macOS Command Line | Hacker News Reader

macOS Command Line

git.herrbischoff.com

385 pointsanimal_spirits114 comments

https://archive.ph/PXzE7 since it's currently receiving the HN hug of death.

Furthermore, looking at almost 26k GitHub stars for this repository, if only about 14% of the people who this resource is useful for, sponsored me with a one time amount of a single Euro, I could order a new machine that probably lasted me another eight years. It's not something I'd expect, nor think I should. It's entirely my issue.

It's unseemly that people have to beg like this. I would be happy to pay a small annual Github subscription if it were distributed back to repo maintainers. Same thing with all the 'buy me a coffee' tags and now the 'buy me a slice of pizza' ones. Instead of funding useful bodies of work with meaningful micro-grants you have projects competing to provide people with better cups to rattle.

That's pretty much what GitHub Sponsors is.

Except it's targeted to whomever I feel like sponsoring. Maybe an additional way could be distribution of a general sponsorship fund based on the number of stars or another measure of merit/popularity? I've sponsored individual projects on GitHub before, but would prefer this way.

Yes, that's my point. It's another market, and while markets have a lot of utility they also tend towards bimodal distributions which disproportionately reward a very successful and visible participants while disproportionately exploiting a much larger group. A pure market system has higher entropy than one which features some redistribution.

https://arxiv.org/pdf/cond-mat/0001432.pdf

If something like this is going to be based on number of stars, imagine how this will turn the current "give us a star on github" to.

a game just like amazon reviews. if there's a system that can generate money that is able to be gamed, it will be. star ratings on git hub will become next to useless as it will no longer be used for anything other than monetization.

I think in France we had some interesting propositions for press funding based on this idea.

Everyone paid a tax each year and gets token that they can attribute to the media they want. The fund are redistributed according to the tokens.

I don't really like the idea of an automated measurement of merit / popularity. Let users decide and choose few projects they want to sponsor each year. Maybe show the current number of sponsor so big projects don't get all the votes.

And Marx and Engles philosophy :D

This kind of knowledge has historically been shared for free. We are not entitled to it, as much as the author is not entitled to compensation.

At the risk of being downvoted to oblivion, I’ll say this: asking for donations is great, complaining that people are “leeching” your CC-licensed content, not so much. If you don’t have the bandwidth to maintain a project, just stop. Humanity will be fine.

This move to monetize everything is a huge step backwards (or forwards into the abyss…).

It’s mind-blowing that the 8th wonder of humanity, the open-source software in general, which runs the entire world, reached other planets and has even taken off from them (the helicopter on Mars ran Linux), has been a giant cooperation and donated for free by its owners. It far outdid the previous wonders and yet, it’s immaterial.

It’s a stretch to call Linux donated imo. A ton of work is paid and the companies only “donate” because the license requires them too

The GPL doesn't require anyone to contribute back to the upstream project, only contribute forward to downstream users, who may or may not care about source code and even if they do, may or may not bother to publish the code they receive, and even if they do, they/others may or may not try to contribute that code back upstream. It is only a culture of contributing back upstream and more importantly the costs of forking vs merging that leads to upstream contributions. Of course the GPL and copyleft are a big part of forming that culture, but not all of it.

It's not only culture, it's also mutually beneficial, the cost of maintaining a forked version of the kernel is not worth it. Meanwhile working on a patch and publishing it upstream is relatively easy "fire-and-forget", the patch is there, other people will use it and contribute back.

Companies still have a ton of private code that never makes its way into the open source world.

I mentioned the forking cost too, my main point though is the license doesn't require people to upstream their stuff.

This move to monetize everything is a huge step backwards

The guy can't afford to replace his 10 year old laptop, meantime rents and profits are at an all-time high. Seems like there's a few segments missing in the virtuous circle.

Or it’s hyperbole and there are many reasons the guy has a 10 year old laptop. People donating a euro for a list of commands is unlikely when 10 euros gives them a billion hours a month on Netflix.

The marginal value of these commands is really, really low when you start comparing it to all the stuff that people could spend.

Most everyone has high rents and old laptops.

Author here. I don't expect people to pay me for anything. My main motivation was (and still is) to just give back, since I profit from free information and software created by others as well. I fully expect this temporary increase in traffic to net exactly what it has netted before: almost nothing. And you know what? I'm fine with that. Donations are not the reason this little project exists.

