Hello Nairalanders, I have been banging my head against the wall trying to understand concurrency. I have made some progress but I still cannot grab the concept fully. So I will appreciate some code samples illustrating the phenomena and some pointers as to how your language tackles the problem.

John

Concurrency in the context means being able to speed up your programs by running them on more than one CPU core, CPU, or computer system at the same time. Python, my favorite programming language, supports concurrency for IO bound programs by providing threading abstractions (threads, locks, queues, event objects, etc). However, the Python interpreter is encumbered by a global interpreter lock which prevents CPU-bound python programs from taking advantage of multiple cores or CPUs. As a result, Python now provides a multiprocessing module that provides thread-like abstractions in a multi-process context (processes, locks, queues, event objects, etc), allowing CPU-bound Python programs to enjoy more concurrency quite easily.

@op, take that and call me in the morning ^^.

@Poster, what you need to do is your own research. Even the wiki page on the subject explains quite a bit.

Seun: Concurrency in the context means being able to speed up your programs by running them on more than one CPU core, CPU, or computer system at the same time. Python, my favorite programming language, supports concurrency for IO bound programs by providing threading abstractions (threads, locks, queues, event objects, etc). However, the Python interpreter is encumbered by a global interpreter lock which prevents CPU-bound python programs from taking advantage of multiple cores or CPUs. As a result, Python now provides a multiprocessing module that provides thread-like abstractions in a multi-process context (processes, locks, queues, event objects, etc), allowing CPU-bound Python programs to enjoy more concurrency quite easily.

You are spot on, however the bolded is more about dealing with the consequence of every computers inability to handle perfectly simultaneous requests.
Its subtle, but concurrency is the simultaneous demand on processing resources by calculations.

@Seun:Thanks for a very clear explanation and sharing some of your languages approach to the issue. I was also hoping you could throw in some code to show the problem and then some more to show how it was solved with the additional library.

@kobojunkie: You're very right research is necessary at this point but sharing what you know on the subject can serve to throw more light on the ongoing research if you understand what I mean Awaiting you code contribution and hoping you speak a different language.

@Beaf: I quite agree with you. Concurrency issues seem to be very subtle. Another interesting observation is that concurrency was necessitated by the increasing demand for distributed computing which implied concurrent requests to your programs so initially some languages did not take this into consideration in their design and so solutions to these concurrency problems seems to be an afterthought as you can quite clearly see from Seun's post. What's even more interesting is that most of these new languages took these problems into consideration in their design

Yeah. Concurrency is hard. However, Java makes it easier for me. I hate pythons concurrency capabilities. Stupid GIL.

Yeah, its hard and its consequences are unavoidable whether a programmer is aware of them or not.

Concurrency prior to now didn't seem to gel with Object Oriented Programming; hence lots of programmers don't think of their solutions in terms of IO cycle/ CPU cycle/ another thread of execution etc. Which was sorta right ... but these days even if you don't explicitly use concurrency, your APIs do, so its best to be very aware of how to write concurrent programs.

worldbest: Yeah. Concurrency is hard. However, Java makes it easier for me. I hate pythons concurrency capabilities. Stupid GIL.

concurrency is not that easy in java either although a new higher level of concurrent abstraction was added to java 5 it is still bloody difficult in non trivial apps.

i heard and have seen examples in Scala and Erlang, both are functional programming languages, that handles concurrency with few lines of code using Actor model. Good new is Scala interoperate with java and compiles to JVM bytecode i.e java classes. Check out Scala concurrency examples on google,

Can you be more specific about the concurrency problem you're trying to solve?
New Nairaland uses DB transactions for concurrency, and it works well and scales nicely, too.

candylips:

concurrency is not that easy in java either although a new higher level of concurrent abstraction was added to java 5 it is still bloody difficult in non trivial apps.

Well. I am working on an enterprise project that involves intelligent use of concurrency. I haven't been struggling that much as I would if I was working with python. I don't think concurrency in Java is bloody difficult though.