RunWithScissors() (2009) | Hacker News Reader

RunWithScissors() (2009)

android.googlesource.com

43 pointstentacleuno22 comments

A less amusing name, but similar effect – `dangerouslySetInnerHTML` from React.

https://legacy.reactjs.org/docs/dom-elements.html#dangerousl...

Or ‘accursedUnutterablePerformIO’ from Haskell https://hackage.haskell.org/package/bytestring-0.11.4.0/docs...

The documentation is itself fairly funny, for those who don’t care to click ahead:

> This "function" has a superficial similarity to ‘unsafePerformIO’ but it is in fact a malevolent agent of chaos. It unpicks the seams of reality (and the IO monad) so that the normal rules no longer apply. It lulls you into thinking it is reasonable, but when you are not looking it stabs you in the back and aliases all of your mutable buffers. The carcass of many a seasoned Haskell programmer lie strewn at its feet.

> Witness the trail of destruction:

    https://github.com/haskell/bytestring/commit/71c4b438c675aa360c79d79acc9a491e7bbc26e7
    https://github.com/haskell/bytestring/commit/210c656390ae617d9ee3b8bcff5c88dd17cef8da
    https://ghc.haskell.org/trac/ghc/ticket/3486
    https://ghc.haskell.org/trac/ghc/ticket/3487
    https://ghc.haskell.org/trac/ghc/ticket/7270
    https://gitlab.haskell.org/ghc/ghc/-/issues/22204

> Do not talk about "safe"! You do not know what is safe!

> Yield not to its blasphemous call! Flee traveller! Flee or you will be corrupted and devoured!

Originally, that function was just called "inlinePerformIO", with the relatively tame warning of:

> Just like unsafePerformIO, but we inline it. Big performance gains as it exposes lots of things to further inlining. Very unsafe. In particular, you should do no memory allocation inside an inlinePerformIO block. On Hugs this is just unsafePerformIO

https://hackage.haskell.org/package/bytestring-0.9/docs/Data...

At some point, this function was moved to the newly created "Deprecated and unmentionable" group of functions with the comment

> Deprecated: If you think you know what you are doing, use unsafePerformIO. If you are sure you know what you are doing, use unsafeDupablePerformIO. If you enjoy sharing an address space with a malevolent agent of chaos, try accursedUnutterablePerformIO.

At this point, the two depreciated and unmentionable functions were actually just aliases of each other, with InlinePerformIO being depreciated, and accursedUnutterablePerformIO being merely unmentionable.

https://hackage.haskell.org/package/bytestring-0.10.10.1/doc...

It wasn't until version 0.11.0.0 that the friendly sounding name was finally removed.

This function name amuses me because it implies there is a “safelySetInnerHTML” method

Also shouts out to the UNSAFE_component methods, you’re a real hipster if you used those before the prefix

Not exactly. If you are putting something inside the tag in React, you are implitly setting inner HTML the React way, managed by the virtual DOM.

This method (actually not even a method but an attribute that probably calls a method if set) "bypasses" React to set the inner HTML of the element, hence "dangerous".

(But I get what you mean of course, the reply was not to you but anyone who isn't familiar with React)

You’re right it’s not even a method lol it’s a custom dom attribute. I had no idea

Fun memories: we had some bug in some application, and we reviewed the possible fixes in a meeting. "The fix is simple and relatively safe, the only issue is that we need to use a function called runWithScissors".

IMHO this is the effect expected by the people who named the function, and the reason why the name is very well chosen.

Yes, it's funny but that's not the point. The important thing is that the name clearly conveys the message that you really shouldn't be calling it, as explained in the comments above the function definition.

The "funny but there's a real reason behind" aspect is a bit like https://xkcd.com/radiation/ : "It's for general education only. If you're basing radiation safety procedures on an internet PNG image and things go wrong, you have no one to blame but yourself." The sentence is funny but has a real purpose: protect readers and author.

Obligatory: Face your fears, from Crazy Ex-girlfriend.

https://www.youtube.com/watch?v=brzZQBSVMX0

Wow the quality of code comments is over the top!

> If we ever do make it part of the API, we might want to rename it to something less funny like runUnsafe().

I wonder, when you write that, why you wouldn't chose to rename it right away rather than using a funny name? Is it a strategy to avoid people using it? (runUnsafeAvoidUsing) or a strategy to make sure it doesn't get into the API? (wouldn't a flag or runUnsafeDontMakeItIntoTheApibe better )?

The comment could've been written after the function was named?

This in some ways evokes for me the worst of Google engineering and culture. It’s documented in a nice way and with a gentle tone, but it’s fundamentally a method on top of broken behavior and a ton of tech debt. Yes, it’s not uncommon for folks engage like this in software; but once you’ve worked with individuals that think deeply about these kinds of issues and regularly, you start to see there’s another way.

As an xoogler, I don't agree at all. Google has a ton of problems, but a reluctance to think deeply about and fix tech debt isn't one of them. In fact, I think the eagerness to fix fundamentally broken abstractions is a big reason Google has so much code churn and is so eager to irresponsibly deprecate things that work!

What? Can you explain what broken behavior you’re talking about?

Some things in concurrent programming can result in deadlocks. That doesn’t mean that you round off all the corners and take away all the dangerous tools until you’re left with a kiddie playground as a programming environment. No, instead it means that the more experienced programmers on your team keep a sharp eye on how concurrency / multithreading is used in your program.

When I needed to use this function, it was because the Android API to run things asynchronously using Handler.post()/remove() seems to be incomplete. You want to remove() a runnable, but maybe that runnable is executing in parallel exactly when you want to remove it. Depending on what you application does, this will cause various issues. The simplest way to fix issues in this case is to synchronize the threads using RunWithScissors. In short: the API is simple and easy to use, but it allows for strange cornercases, and you need a hack like RunWithScissors. That function is a hack because for various other reasons related to the rest of the API and the way it is implemented, it may block forever in some other cornercase. So maybe we can say this is broken behavior and an bad/incomplete API.

That makes sense.

Not making excuses for the API—but this does seem like a difficult design space to work in.

I've never been a googler, but looking through this file, I would be really pleased if my company wrote that. I left Java altogether because "Java" became synonymous with "nightmare code" but I could live in the world of this source file.

This reminds me of Running with Scissors, the game company who made Postal games. Maybe the author was a fan of those games?

I'm more thinking of Weird Al Yankovic's album "Running with Scissors" (1999).

I was very surprised to see that there was an inconsistency between the constructors as well as the fields placed after all the methods and private and public methods mixed and matched in order, trying to see what mLooper was about.

I would not have approved this code.

stupid question: are the limitations testable in an assert to reduce the risks?

Reply on news.ycombinator.com