What I am is struggling to make ends meet for some time. A new laptop is near the bottom of my priorities right now. Money will run out around the beginning of February. It makes me sad to read cynical comments like this. That's all I have to say.

Thanks for putting the list together, the commands have really helped me.

I hope your fortune turns and you soon find a job.

The offputting thing to me was the faulty logic of 14% of people starred donated. That’s a phenomenal donation rate that’s unprecedented. So it’s like prefacing my work with “if money rained on my front yard” or something along those lines.

And I don’t like the discussion that guilts people who don’t donate. I donate to projects on GitHub. Not much and not to yours. But it’s something. But I contribute in other ways and would like to discourage nagging people (not by you as your preamble comment was pretty benign) who don’t donate.

Meanwhile, Dwarf Fortress made a new release, is selling it for $30, and everyone is rushing just to give the developer $30 because they want to support them.

> have all those ppl been playing free DF for 25 years and just waiting for an opportunity to pay $30?

> Yes

* times 50 pages

https://steamcommunity.com/app/975370/discussions/0/37093075...

Something about selling tools to developers just brings out the worst of "bah, I could build that in a weekend so it's not worth anything so I'm not going to pay for it"-ness.

It is well over 100 pages now.

That is amazing!

The flipside would be if the author put it on steam, had barely any sales, and then started complaining about how people leeched on his work for 25 years.

I think the author doesn't get GitHub stars. They don't mean "I use this and like it", they are simply bookmarks. A way to find something interesting again. I don't think I actively use anything I starred. Or am I the odd one out?

I use it the same way - "oh this is interesting and I may want to use this later" presses star button

presses star button, never read it again

This is what bookmarking in your browser is for. Not here for whines about convenience, if the maintainer can see it then you're sending a signal.

People place way too much importance on github stars.

I will continue "bookmarking with stars" and not using half the things I star :)

This is a misunderstanding but it is far from unusual. For one, the JavaScript package ecosystem is incredibly vain and is absolutely obsessed with these sorts of deeply flawed measurements of popularity.

The university system is a mechanism to fund public goods. The problem is the professors are judged according to their production of novel knowledge, and maintaining useful software does not produce novel knowledge. We should endow chairs of "open-source professors", who are judged according to their production/maintenance of useful software, in order to re-use the infrastructure of the university system to fund this other type of public good.

If you were similarly curious about the mention of "GitHub stars" in the text hosted on their own cgit, this appears to be the back-story: https://github.com/herrbischoff/awesome-macos-command-line

I've found that the trick to getting $3000 for a wiki page is to go get a job and write a wiki page while I'm there

(Or come on board as a consultant, either approach can work wonders)

How do people discover the `defaults` keys?

I occasionally stumble upon such secret troves of knowledge, but I'd love to find out how these are discovered.

Is there, like, a way to intercept all `defaults`/settings reads to discover what keys are checked?

I think what you’re looking for is “defaults read”

I suspect that they are asking about finding settings which get used, but aren't documented. Like, how to find hidden settings that programs will use.

In my experience with plist settings is that they are rarely sparse. But sure finding hidden settings is a tricky issue.

Unfortunately, that only seems to print those defaults that are set, not all defaults that are possible.

For example, the NSToolbarTitleViewRolloverDelay setting, while valid per the article, isn't listed in the output of `defaults read`

Per sibling comments, you run strings on binaries, find out informally from Apple engineers, or dig through settings in the System Preferences and then use `defaults` to look for changes in output.

Defaults is a tool which can read and write system and other application settings. By default it works on the system settings. One can either pass a bundle id or a path to a plist file to read and write other settings.

You can see a huge list of settings by typing „defaults read“

Theoretically you could attach to a process with LLDB, and use objc swizzling to intercept calls to NSUserDefaults. However to do this, I think you would need to disable SIP. Even then i'm not sure this would work.

In practice, people probably just read user defaults and try to reverse engineer what they do.

defaults logs what it is doing if you turn on debug logs for that subsystem.

AIUI some of them are communicated informally by Apple engineers to individuals, some of them are found by using `strings` on the binary and looking at reasonable-seeming keys, any key that actually has a UI to set it is typically found just by setting it in the UI and seeing what changed, some of this stuff may have had UI's in the past that were removed while still leaving the defaults key, some of this stuff may just write the keys into the preferences files automatically and so you can find them just by looking at what's already there, etc.

Skimming this list right now a lot of the stuff I'm seeing definitely has UI settings, so a lot of these are just useful for e.g. automating the setup of a new user.

