Posts
1150
Comments
891
Trackbacks
1
Developer Friction is Relative

So, my buddy Derik Whittaker posted about MSTest and some problems he had with configuring stuff when using StructureMap.  The post was entitled "MSTest, why I hate you.... You cause me too much friction."

Now, my main point is not specifically connected with the details of Derik's post, so I could just leave that alone, but I can't resist.

Basically, Derik (who is used to using NUnit and MBUnit) was running into problems with getting MSTest to work nicely with StructureMap.config files.  He even asked the Austin Bobbsey Twins (one of whom wrote StructureMap, the other knowing as much about it as anyone I know other than the one who wrote it) for help, but in the end, basically moved the config data to app.config and went with that.  Read the posts for the details.

Now, if you read the comments, you'll discover that there are easy solutions to the problems Derik had.  Not to put too fine a point on it, but basically, he didn't RTFM, and incorrectly expected MSTest to work as NUnit did, and when it didn't, it caused what is called 'Friction' (which I don't know is ever defined anywhere, but it's basically those things that make you stop your productive work and have to deal with other things...there, that's vague enough). 

He could, of course, have entitled his post something like "Problems I've Had With MSTest and StructureMap.Config" but that would have been boring.  Better to talk about hating MSTest and get some extra kudos at the next 'Alt.NET Self-Congratulatory Back-slapping Workshop' or what-have-you.

Leaving all that aside, as one of the comments makes clear, the cause of the friction was because he expected MSTest to act like NUnit, as he had a lot of extensive experience with NUnit.  Without that previous experience, another user was able to get the problem solved much much quicker.

And I think this is actually a much more pervasive thing.  If you have followed Rob Conery's MVC series, you have watched an episode (one of the best, if longest) with Jeremy Miller, once again about StructureMap, and there is a moment where he is supposed to grab control of the session and start coding, but Rob's Visual Studio install doesn't have ReSharper, and so things comes to a screeching halt, since Jeremy is one of those luminaries who are self-addicted to ReSharper (which shouldn't be mis-read as a criticism, I understand the issue).  For someone like myself who has tried and tried and tried (and I'm giving it one more shot with ReSharper 4.0) ReSharper but hated it every time (I think because of how it reacts with aspx pages for a large part), working in VS 'nakedly' is easy.

I think this is important because Alt.NET people LOVE to talk about reducing friction, but always seem to think that this is some objective thing, as opposed to understanding that friction is often (almost, but not quite) always tied to their previous practices and experiences.

None of this should be understood as a post saying that MSTest is as good or better than NUnit.  AFAIKT, it isn't.

posted on Saturday, July 26, 2008 6:33 PM Print
Comments
Gravatar
# re: Developer Friction is Relative
Brian
7/27/2008 7:36 PM
Good is subjective.

Java is not as good as .NET.
VB is as good as C#.
Oracle is as good as SQL Server.
MSTest is as good as NUnit.

All these things are not quantitatively measurable and can start a pissing match in a heart beat.

Good, like beauty, is in the eye of the beholder; the term 'developer preference' has gone AWOL over the last couple of years and now everything is 'friction'.

Post Comment

Title *
Name *
Email
Url
Comment *  
Please add 4 and 7 and type the answer here: