Comments on: Peterson’s n-Process Protocol http://www.alper.net/programming/petersons-n-process-protocol/ Eyup Alper Yoney Sun, 28 Aug 2011 09:14:25 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Raheel http://www.alper.net/programming/petersons-n-process-protocol/comment-page-1/#comment-5 Raheel Wed, 25 May 2011 15:34:22 +0000 http://www.alper.net/?p=171#comment-5 Hi good work done by Mr.Alper... Hi
good work done by Mr.Alper…

]]> By: Alper http://www.alper.net/programming/petersons-n-process-protocol/comment-page-1/#comment-4 Alper Sat, 27 Mar 2010 17:57:01 +0000 http://www.alper.net/?p=171#comment-4 Hi Peter, Thank you so much for your comment. You are absolutely right on both; - The turn[j] should be turn[i], there was a typo. - For the second problem, I had missed the case FALSE and the first stage 0. I had been thinking to assign -1 as a not-interested but forgot somehow. Anyway updating the loops is much clear and readable. I have corrected the code as you suggested. Best Regards, Alper Hi Peter,
Thank you so much for your comment.
You are absolutely right on both;
- The turn[j] should be turn[i], there was a typo.
- For the second problem, I had missed the case FALSE and the first stage 0. I had been thinking to assign -1 as a not-interested but forgot somehow. Anyway updating the loops is much clear and readable.
I have corrected the code as you suggested.

Best Regards,
Alper

]]> By: Peter http://www.alper.net/programming/petersons-n-process-protocol/comment-page-1/#comment-3 Peter Fri, 26 Mar 2010 14:23:34 +0000 http://www.alper.net/?p=171#comment-3 Sorry, I wanted to write, that the i-cycle should run from 1 to N-1 and a stages should be 0 to N-1. Sorry, I wanted to write, that the i-cycle should run from 1 to N-1 and a stages should be 0 to N-1.

]]> By: Peter http://www.alper.net/programming/petersons-n-process-protocol/comment-page-1/#comment-2 Peter Fri, 26 Mar 2010 14:21:15 +0000 http://www.alper.net/?p=171#comment-2 Hi, in my humble opinion, there are two significant errors in the code of the n-process solution. Firstly, in the while cycle, there should be turn[i] instead of turn[j]. That's because turn is a value associated to a stage, not to a process. Secondly, I am not very sure if a solution with N-1 stages works. When a process leaves a critical region, it's stage is being set to FALSE, which is 0. But, when a process tries to enter a critical region, it's stage is set to 0, as a first try. This implies that processes which aren't interested to enter a critical region can block processes that are interested (in the first condition of the while cycle). In my opinion, there should be stages 0 to N, and the i-cycle should run from 1 to N. This would avoid such not so good situations. Sincerely, Peter. Hi,

in my humble opinion, there are two significant errors in the code of the n-process solution.

Firstly, in the while cycle, there should be turn[i] instead of turn[j]. That’s because turn is a value associated to a stage, not to a process.

Secondly, I am not very sure if a solution with N-1 stages works. When a process leaves a critical region, it’s stage is being set to FALSE, which is 0. But, when a process tries to enter a critical region, it’s stage is set to 0, as a first try. This implies that processes which aren’t interested to enter a critical region can block processes that are interested (in the first condition of the while cycle). In my opinion, there should be stages 0 to N, and the i-cycle should run from 1 to N. This would avoid such not so good situations.

Sincerely,
Peter.

]]>