Add that other blog post #132

faho · 2024-12-23T18:27:44Z

Should be in a reasonable state.

I've targeted it for the 25th of December because I had to pick something, but I'd be fine whenever.

I don't know if it's smarter to wait until after Christmas, or into the new year?

site/_posts/2024-12-25-rustport.md

krobelus · 2024-12-23T19:53:40Z

site/_posts/2024-12-25-rustport.md

+Truth be told, we did not quite expect that to be as popular as it was.
+It was written as a bit of an in-joke for the fish developers first, and not really as a press release to be shared far and wide.
+We didn't post it anywhere, but it was commented far and wide.


This sounds like humblebrag and doesn't seem super honest given that "attract (new) contributors" is the primary goal.

Attracting them with the code, sure. But with the PR? Did you expect that sort of reaction?

(plus the confusion at some of the tone)

And I think it's fair to be a bit proud, if you don't let it go to your head.

Like, yeah, it was pretty cool to see how many were interested in that?

And we did, as far as I know, not post it anywhere ourselves.

Yeah I'm also fine with this as written.

Bragging is appropriate in a technical blog post, that's not my point.

zanchey · 2024-12-24T03:29:52Z

@faho, I'm happy to review but it may not be until 26th/27th if you're happy to hold until then. Did you want me to try and get the word count down a bit, or is it mostly for tone/accuracy?

faho · 2024-12-24T07:37:02Z

Did you want me to try and get the word count down a bit, or is it mostly for tone/accuracy?

We're at about 4400 words in the main text plus a little under 400 in the footnotes. That seems fine and I don't see a lot to remove anymore.

I think the priorities are:

Is anything wrong, misleading or confusing? Any points we made that don't come across well?
Is anything missing - did we entirely forget to talk about a kind of thing, not "are there more examples" (unless there's a really good example)
Do we want links to commits or code examples? I want to go easy on examples because they can distract from the point
Wording
Typos
More jokes?
A header image of Ferris and Shelly eating spaghetti

ridiculousfish

I like the structure and flow of this, and the lighthearted casual tone. The jokes are great ("note to editor").

I would like to cut down on the "pain with C++" section so as not to come across as overly negative.

site/_posts/2024-12-25-rustport.md

ridiculousfish · 2024-12-26T23:47:09Z

site/_posts/2024-12-25-rustport.md

+
+We've experienced some pain with C++. In short:
+
+- tools


I suggest getting this list down, maybe even to three items, and keep it short. I don't want this post to be a dump-fest on C++. The three in my mind are "community, packaging (i.e. tooling), and thread-safety" but I'm not married to that.

I condensed it to

tools and compiler/platform differences

ergonomics and (thread) safety

community

ridiculousfish · 2024-12-27T00:13:33Z

site/_posts/2024-12-25-rustport.md

+experience with a roughly C++-shaped language should help.
+
+## Why are we doing this again?
+


This section is necessarily critical of both C++ and the current state of fish. For that reason I think we should try to make it significantly shorter. No need to dwell on the past.

I don't think it is too critical of the current state of fish (we managed to make it as good as it is despite these flaws!), and I think we do need to make our arguments against C++ clear.

That being said, I removed a paragraph that was essentially restating "compilers are hard to upgrade, we need to stick to old standards"

site/_posts/2024-12-25-rustport.md

ridiculousfish · 2024-12-27T02:23:18Z

site/_posts/2024-12-25-rustport.md

+Here's a dirty secret: fish's script execution is currently entirely serial - you can't background functions,
+and you can't even run builtins in parallel. This code:
+
+```fish
+for i in 1 2 3 4 5
+    sleep 1
+end | while read -l num
+    break
+end
+```
+
+takes 5 seconds, because the `while read` loop can't even run before the `for` loop completes.
+
+POSIX shells use subshells to get around this, but subshells are a leaky abstraction that can bite you in the behind when you least expect it.
+For instance you can't set variables from inside a pipe like that (except on some shells, but only in the last part of the pipe, maybe, if you have enabled the correct option).
+We would like to avoid that, and so the heavy hand of forking off a process isn't appealing.
+
+This involves a lot of careful handling of shared state, and C++ famously does not help - thread safety is your responsibility as the programmer.


I think it's OK to elaborate somewhat, but this "currently entirely serial" might make readers believe that fish runs all commands one at a time. The distinction between builtin, function, and external command isn't that widely known. I think it's OK to drop the example here, and instead discuss the benefits that multithreaded execution will bring. That is, focus on future capabilities rather than current limitations.

Here's my attempt, partially cribbed from my first post. Take or leave:

Here’s a dirty secret: while external commands run in parallel, fish’s execution of internal commands (builtins and functions) is currently serial. Lifting this limitation will enable features like asynchronous prompts or non-blocking completions.

POSIX shells use subshells to get around this, but subshells are a leaky abstraction that can bite you in the behind when you least expect it. For instance you can't set variables from inside a pipe (except on some shells, but only in the last part of the pipe, maybe, if you have enabled the correct option).

We prototyped true multithreaded execution in C++, but it just didn't work out. For example, it was too easy to accidentally share objects across threads, with only post-hoc tools like Thread Sanitizer to prevent it.

(this is going to bring out the "git gud" squad, but whatever, I'm taking it)

site/_posts/2024-12-25-rustport.md

ridiculousfish · 2024-12-27T19:03:53Z

site/_posts/2024-12-25-rustport.md

+
+## The story of the port
+
+We had decided we were gonna do a "Fish [Of Theseus](https://en.wikipedia.org/wiki/Ship_of_Theseus)" port - we would move over, component by component, until no C++ was left.


I really like the time-based narrative flow of this section, well done

site/_posts/2024-12-25-rustport.md

ridiculousfish

LVGTM!

faho · 2024-12-27T22:15:36Z

Okay, I'd like to publish this tomorrow, i.e. in 12-18 hours.

ccoVeille

please accept my humble contribution to your article

docs/blog/rustport/index.html

zanchey · 2024-12-29T13:01:12Z

This is great, well done!

tanriol · 2024-12-29T16:12:12Z

docs/blog/rustport/index.html

+
+<p>But listing targets in our code is also fundamentally duplicating work that the libc crate (in our case) has already done. If you want to call libc::X, which is only defined on systems A, B and C, you need to put in that check for A, B and C yourself and if libc adds system D you need to add it as well. Instead of doing that, we are using our own <a href="https://github.com/mqudsi/rsconf">rsconf</a> crate to do compile-time feature detection in build.rs.</p>
+
+<p>Most of this would be solved if Rust had some form of saying “compile this if that function exists” - <code class="language-plaintext highlighter-rouge">#[cfg(has_fn = "fstatat")]</code>. With that, the libc crate could do whatever checks it wants and fish would just follow what it did, and we could remove a lot of the use for rsconf. It would not really help support older distributions that lack some features, tho. That could be solved by something like the <a href="https://github.com/rust-lang/rfcs/pull/3036">min_target_API_version</a> cfg.</p>


Sounds like the correct issue here would be rust-lang/rust#64797 (#[cfg(accessible(::path::to::thing))])

Ah, that's probably the one I was looking for.

I knew something here had been proposed but I forgot what it was exactly.

Add riir post

2db332b

faho requested review from zanchey, mqudsi, ridiculousfish and krobelus December 23, 2024 18:27

krobelus reviewed Dec 23, 2024

View reviewed changes

link ship of theseus

c7eed18

faho added 4 commits December 24, 2024 15:42

tighten wording

9ec87ef

Some more about autocxx/copying/performance during the port

8a3c801

Add note on widecharwidth

412fc76

Add rust-terminfo link

04fb5aa

ridiculousfish reviewed Dec 27, 2024

View reviewed changes

faho force-pushed the 🦀🦀🦀 branch from b6b114c to 2d5663b Compare December 27, 2024 20:45

faho added 3 commits December 27, 2024 22:05

Address feedback

89d0e08

Add some more to parallel execution

230bfcd

Remove duplicate "upgrading compilers is hard"

a47512c

ridiculousfish self-requested a review December 27, 2024 21:46

ridiculousfish approved these changes Dec 27, 2024

View reviewed changes

faho added 3 commits December 27, 2024 22:49

Titles

0f366ba

Adjust date

93a0cba

slight rewording

e0100a5

faho force-pushed the 🦀🦀🦀 branch from 2d5663b to e0100a5 Compare December 27, 2024 21:57

Build

21d8035

ccoVeille reviewed Dec 28, 2024

View reviewed changes

faho added 2 commits December 28, 2024 10:43

Typos

d6c70a6

builtin builtin builtin builtin builtin builtin

8be87f3

faho merged commit 21222b4 into fish-shell:master Dec 28, 2024

tanriol reviewed Dec 29, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add that other blog post #132

Add that other blog post #132

faho commented Dec 23, 2024

krobelus Dec 23, 2024

faho Dec 23, 2024

faho Dec 23, 2024

ridiculousfish Dec 27, 2024

krobelus Dec 28, 2024

zanchey commented Dec 24, 2024

faho commented Dec 24, 2024 •

edited

Loading

ridiculousfish left a comment

ridiculousfish Dec 26, 2024

faho Dec 27, 2024

ridiculousfish Dec 27, 2024

faho Dec 27, 2024

ridiculousfish Dec 27, 2024

faho Dec 27, 2024

ridiculousfish Dec 27, 2024

ridiculousfish left a comment

faho commented Dec 27, 2024

ccoVeille left a comment

zanchey commented Dec 29, 2024

tanriol Dec 29, 2024

faho Dec 29, 2024

		experience with a roughly C++-shaped language should help.

		## Why are we doing this again?


		## The story of the port

		We had decided we were gonna do a "Fish [Of Theseus](https://en.wikipedia.org/wiki/Ship_of_Theseus)" port - we would move over, component by component, until no C++ was left.


		<p>But listing targets in our code is also fundamentally duplicating work that the libc crate (in our case) has already done. If you want to call libc::X, which is only defined on systems A, B and C, you need to put in that check for A, B and C yourself and if libc adds system D you need to add it as well. Instead of doing that, we are using our own <a href="https://github.com/mqudsi/rsconf">rsconf</a> crate to do compile-time feature detection in build.rs.</p>

		<p>Most of this would be solved if Rust had some form of saying “compile this if that function exists” - <code class="language-plaintext highlighter-rouge">#[cfg(has_fn = "fstatat")]</code>. With that, the libc crate could do whatever checks it wants and fish would just follow what it did, and we could remove a lot of the use for rsconf. It would not really help support older distributions that lack some features, tho. That could be solved by something like the <a href="https://github.com/rust-lang/rfcs/pull/3036">min_target_API_version</a> cfg.</p>

Add that other blog post #132

Add that other blog post #132

Conversation

faho commented Dec 23, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

zanchey commented Dec 24, 2024

faho commented Dec 24, 2024 • edited Loading

ridiculousfish left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ridiculousfish left a comment

Choose a reason for hiding this comment

faho commented Dec 27, 2024

ccoVeille left a comment

Choose a reason for hiding this comment

zanchey commented Dec 29, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

faho commented Dec 24, 2024 •

edited

Loading