Comments for the sleepy dev

Comment on Static methods: feel the fear and do it anyway by Mark

Mark — Tue, 27 Jul 2010 08:10:22 +0000

I like this idea a lot. I usually follow ReSharper’s suggestion, since I believe that marking a method as static lets the reader know that it won’t change the object’s state (which is valuable information) and also it prevents a future maintainer from accidentally causing some state-changing side-effect.

However, I don’t usually think about reviewing the static methods I have so I can turn them into instance methods on a new class (or an existing class, if appropriate). From now on, I will!

Comment on Spark codec for OpenRasta by thesleepydev

thesleepydev — Mon, 26 Jul 2010 07:33:48 +0000

Hi Howard,

Thanks!

There is an updated version on Github (see note at top of the screen). I would suggest looking at that. If you have stuff to contribute do send me a pull request!
thanks

jen

Comment on Spark codec for OpenRasta by Howard van Rooijen

Howard van Rooijen — Sat, 24 Jul 2010 08:54:03 +0000

Hello,

Wonderful work on the codec – I tried to get it working this week – but it seems that it is slightly out of sync with the latest version of OpenRasta 2.0.3.0 – I’ve made the changes required. What’s the best way of submitting this patch – I’ve far too used to working in Git rather than SVN. Actually if you had the inclination – moving the project to GitHub might make it easier for the community to help support the codec.

Thanks for all the hard work,

Howard

Comment on Spark codec for OpenRasta by RobertTheGrey

RobertTheGrey — Wed, 21 Jul 2010 17:14:13 +0000

This is top stuff!

Your extentions are almost identical to what the Bindings features gives ASP.NET MVC users. Fantastic use of the framework.

Comment on Spark codec for OpenRasta by thesleepydev

thesleepydev — Wed, 28 Apr 2010 18:50:47 +0000

Erm, as it works in OR it should manage both, but having a try is probably the best thing!

Comment on Spark codec for OpenRasta by DaRage

DaRage — Fri, 09 Apr 2010 14:20:21 +0000

Very interesting. Does this run outside IIS or it requires IIS like the webform codec?

Comment on Spark codec for OpenRasta by thesleepydev

thesleepydev — Sun, 31 May 2009 15:44:42 +0000

Thanks! It wasn’t hard to figure out after I followed the word ‘extension’ around in the code for a while to see what it did – that’s the beauty of open source code – it documents itself! Tried to override some of the behaviour of WCF once (as you are supposedly able to) and it just had me tearing my hair out, documentation or no documentation…

Comment on Spark codec for OpenRasta by Louis DeJardin

Louis DeJardin — Fri, 29 May 2009 23:48:28 +0000

Remarkable!

Excellent work hooking into the undocumented parser extension interface. It was added originally for the MonoRail Spark view engine to weld ViewComponent instances into place by the element name without the core library’s knowledge of their implementation details. ViewComponent is sort of like MonoRail’s approach to web controls.

Never guessed the extension interface would be used to augment the parser with framework-specific markup syntax. Very clever.

Comment on Adventures in BDD style testing – too many scenarios! by thesleepydev

thesleepydev — Wed, 06 May 2009 16:31:11 +0000

Code smell? I suppose it is the programming equivalent of stinking bishop!

Anyway, you are right really – with those three bits of assertion I was testing, I was still measuring just one real behaviour – that the save operation failed with a nice error message. I guess I was just getting carried away with tests.

Only nice thing about splitting it out into separate tests is that you get the benefit of having three separate failure points – if the test fails cos it calls save AND doesn’t return the right message then in the original scenario I would get two failing tests whereas here I would just get one failing test.

Comment on Adventures in BDD style testing – too many scenarios! by Jon Neale

Jon Neale — Tue, 05 May 2009 14:26:03 +0000

Having to repeat only slightly different test code seems like a smell to me. If your test code becomes a nightmare to maintain, you’ll lose confidence in it, and therefore half the reason for having it.

Personally, I’d try to divide and conquer. Couldn’t you have something like

[Test]
public void Result_is_failure()
{
Result_should_fail_with_message(“Moon is not in the seventh house);
}

Result__should_fail_with_message(string message)
{
Result.Success.ShouldBeFalse();
Result.Message.ShouldEqual(message);
MyRepository.SaveWasNotCalled();
}

It would make thing much more maintainable if you should decide later on that maybe you want your conditions to change slightly