Same question. I also want to read a value, but can't seem to get it right. For example, I'd like to know what the current value of

   CGFontRenderingFontSmoothingDisabled

is. I tried

   defaults read CGFontRenderingFontSmoothingDisabled

but got:

   Domain CGFontRenderingFontSmoothingDisabled does not exist

thankfully defaults read spits out all of the vars, so you can do

    defaults read | grep CGFontRenderingFontSmoothingDisabled

Thank you

Yes! It doesn't work 100% of the time, but what I've been doing for my setup script is:

`defaults read > old` -> change setting in GUI -> `defaults read > new` -> `diff old new`

If nothing is there, you're out of luck, most likely. (I think I've had luck using -currentHost or maybe sudo sometimes though...) If there is something, `less new` and search for the string, then find the path to the key. Do a test read with `defaults read <domain> <item>` to see if you got the path right.

You can also copy plist files (cp -r /Library/Preferences, or wherever) and then diff those. A useful alias to convert to xml:

alias plist='plutil -convert xml1 -o /dev/stdout'

For mapping UI preferences to corresponding plist keys, https://github.com/catilac/plistwatch will monitor and output real-time changes. May be easier than diffing snapshots of `defaults read`.

This doesn't help with "secret" settings that aren't exposed through the UI, but can be handy for creating setup scripts.

An exhausting list of mac os specific settings, that can only be changed from terminal. Quite the effort!

Thanks for sharing this! The “Remove proxy icon hover delay” setting is the first thing I’ll try. Many times I use the proxy icon in Preview to move a pdf file… the delay introduced in Big Sur is extremely annoying.

I believe since Monterey you can reenable proxy icons, in the Accessibility panel of System Preferences. No idea about Ventura, couldn't update.

I'm having a lot of trouble with getting fonts to look clear and unfuzzy on my M1 Pro with a new LG 38WN95C-W 38" 21:9 (3840 x 1600) monitor I just bought. It's driving me nuts. I've tried different picture settings on the monitor, with and without HDR, tried different resolutions, tried every possible value of `defaults -currentHost write -g AppleFontSmoothing -int ${n}` but no dice.

Anyone have any ideas what I can do to resolve?

Machine details below:

   Software:

    System Software Overview:

      System Version: macOS 12.6 (21G115)
      Kernel Version: Darwin 21.6.0
      Boot Volume: Macintosh HD
      Boot Mode: Normal
      Secure Virtual Memory: Enabled
      System Integrity Protection: Enabled
      Time since boot: 4:14

Hardware:

    Hardware Overview:

      Model Name: MacBook Pro
      Model Identifier: MacBookPro18,1
      Chip: Apple M1 Pro
      Total Number of Cores: 10 (8 performance and 2 efficiency)
      Memory: 32 GB
      System Firmware Version: 7459.141.1
      OS Loader Version: 7459.141.1

Are you sure you're running the monitor at its native resolution and not scaled? And do they provide a display profile that you can load?

Edit: This post would indicate that it's a bug in the interaction between the Mac and the monitor and is not fixed: https://forums.macrumors.com/threads/fuzzy-text-with-lg-38wn...

I am pretty sure I'm running it at the native resolution and not scaled, yes. At least, I'm pretty sure:

   > system_profiler SPDisplaysDataType | grep Resolution
             Resolution: 3840 x 1600 (Ultra-wide 4K)

And ah dang, thanks for the link. Someone there mentioned that getting a higher PPI monitor could help and might be the only way to get HiDPI modes. I did notice when I used switchResX for a bit that the HiDPI mode was greyed out and unavailable. Guess I'm going to have to return this monitor :(

BetterDisplay can get you supersampling on <4K monitors. I run a 4K monitor as my main display that did this out of the box with macOS, but I have side 1440p monitors that didn't. This fixed it: https://github.com/waydabber/BetterDisplay

Try this https://www.reddit.com/r/ultrawidemasterrace/comments/sve8jf...

Switch to Linux, which has actual subpixel AA, unlike macOS and increasingly Windows.

I mainline NixOS, but sometimes switching to Linux is... just not an option unfortunately. It's sad tbh that macOS dropped subpixel AA

I replied in a child comment too, but to put this on the original. BetterDisplay can fix this: https://github.com/waydabber/BetterDisplay#fully-scalable-hi...

