synchronization

I have a collection of Mercurial repositories on a network share. To enable offline work, I want a local copy of this collection on my laptop, and an easy way to synchronize the two when I'm online. For this, I wrote a quick script that automatically synchronizes each local repository with the corresponding remote repository (push and pull), but it's missing a couple of desirable features: automatic cloning of new repositories from the local to the remote collection (and vice versa) the ability to organize (move/rename) a local repository and have the change being applied on the remote side as

I have two clients A and B. Both clients have identical, synced local data caches. If client A makes an offline edit to record X and then client B also offline edits record X and syncs with the server, when client A syncs with the server, the change that client B made is not reflected and no amount of bidirectional syncing makes the two clients correctly synced with the server. Other than this, my sync app works great for deletions, additions edits etc. It's only simultaneous offline edits that destroy integrity. I am using SQL 2005 (with the necessary SQL scripts for sync support applied) Is

Process A forks say 4 child processes. exec() is used to replace the code of the children. The children initialize and have to wait for the parent to create all of the 4 of them. Then the parent sends a sigusr1 to each child process in order for them to start processing. The parent waits for all the children to complete procesing. When a child finishes its work it sends a sigusr2 to the parent. When the parent receives all the sigusr2 signals it continues with execution, merging the calculations of the child processes. This is a university exercise, and it was stated in class that process A

If I have a synchronized block and somewhere inside that block an exception is thrown that is not caught within the synchronized block, would the lock be relinquished when the exception propagates out of it?(the synchronized block) synchronized( mutex ) { throw new Exception( "" ); } The lock is always released. From JLS §14.19 : "If execution of the Block completes normally, then the lock is unlocked and the synchronized statement completes normally. If execution of the Block completes abruptly for any reason, then the lock is unlocked and the synchronized statement then completes abruptly

I'm in the initial phase of designing an application that will have a backend implemented in C# that will provide data for other platforms using WCF web services hosted on IIS. One of the platforms will the the iPhone. Since it's a personal project, I want to use it to learn MongoDB. I already know that there are community developed drivers for MongoDB and C#, so I could handle the persistence on the server side using MongoDB. Without even knowing the replications models offered by MongoDB, I was thinking about some kind of simple synchronization model to keep data local if the iPhone is not

I need to sync text and audio on many web pages. While the audio is played the text should be highlighted phrase by phrase (not word by word or by character so sync is necessary only for the start of the phrase). I'd like not to use a flash only solution and I'd prefer to use something more HTML friendly. I thought to use a combination of two plugins: JPlayer for the audio Jquery highligheter for highlighting the text I could then write a jquery script to cycle on the various phrases. The first problem is that JPlayer has only percentage played events and not precise timeline events so I would