Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Trainwreck Design

j3s.sh

Post reply on HN
Print view View original

91–100 of 107 posts

Re: Trainwreck Design

#91

Post by hinkley »

nottorp wrote: post #76
topspin wrote: post #65

Earlier quoted context omitted.

> Why are you even looking at this with the naked eye? My first instinct was to answer your question with a question. I'll avoid that. In the real world of devops, whatever you might have in mind as not "naked eye," fails. Deeply isolated systems, embedded systems, development environments, headless systems, virtualized systems, etc., don't have working GUIs, complex input devices or other high fidelity tools. You ge…

Yeah, I keep forgetting HN needs explicit sarcasm tags.

People knew sarcasm doesn’t translate on the internet a decade before Poe’s Law was named.

Re: Trainwreck Design

#92

Post by hinkley »

topspin wrote: post #65
nottorp wrote: post #56

Earlier quoted context omitted.

But... you have to use systemd tools to parse the output these days. Why are you even looking at this with the naked eye?

> Why are you even looking at this with the naked eye? My first instinct was to answer your question with a question. I'll avoid that. In the real world of devops, whatever you might have in mind as not "naked eye," fails. Deeply isolated systems, embedded systems, development environments, headless systems, virtualized systems, etc., don't have working GUIs, complex input devices or other high fidelity tools. You ge…

Also the point of the article is that all of this stuff used to work just fine. Abandoning tools people used just fine for decades is a cop out wrt to the context of this conversation.

Re: Trainwreck Design

#93

Post by hinkley »

Xenoamorphous wrote: post #58
benrutter wrote: post #39

Earlier quoted context omitted.

I think it's an intentional aesthetic choice[0]. Hyperlinks would be convenient, but something about the raw text / ascii art vibe makes me happy everytime I read a blog post from j3s even if it doesn't have the conveniences of the modern web. [0] https://j3s.sh/about.html

Hyperlinks are not modern web though.

What?

Re: Trainwreck Design

#95

Post by SSLy »

larodi wrote: post #85
atoav wrote: post #34

Earlier quoted context omitted.

As a youngling I have to say, I always liked the idea of the shell as something where you can pipe stuff inbetween programs. It is a genius concept. The problem is that in the real world implementation of the idea you have something like bash, that sucks for text processing with all its footguns of unescaped strings, spaces, wildcard operators and such. Maybe it has to do with strings not really having been a first c…

And eventually you come to conclusion that you need powershell like pipes of objects. Btw someone tried to implement a shell in Python which feeds structures, not sure the name though… was in HN front a while ago

https://www.oilshell.org/cross-ref.html?tag=YSH#YSH https://www.nushell.sh/

Re: Trainwreck Design

#96

Post by kantapproves »

Mali- wrote: post #69
kantapproves wrote: post #63

Surprise, surprise. Let me be the contrarian: I dislike this attitude of dumbing everything down just because some people can't be bothered to learn how and why their tools work the way they do. They'll invoke "2024" like it means something. Not a great look IMO. The output of df and mount shows the truth as it is: raw and shitty. Don't dumb it down. Your version of usability might be my idea of a nightmare. I like t…

Why should the default of the tool be totally unintelligble useless (for 90% of single usages) rubbish? You can have your raw truth - strange way of putting it, its a tool, this is not a detective drama - it would be a more useful tool for the average use-case if this was a flag though (or some other option). We can't handle the truth!

It is not unintelligible, it is exceedingly simple. If you are not knowledgeable about your system at this level, then sure, it is quite chaotic. But that’s a different problem and why are you at this level anyway?

Also, a substantial amount of this problem is just systemd. Don’t shit on df, which looks fine on, say, BSD.

This is not for the “average user” IMO.

Re: Trainwreck Design

#97

Post by kantapproves »

creesch wrote: post #73
kantapproves wrote: post #63

Surprise, surprise. Let me be the contrarian: I dislike this attitude of dumbing everything down just because some people can't be bothered to learn how and why their tools work the way they do. They'll invoke "2024" like it means something. Not a great look IMO. The output of df and mount shows the truth as it is: raw and shitty. Don't dumb it down. Your version of usability might be my idea of a nightmare. I like t…

I feel like you missed the point of the article. As this > The output of df and mount shows the truth as it is: raw and shitty. Is pretty much the point they are making about Trainwreck Design. They are using their output of df as a talking point to illustrate the issue.

Maybe. But they explicitly mention: “this results in programs like df. or mount.”

If you want to call out design shit then just call it what it is: systemd. Leave df out of it, the poor fellow. Mount also has nothing to do with it.

Re: Trainwreck Design

#98

Post by kantapproves »

topspin wrote: post #83
kantapproves wrote: post #63

Surprise, surprise. Let me be the contrarian: I dislike this attitude of dumbing everything down just because some people can't be bothered to learn how and why their tools work the way they do. They'll invoke "2024" like it means something. Not a great look IMO. The output of df and mount shows the truth as it is: raw and shitty. Don't dumb it down. Your version of usability might be my idea of a nightmare. I like t…

> Don't dumb it down Where was that demanded? > It is exceedingly easy to filter the output of said tools. This is provably false. Consider the content of Linux file systems. There is no one tool to simultaneously observe conventional POSIX properties, ACLs, capability attributes, selinux properties, AppArmor references, etc., to which you might apply a filter. So you have to repeatedly create your own, half-baked, i…

“this results in programs like df. or mount.”

This implies both tools need to adjust to reality - aka systemd bs - to make it easier for the user. Also known as dumbing down. I concede it is not explicitly requested, but it’s not that much of a reach TBH.

It is provably true that the output of df and mount is exceedingly easy to filter. You don’t even need tools to do it.

I don’t care about AppArmor and other BS because I am not on Linux. Those weren’t part of the discussion anyway. It is about mount and df and the author explicitly called them out for failure to be “easy”.

Re: Trainwreck Design

#99

Post by kantapproves »

marginalia_nu wrote: post #75
kantapproves wrote: post #63

Surprise, surprise. Let me be the contrarian: I dislike this attitude of dumbing everything down just because some people can't be bothered to learn how and why their tools work the way they do. They'll invoke "2024" like it means something. Not a great look IMO. The output of df and mount shows the truth as it is: raw and shitty. Don't dumb it down. Your version of usability might be my idea of a nightmare. I like t…

The problem is the raw shitty truth. Not that it isn't hidden, but the mess that it shows. The underlying reason for this mess is that several popular tools, especially in the systemd and containerization-adjacent space, are built on a bunch of non-standardized userspace hacks (e.g. per-process tmpfs chroot jails) that pollute the mount table of the host system. That is the trainwreck.

I agree. The trainwreck is systemd and Linux in general, but not df and mount.

Re: Trainwreck Design

#100

Post by anal_reactor »

huem0n wrote: post #89
topspin wrote: post #36

This complaint about df/mount/etc. recurs with noticeable frequency. Another ugly spot is iproute2's ip link/addr. Between containers and VMs and their virtual interfaces, link bonding, virtual bridges, macvlan/tap, etc., the list of "interfaces" proliferates. Trying to find an actual physical interface is difficult, and interpreting what is there is a challenge. Improving this stuff isn't happening because no one is…

> It's just that no one is doing that. There's a new wave of Rust based tools. Things like sd instead of sed, and fd instead of find. I think they have an opportunity to fix these things. They're not built in right now, and probably won't be anytime soon. But maybe eventually they will be. Although, on a different note, I don't think they follow the hive or cathedral design that this author discusses so maybe they wi…

> Things like sd instead of sed, and fd instead of find.

I really wish we'd stop naming things `a`

Post reply on HN