Thanks so much. I'll give that a try. I tried switchResX and easyres without much luck. Giving this one a try soon

I followed those steps in the link you provided and played around with other resolutions and configurations as well, but still no dice. Everything is so fuzzy, I'm beginning to get a headache :(

* Is the monitor running in RGB or YPbPr? With M1 Macs some monitors are getting forced into YPbPr mode and fonts look worse because of it. (https://gist.github.com/GetVladimir/c89a26df1806001543bef4c8...)

* Remember to log out and log back in after you set font smoothing to 0. For me, this tweak helped a lot with clarity but I initially didn't think it was working because I didn't know you had to log out.

* In general I've found Macs just suck at DPI scaling. If I'm not mistaken, your monitor is around 109 PPI, which isn't really that high compared to the internal screen (~250ppi), and firmly in the "non-Retina" UI elements zone. Check out these blog posts (https://bjango.com/articles/macexternaldisplays/, part 2: https://bjango.com/articles/macexternaldisplays2/) for more info on how this stuff works.

Imagine my surprise when I found out my 27" 4K screens that look incredibly sharp on Windows at 163ppi are actually in the "bad zone" for Macs, and look bad at native resolution. As a workaround I run them scaled @ 3360 x 1890 and then set my browser to permanent 90% zoom level. Kind of embarrassing that Windows has completely lapped Apple in terms of good sharp UI at flexible scaling levels. I mean you can set Windows DPI scaling down the single % point, it's great.

As someone who is dealing with this problem (27" 1440p monitor), I read that macOS renders the whole screen as a whole, instead of having a separate path for text. And there is a threshold PPI before it uses 2x assets. I had to use either a blurred rendering for bigger UI elements or the sharp, but small, representation at the native resolution. BetterDisplay[0] helped by tricking the OS to use the 2x assets at a smaller resolution. Still not as clear as my MBA's screen.

[0]: https://github.com/waydabber/BetterDisplay

> Is the monitor running in RGB or YPbPr?

I am not sure...? How can I find out? Thanks for that gist, I'll look into it after checking whether BetterDisplay resolves my issues as another commenter pointed me to.

> Remember to log out and log back in after you set font smoothing to 0

Hehe yeah. I have actually been operating under the assumption that I needed a full restart, which I have done several times :P (though this isn't so bad since the machine starts pretty dang quick).

> In general I've found Macs just suck at DPI scaling

I'm learning this :( It's very frustrating and disappointing.

Followed along with the video to Force RGB Color on M1 Mac, but seems I already had PixelEncoding set to 0 in com.apple.windowserver.displays.plist

Those posts were helpful, thanks. From their table that lists the various displays[0], it seems like two smaller displays are more likely to work well vs a larger display — 25 - 29 inch rather than the 38" I have.

Are there truly no displays in the 30"+ category that are good for retina?

[0]:https://bjango.com/images/articles/macexternaldisplays/displ...

EDIT: I'm seriously wondering whether I ought to just get the Studio Display. It's $1,599 (or $1,899) which is not that much more than I spent on the LG (~$1,160)

You could also check out the ultrawides from Dell/LG that have 2160 vertical res. They're $1500-$2000 from what I recall and will help with sharper text The PPI will be closer to what a standard 16:9 4k offers, and though won't be as good as the studio display @ 5k, debatable whether that matters at desktop viewing distance.

Personally I won't be upgrading from my 2x 27" 4k displays until

* I can get 2x 27" 4k with 144hz+ at a reasonable price

* I can get an ultrawide that matches the PPI and is just effectively 2x screen space

* 5k+ gets cheaper/commonplace

Thanks a lot. This makes sense. It's really too bad because ultrawide screen was kind of a game changer to me in terms of productivity. I'm not even sure why it's served me so much better than using dual monitors. It does seem I will need to consider going to back to dual screens.

Thanks. So I went out and picked up a Studio Display. It's a big downgrade in terms of size obviously, but text is extremely sharp and colors are great. Beyond that, it's more tidy on my desk with the great stand and I can get rid of my external speakers because the built-in speakers are pretty dang good. We'll see how this feels.

Thanks again for the replies.

Nice, glad to hear you're enjoying it, maybe I'll look into one myself.

I don't know if I have the same display (it is also 21:9, in any case) but one issue I found is when I'm in some "gaming mode", I get ugly font rendering. It looks great if I switch to "user mode", but I don't know exactly what is different.

This is all configured in the display's own settings, not in the OS.

