Expected result of 'Dining Philosophers' example is not possible by following the instructions #30471
Comments
steveklabnik
added a commit
to steveklabnik/rust
that referenced
this issue
Dec 28, 2015
Some history: While getting Rust to 1.0, it was a struggle to keep the book in a working state. I had always wanted a certain kind of TOC, but couldn't quite get it there. At the 11th hour, I wrote up "Rust inside other langauges" and "Dining Philosophers" in an attempt to get the book in the direction I wanted to go. They were fine, but not my best work. I wanted to further expand this section, but it's just never going to end up happening. We're doing the second draft of the book now, and these sections are basically gone already. Here's the issues with these two sections, and removing them just fixes it all: // Philosophers There was always controversy over which ones were chosen, and why. This is kind of a perpetual bikeshed, but it comes up every once in a while. The implementation was originally supposed to show off channels, but never did, due to time constraints. Months later, I still haven't re-written it to use them. People get different results and assume that means they're wrong, rather than the non-determinism inherent in concurrency. Platform differences aggrivate this, as does the exact amount of sleeping and printing. // Rust Inside Other Languages This section is wonderful, and shows off a strength of Rust. However, it's not clear what qualifies a language to be in this section. And I'm not sure how tracking a ton of other languages is gonna work, into the future; we can't test _anything_ in this section, so it's prone to bitrot. By removing this section, and making the Guessing Game an initial tutorial, we will move this version of the book closer to the future version, and just eliminate all of these questions. In addition, this also solves the 'split-brained'-ness of having two paths, which has endlessly confused people in the past. I'm sad to see these sections go, but I think it's for the best. Fixes rust-lang#30471 Fixes rust-lang#30163 Fixes rust-lang#30162 Fixes rust-lang#25488 Fixes rust-lang#30345 Fixes rust-lang#29590 Fixes rust-lang#28713 Fixes rust-lang#28915 And probably others. This lengthy list alone is enough to show that these should have been removed. RIP.
Manishearth
added a commit
to Manishearth/rust
that referenced
this issue
Dec 29, 2015
…ankro Some history: While getting Rust to 1.0, it was a struggle to keep the book in a working state. I had always wanted a certain kind of TOC, but couldn't quite get it there. At the 11th hour, I wrote up "Rust inside other langauges" and "Dining Philosophers" in an attempt to get the book in the direction I wanted to go. They were fine, but not my best work. I wanted to further expand this section, but it's just never going to end up happening. We're doing the second draft of the book now, and these sections are basically gone already. Here's the issues with these two sections, and removing them just fixes it all: // Philosophers There was always controversy over which ones were chosen, and why. This is kind of a perpetual bikeshed, but it comes up every once in a while. The implementation was originally supposed to show off channels, but never did, due to time constraints. Months later, I still haven't re-written it to use them. People get different results and assume that means they're wrong, rather than the non-determinism inherent in concurrency. Platform differences aggrivate this, as does the exact amount of sleeping and printing. // Rust Inside Other Languages This section is wonderful, and shows off a strength of Rust. However, it's not clear what qualifies a language to be in this section. And I'm not sure how tracking a ton of other languages is gonna work, into the future; we can't test _anything_ in this section, so it's prone to bitrot. By removing this section, and making the Guessing Game an initial tutorial, we will move this version of the book closer to the future version, and just eliminate all of these questions. In addition, this also solves the 'split-brained'-ness of having two paths, which has endlessly confused people in the past. I'm sad to see these sections go, but I think it's for the best. Fixes rust-lang#30471 Fixes rust-lang#30163 Fixes rust-lang#30162 Fixes rust-lang#25488 Fixes rust-lang#30345 Fixes rust-lang#29590 Fixes rust-lang#28713 Fixes rust-lang#28915 And probably others. This lengthy list alone is enough to show that these should have been removed. RIP.
