Ayende @ Rahien

Turkey commented on The Common Service Locator library

Fri, 17 Oct 2008 00:36:42 GMT

great an article, I'm very thank you, I think really usefull, I'm following comments, thanks

Andrey Shchekin commented on The Common Service Locator library

Fri, 03 Oct 2008 21:04:05 GMT

Sure. I would have said that 'Base' is enough, but I see your point. However, what about ServiceLocatorImplementationBase? It is long enough so no one sane will use it as reference (having interface), and it is not breaking .NET guidelines.

Chris Tavares commented on The Common Service Locator library

Fri, 03 Oct 2008 15:52:59 GMT

@Andrey: Yeah, the name of that class is ugly. The choice of name was actually deliberate on my part - I wanted to make it abundantly clear that it was only a convenience for implementation of the interface, and NOT something that people should be referring to otherwise. Did it succeed in that goal?

Will Shaver commented on The Common Service Locator library

Fri, 03 Oct 2008 14:18:32 GMT

There's a big discussion regarding this topic on the MVC google groups. [groups.google.com/.../b86881b920cba0b4](http://groups.google.com/group/mvccontrib-discuss/browse_thread/thread/b86881b920cba0b4) The problem being how to handle things that implement IDisposable, and how to know when to release them.

Andrey Shchekin commented on The Common Service Locator library

Fri, 03 Oct 2008 06:16:46 GMT

Good job. I was going to create a container abstraction as a part of net-ioc-frameworks ( [http://code.google.com/p/net-ioc-frameworks/](http://code.google.com/p/net-ioc-frameworks/)), now I can do the locator part in a most interoperable way. The only thing that I really dislike is seeing a class with 'Impl' in its name. What's up with .NET use-full-words policy?

tasarım commented on The Common Service Locator library

Thu, 02 Oct 2008 22:08:00 GMT

It is great. Thanks for your efforts

Chris Tavares commented on The Common Service Locator library

Thu, 02 Oct 2008 21:56:26 GMT

And this, ladies and gentlemen, is why it took two months to write an interface with 6 methods on it. I feel like I just earned a degree in cat herding. ;-)

hammett commented on The Common Service Locator library

Thu, 02 Oct 2008 20:48:31 GMT

I'm not looking for agreement. I'm just saying that your use case isn't necessarily the one on the project.

Ayende Rahien commented on The Common Service Locator library

Thu, 02 Oct 2008 20:47:10 GMT

I don't think that we can agree on that, because I don't think that a factory has to create a new instance. In fact, since most of the time, the factories that I am thinking about looks like this: IFoo Create(string param); I see no reason to add unnecessary level of abstraction just to be pattern complete.

Neil Mosafi commented on The Common Service Locator library

Thu, 02 Oct 2008 20:47:04 GMT

Interesting, I actually created my own IServiceLocator interface in a project a while ago to allow my factory classes to locate classes to provide to consumers. It only had one method - Resolve

hammett commented on The Common Service Locator library

Thu, 02 Oct 2008 20:42:43 GMT

That's different from saying that people wont need to implement Factories anymore. Factories carry semantics, a well defined contract. The Service Locator wont always match up. Starting off with the fact that a factory is suppose to create things - always - whilst the service locator is to locate (maybe create) things.

Ayende Rahien commented on The Common Service Locator library

Thu, 02 Oct 2008 20:38:43 GMT

From CodePlex's site: "The library provides an abstraction over IoC containers and service locators. Using the library allows an application to indirectly access the capabilities without relying on hard references. The hope is that using this library, third-party applications and frameworks can begin to leverage IoC/Service Location without tying themselves down to a specific implementation."

Ayende Rahien commented on The Common Service Locator library

Thu, 02 Oct 2008 20:36:55 GMT

Hammett, From my first email in the subject: "We need some sort of a common way to refer to a container that isn't tied to a specific implementation. This is important in all the situations where you are writing a library that should make use of a container but not a particular one."

Louis commented on The Common Service Locator library

Thu, 02 Oct 2008 20:33:47 GMT

Very cool. No release/dispose? Looks like IServiceProvider is baked in there. [www.codeplex.com/.../FileView.aspx](http://www.codeplex.com/CommonServiceLocator/SourceControl/FileView.aspx?itemId=332676&changeSetId=24262) And a static ServiceLocator.Current with a delegated implementation. Excellent. [www.codeplex.com/.../FileView.aspx](http://www.codeplex.com/CommonServiceLocator/SourceControl/FileView.aspx?itemId=332683&changeSetId=24262)

hammett commented on The Common Service Locator library

Thu, 02 Oct 2008 20:09:48 GMT

Could be yours, but it's not mine, nor we have consensus that this is a good idea. So please don't mix the message of the goals, which should be clear on the codeplex page.

Jimmy Zimms commented on The Common Service Locator library

Thu, 02 Oct 2008 20:08:59 GMT

While it is for-suck, hopefully the abstraction is a strongly typed extension to IServiceProvider from mscorlib. It's the lowest common demoniator in .net for service locators. L'Shana Tova!

Ayende Rahien commented on The Common Service Locator library

Thu, 02 Oct 2008 20:07:28 GMT

That is my own use case for this

hammett commented on The Common Service Locator library

Thu, 02 Oct 2008 20:03:25 GMT

It is not intended as a replacement for specific factories.

Lucas Goodwin commented on The Common Service Locator library

Thu, 02 Oct 2008 19:45:18 GMT

I remember when this was being bantied about a while back as an idea. I'm impressed it turned into reality. Good job everyone who worked on it.

chrissie1 commented on The Common Service Locator library

Thu, 02 Oct 2008 19:30:46 GMT

Cool.