How is the monitor connected? USB-C/DisplayPort or HDMI?

I have tried the Thunderbolt cable as well as the HDMI cable (which seems to limit the resolution and refresh rate) that the monitor came with. The Macbook Pro doesn't have a DisplayPort input. The monitor did come with a DP cable. Should I try a DP to USB-C converter? Is there any reason that would work better than TB?

Upgrading to Ventura and resetting resolved the issue for me. I had previously been using the beta version due to this issue. The better display app did not resolve the issue for me, and there are other known issues with external monitors on Ventura that you may encounter.

Oh wow, I was going to ask about this. Thank you!

What do you mean by "reset" exactly? You mean any and all display changes I've made?

I wiped it clean during installation so no previous settings.

The HDPI icon pop ups automatically on ventura in display settings.

I would also change the cable and try a different one if that doesn't work. Sometimes the DP cable end up causing issues with monitor recognition.

Got it, thank you. Unfortunately I'm not in a position to do a clean install at this moment without some interruption to my work, but I'll try that later if I need to.

Upgraded to Ventura, but didn't help display issues unfortunately. I've got a thunderbolt cable that came with the monitor. I've tried that as well as two different HDMI cables. My MBP doesn't have a DP input. Is it worth trying DP with a DP to USB-C adapter? I may have one laying around somewhere.

> Is it worth trying DP with a DP to USB-C adapter?

No. I meant if you have a DP to USB type C cable, you may switch to see if issue resolved.

Try plugging it in and out a few times (yes, it worked for me when it went back to not recognizing).

I have the same mac but a different 4K monitor. Sadly, I don't know any other solution.

Macbook display and scaling handling on external monitors is one of the worst I have experienced by a long margin so know you are not alone.

Great resource. You can find many more examples of other people’s efforts with this GitHub search: https://github.com/search?q=defaults+write+com.apple&type=co...

> There's really only one thing I'd like to note here: man pages. Man pages. Man pages. Okay, three things. But this one thing seemed so important, I had to mention it multiple times. If you're not doing it already, you should get into the habit of consulting man pages before searching anywhere else. Unix-style man pages are an excellent source of documentation. There's even a man page for the man command itself:

YES YES YES. It's how I learned my way around Unix back in SunOS days. I literally started with "man intro" and worked my way out from there. Sun's documentation was _excellent_.

Sadly, Linux and even BSD man pages are more hit or miss compared to Sun's documentation. But they are still a great place to start.

Many macOS utilities and facilities have man pages, including things like launchd plists (man launchd.plist).

I’d love a tweak to change the speed of the Spaces animation when you switch desktop, it’s so frustrating slow that I just don’t use it. I’ve searched a few times and not found anything but thought I’d mention it here in case anyone does know of one!

If you tick "Reduce motion" in Accessiblity, it fades in an out instead of scrolling, a lot faster! I wish there was a way to disable transitions altogether like in earlier OSX versions...

Ugh, preach. I haven't found anything, and I did search around

Yabai (https://github.com/koekeishiya/yabai) with SIP disabled is the only way to do that nowadays.

I personally gave up on Spaces completely because of this. Now I just have everything on a single space, and move through apps instantly using my rcmd app (https://lowtechguys.com/rcmd)

TotalSpaces can do it, you can switch instantly.

If you use Ctrl+N (or whatever the shortcuts for 'Switch to Desktop N' are set to in the settings) instead of swiping or using Ctrl+Left/Right it's much faster, and input is re-enabled for the desktop you're switching to basically instantly

This resource is great, but the saddest part is removing it from GitHub. I think it will hugely hit its visibility and create more of low quality copycats. For example, I rarely use my bookmarks or even check them out, but I constantly use my categorized GitHub stars to read the content I need. I suspect I’m not the only one doing it like this.

What I would really like to know is how to mute an application. I have sounds disabled for outlook notifications in System settings and do not disturb on. Yet, I hear chimes/bells from outlook every time I receive an email.

Anybody else experience similar things?

Don’t have Outlook, but the setting is supposedly in Outlook’s Personal Settings>Notifications and Sounds>Sounds, uncheck all.

For Outlook specifically, look in Outlook's settings under "Notifications and Sound".

Thanks, I will check it. But how can it override the system settings, that is an oversight I think…

https://rogueamoeba.com/soundsource/

Now, I haven't tested your issue specifically, because I do not use Outlook.

