way left to explore what is actually on the network now.

And technical issues aside, there is also the legal problem that IPFS conflicts with copyright. Redistributing anything is illegal by default, unless somebody gives you permission, IPFS provides no means to track that permission. You can't attach a GPL or an author to a hash.

grumbel··on AI Is Already Killing Books
But you could feed the text through the LLM, generate a summary and a list of keywords, characters and locations and search through that the regular way. And as far as I can tell, that doesn't really exist yet.

IMDb has a keyword feature[1], but that's still completely human-curated and quite useless due to that, as it just doesn't have the necessary depths and completeness to discover interesting new content.

[1] https://www.imdb.com/search/keyword/

grumbel··on AI Is Already Killing Books
It's weird, I always expected AI categorization and search to arrive long before generative content. But despite a lot of research going into that direction, it never really arrived in the consumer space. Search today feels not fundamentally different than search 20 years ago and if I want to know about a movie I still use IMDb like it's the 90s.

The already enormous mountain of content out there keeps growing, fuel by AI now, yet ways to explore that mountain keep staying the same or even diminishing, as the clutter in the results just keeps increasing.

← PreviousPage 8 of 22Next →