steveklabnik
added a commit
to steveklabnik/rust
that referenced
this issue
Dec 29, 2015
Some history: While getting Rust to 1.0, it was a struggle to keep the book in a working state. I had always wanted a certain kind of TOC, but couldn't quite get it there. At the 11th hour, I wrote up "Rust inside other langauges" and "Dining Philosophers" in an attempt to get the book in the direction I wanted to go. They were fine, but not my best work. I wanted to further expand this section, but it's just never going to end up happening. We're doing the second draft of the book now, and these sections are basically gone already. Here's the issues with these two sections, and removing them just fixes it all: // Philosophers There was always controversy over which ones were chosen, and why. This is kind of a perpetual bikeshed, but it comes up every once in a while. The implementation was originally supposed to show off channels, but never did, due to time constraints. Months later, I still haven't re-written it to use them. People get different results and assume that means they're wrong, rather than the non-determinism inherent in concurrency. Platform differences aggrivate this, as does the exact amount of sleeping and printing. // Rust Inside Other Languages This section is wonderful, and shows off a strength of Rust. However, it's not clear what qualifies a language to be in this section. And I'm not sure how tracking a ton of other languages is gonna work, into the future; we can't test _anything_ in this section, so it's prone to bitrot. By removing this section, and making the Guessing Game an initial tutorial, we will move this version of the book closer to the future version, and just eliminate all of these questions. In addition, this also solves the 'split-brained'-ness of having two paths, which has endlessly confused people in the past. I'm sad to see these sections go, but I think it's for the best. Fixes rust-lang#30471 Fixes rust-lang#30163 Fixes rust-lang#30162 Fixes rust-lang#25488 Fixes rust-lang#30345 Fixes rust-lang#28713 Fixes rust-lang#28915 And probably others. This lengthy list alone is enough to show that these should have been removed. RIP.
bors
added a commit
that referenced
this issue
Jan 5, 2016
Some history: While getting Rust to 1.0, it was a struggle to keep the book in a working state. I had always wanted a certain kind of TOC, but couldn't quite get it there. At the 11th hour, I wrote up "Rust inside other langauges" and "Dining Philosophers" in an attempt to get the book in the direction I wanted to go. They were fine, but not my best work. I wanted to further expand this section, but it's just never going to end up happening. We're doing the second draft of the book now, and these sections are basically gone already. Here's the issues with these two sections, and removing them just fixes it all: // Philosophers There was always controversy over which ones were chosen, and why. This is kind of a perpetual bikeshed, but it comes up every once in a while. The implementation was originally supposed to show off channels, but never did, due to time constraints. Months later, I still haven't re-written it to use them. People get different results and assume that means they're wrong, rather than the non-determinism inherent in concurrency. Platform differences aggrivate this, as does the exact amount of sleeping and printing. // Rust Inside Other Languages This section is wonderful, and shows off a strength of Rust. However, it's not clear what qualifies a language to be in this section. And I'm not sure how tracking a ton of other languages is gonna work, into the future; we can't test _anything_ in this section, so it's prone to bitrot. By removing this section, and making the Guessing Game an initial tutorial, we will move this version of the book closer to the future version, and just eliminate all of these questions. In addition, this also solves the 'split-brained'-ness of having two paths, which has endlessly confused people in the past. I'm sad to see these sections go, but I think it's for the best. Fixes #30471 Fixes #30163 Fixes #30162 Fixes #25488 Fixes #30345 Fixes #29590 Fixes #28713 Fixes #28915 And probably others. This lengthy list alone is enough to show that these should have been removed. RIP.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I see that the delay between picking up the forks was added later and I think the end of the instructions hasn't been updated to fit.
The text at the end says "With this, our program works! Only two philosophers can eat at any one time, and so you’ll get some output like this:" with a trace showing 2 philosophers eating at a time.
What we actually get is:
With the 150ms delay between picking up forks, when the philosophers go to pick up their second forks, only fork 4 is available - so Emma Goldman can eat, when she finishes Karl Marx can use fork 3.
Without the 150ms delay the behaviour matches the description, with 2 eating at a time.
The text was updated successfully, but these errors were encountered: