Head On - Latest Comments

Re: 1996 revisited

bill gates — Wed, 05 Aug 2009 18:01:43 -0000

it either means the images were inaccessible or safari doesn't render html in a useful way :)

Re: Smidig Videos Are Up

mahnve — Thu, 23 Apr 2009 14:49:16 -0000

Tack!

Re: Smidig Videos Are Up

Magnus — Thu, 23 Apr 2009 09:30:34 -0000

Tycker jag var bra.

Re: The correct Telenor Mac downloads

Marcus Ahnve — Mon, 09 Feb 2009 09:33:25 -0000

I am actually using Globetrotter Connect again, it's working just fine. I am on version 1.3.0 (d163)

Re: The correct Telenor Mac downloads

björn — Mon, 09 Feb 2009 08:32:20 -0000

Agree with Thomas. This is driving me crazy, 1,5 years after your post. How did you solve it?

Re: How To Automigrate the Test Database Using Merb, Datamapper and RSpec

ottobar — Tue, 20 Jan 2009 14:16:20 -0000

It also looks like it is now DataMapper.auto_migrate!

Re: Agile Sweden Christmas Party

Marcus Ahnve — Wed, 10 Dec 2008 18:11:51 -0000

Hej Denise,

Kolla med Ola Ellnestam på Agical om det finns platser kvar. Det är på deras kontor på Västerlånggatan.

Re: Agile Sweden Christmas Party

denise — Wed, 10 Dec 2008 09:55:50 -0000

hej MARCUS!
kan man få komma o lyssna?
vart i såfall?

Re: Words don’t come easy: MoSCoW and BDD

Måns Sandström — Sun, 07 Dec 2008 10:53:30 -0000

The reason for the mix of should and test, in this case, was due to the fact that not all code was written when pairing and some people simply felt more comfortable using test in the beginning of the project (probably due to previous JUnit 3 experience). Old habits die hard. However I think that some of the tests/designs could have been written after the fact.

Re: Words don’t come easy: MoSCoW and BDD

Mark Needham — Fri, 28 Nov 2008 21:51:12 -0000

Interesting idea - I guess there would be value in the Must, Should and Won't, not sure about the Could though, that would be very vague.

I suppose Won't is the opposite of Must but then it seems to close the discussion on whether or not the assertion made after the Must/Won't is actually valid. In the nuclear example you gave clearly that makes sense but the systems I've worked on at least aren't as clearcut and things that seem 'definite' can change, hence the value of the 'Should' idea.

Mark

Re: Words don’t come easy: MoSCoW and BDD

Mathias Stjernström — Thu, 27 Nov 2008 06:12:48 -0000

I like it. I have had this weird feeling about 'should'...

"It should, but i does not have to"

"shall or must" are more static...

interesting...

Re: Words don’t come easy: MoSCoW and BDD

Torbjörn Gyllebring — Thu, 27 Nov 2008 04:34:48 -0000

A slightly better formated version of the SWiM comment can be found here: http://torbjorn-gyllebring.blogspot.com/2008/11...>

Re: Words don’t come easy: MoSCoW and BDD

Marcus Ahnve — Thu, 27 Nov 2008 04:19:48 -0000

Måns, why mix 'should' and 'test'? To me that sounds like the 'test' once are written after the implementation, I am wrong?

Re: Words don’t come easy: MoSCoW and BDD

Marcus Ahnve — Thu, 27 Nov 2008 04:18:23 -0000

Torbjörn, I totally agree with the notion that 'should' encourages dialog. Use it as much as possible when discussing features. However, the specs also acts as documentation of design decisions, and as such they should convey as much information as possible.

Re: Words don’t come easy: MoSCoW and BDD

Marcus Ahnve — Thu, 27 Nov 2008 04:06:12 -0000

Steve,

BDD is all about conversation, and people already have their own definition of 'should'. Let's work with that instead of redefining the meaning.