But Rogue Amoeba is one of the best remaining Mac app developers (also: Panic), and they made an app that provides app-based sound control. Worth checking out if you can't solve it with Outlook's settings.

I got SUPER excited about the NSScrollViewRubberbanding option, because I really hate that rubberbanding animation; sadly it looks like it hasn't worked for several years now.

here's the magic line in .*rc that you didn't know you need:

`set -o vi`

It applies vi rules to your command line.

I love vi bindings but I've found that if you're using oh-my-zsh this will break a bunch of stuff like the command history. If anyone knows of a workaround for that I'd love to hear about it.

Or its more sane friend, control-x+control-e to launch $EDITOR with a file in $TMPDIR containing the current command-line, and then execute it when $EDITOR exits. BTW, I don't mean "$EDITOR" as a placeholder for whatever editor is your favorite, I quite literally mean the shell variable "EDITOR". Strictly speaking, $VISUAL has precedence but I don't think any system sets that one by default, whereas many do set EDITOR. I guess the one drawback to this comment in reply to yours is that the very bottom of that fallback tree is "emacs", but what vimmer doesn't already have EDITOR=$(command -v vi)? :-D

The gory details: https://www.gnu.org/software/bash/manual/html_node/Miscellan...

Does anyone know how to login to iCloud through the terminal?

Glaring omission (no pun intended): Spotlight including mls

Holy cow, TIL 'say "my mac can speak"'!

It's a handy thing to have in your toolbox https://xkcd.com/530

There was a time when you’d get a printed manual with your computer that taught you these things. Now we can’t even get Apple to provide digital documentation for a lot of these hidden features in macOS. I prefer the old way.

Users of my monitor control app (https://lunar.fyi/) feel overwhelmed with the information I provide on the website and in little help hints placed in the app near each setting.

If an app has this effect, I can't even imagine what an OS manual would feel like in these days.

I wish I could provide a proper manual but I can barely keep up with updating the app website with the rate of change in technologies. A major OS upgrade would need an army of writers to update manual on time.

Not that I wouldn't love it. It would be tremendously useful to have a PDF where I can Cmd-F obscure tidbits which I often need in building MacOS apps.

I just think it's impractical given the complexity of software in this age.

I don't even think it's the complexity, necessarily. More of the fundamental instability of it all. Why exactly does Mac or Windows need to be releasing entirely new OSes every 2 years, except to maintain a new product for the demands of a corporate behemoth to chirp around. People have PERL scripts from 40 years ago that still execute fine, but I doubt anyone has a PowerShell file from the early 2010s that maintains behavior.

I can't comment on Windows, but with regards to apple, they really don't. Sure they strut around like they have but really, they've tweaked some user apps like Mail and Safari, and made bug fixes under the covers. Every now and then they do something like muck with background scheduling, but mostly... it's just bug fixes and new APIs that don't invalidate old ones.

The only problem I've had is dylib files relentlessly moving forward in versions and old ones not being available anymore, but i don't _think_ that's really an apple problem. That and the M1 architecture transition has moved some folders around for reasons I don't understand.

My ancient scripts still work just fine.

It’s not impractical for Apple and Microsoft literally does it. I’m not saying MSDN is perfect, but it’s a very expansive and thorough documentation of (almost) everything in Windows.

The only difference is that people like to make excuses for Apple.

love Cascadia Code as terminal and coding font. author might want to take a look.

folks should probably avoid linking their own personal git forge (or others, if this is the case) to HN and just use GitHub to share their code. or deploy the content to a static site host.

Why even have any sites in other places? Why not just put everything on Github, then it's all in one place within reach at us-east-2

I'm thinking in terms of etiquette and experience. It's not really polite to throw a bunch of unexpected traffic at a little lowly app server

Since I host a little lowly app server with a public listener on 80, I’m cool with whatever traffic is thrown. And that’s good etiquette.

The web was built for lots of distributed servers hosting and users hitting them. Only using “big servers” is bad for the world.

Postel's law, I suppose, is the right way to look at it.

Agreed, even as a SRE myself it is just easier to link to a GitHib repo. GH, with all of its warts, is still likely better at handling a traffic spike than you.

Link elsewhere in this thread[0] shows that this user removed this repo from GitHub deliberately, so I doubt they're the account who shared it.

[0]: https://news.ycombinator.com/item?id=33897373

People are free to host however they like. They can't control who posts links on popular aggregators.

Reply on news.ycombinator.com