Re: Words don’t come easy: MoSCoW and BDD

Torbjörn Gyllebring — Thu, 27 Nov 2008 03:56:53 -0000

I propose that we start advocating that our tests should "SWiM" that is they should specify theese things:

Should - the SUT "should" do this. "Should have page title"
Won't - the SUT Won't do this. "Won't call the database"
instruct - the SUT is used in this way. Learning tests, samples.
Must - the SUT Must do this. "Must shutdown reactor when overheating"

Re: Words don’t come easy: MoSCoW and BDD

Torbjörn Gyllebring — Thu, 27 Nov 2008 01:31:58 -0000

If I remember correctly Dan North is very fond of saying that language matter, a lot, and from what I remember of him describing the history of BDD and the epiphany of "Should" it had quite much do to with the fact that "should" evokes dialogue. As you put it in your post "“the title of the index page should be ‘blabbr.com - the worlds largest social network’" reading the test name ""Shouldhaveindexpagewithtitleblabbrcomtheworldslargessocialnetwork()" you can in essence say, "No it shouldn't". And change both the test and the implementation to match, the testname "TestIndexPageTitle" needs much more context to figure out what it's doing, and even when you do the whole test business seems to make us inclined to read "MUST" into the whole thing and make the majority of us qutie reluctant to change it. Should seems to be vauge enough for us as people to question and discuss about. It also has the fantastic property that it's quite easy to figure out what you're supossed to be driving at when you're set to write the next test, and ask your pair "so what should this fella do next" and you get the answer "Well now the Frooble should Brobble the Thingmajig", if you're a fast typist you already have the corrosponding "ShouldBrobbletheThingmjigwhen_something()" and you're set to go. The question "Wha't the next test?" seldom leads to "BrobbleThingmajig()" in the same fluent manner. This also works if you're alone, any time you need to start thinking about where to go next it's often quite convinent to just start with well now Frooble should... and something almost always appears.

Basicly my view on "Should" is that it's a catalyst for dialouge and an often needed invitation to discuss and explore the domain, the actors and their responsibilities. And I guess that sometimes we arrive at things that really "must" happen or "won't" happen, and I can't find a sensible reason to not use that phrasing if that's what we really mean, it really seems silly to get stuck in dogmatic use of should, espicially given that the outset of the whole thing is that language matters, limiting ourself to a single epxression seems to me, very strange in that situation.

Re: Words don’t come easy: MoSCoW and BDD

Steve R — Wed, 26 Nov 2008 19:56:23 -0000

Or you could define 'should'...

http://www.ietf.org/rfc/rfc2119.txt

Re: Words don’t come easy: MoSCoW and BDD

Måns Sandström — Wed, 26 Nov 2008 17:56:35 -0000

An interesting notion. Though I like the "should"-convention it can at times feel a little forced. I can see your point in Must and Won't, but I have a little problem with Could. It seems too weak to be of any value.
Looking through the code of one of my current assignments I realized that we've used "should" in aprox. 90% of the cases and "test" in the other 10%. Slightly surprised. I didn't think that we were so steadfast in our naming. Perhaps we shouldn't be?

Re: How To Automigrate the Test Database Using Merb, Datamapper and RSpec

Dylan — Sun, 02 Nov 2008 21:45:06 -0000

Seems like you need to make it rake MERB_ENV=test db:automigrate (without the dm:) now.

Re: New Job in Startup Land

Nenad Paunovic — Fri, 03 Oct 2008 08:58:50 -0000

Det är alldeles utmärkt att det idag finns internetbehandlingar med inriktning mot kognitiv beteendeterapi (KBT). Det finns vetenskaplig evidens för dess effektivitet vid ett flertal allvarliga psykiska störningar.

Re: ThoughtWorks Sweden Is Open For Business

Marcus — Tue, 26 Aug 2008 07:41:00 -0000

Kedar, thanks! I sincerely hope that we will need your help soon :).

Re: ThoughtWorks Sweden Is Open For Business

Kedar — Mon, 25 Aug 2008 13:31:16 -0000

All the best Marcus! Let me know if you need help from India :-)

Kedar

Re: Not everybody is into Euro 2008

Peter Krantz — Mon, 30 Jun 2008 02:08:27 -0000

So, what sport was it? :-)

Re: Merb, Leaner And Meaner Than Rails

Aaron Qian — Fri, 27 Jun 2008 05:21:20 -0000

First I met Rails about 3 years ago, that time I was using C#. I was wondering if there was a good ORM that can dynamically adapt to the database. When I found Rails, I was so happy.

Now when I tried Merb 0.9 and Datamapper 0.9. I was overjoyed!!

Go go go, Merb and Datamapper!!! :D