Ignoring JUnit

In the past I have been against ignoring broken tests (usually accomplished by commenting-out the @Test annotation) - usually looking something like this:

//@Test
public void brokenTest()
{
  . . .
}

Generally when I have been involved in a project where this happens those tests get ignored for so long that they will eventually be thrown away.

Recently, however, overwhelming demand for this feature in my etl-unit testing framework (I.E., it was suggested by one developer and I liked it) meant I had to do some refactoring. When I was in the middle of  changing some code in a common module I wanted to get an idea for where I was in the bigger project.  I suspected JUnit had an @Ignore annotation and I was right.  So, instead of the above, this is what I had:

@Ignore
@Test
public void brokenTest()
{
  . . .
}

I liked this much more, for a few reasons:  (1) my IDE, Intellij IDEA, will run all the tests, but leave the ignored ones highlighted to call attention to them, (2) a simple search of the project will reveal everywhere this has been used, (3) this serves as a reminder to me of how much work I have left to do, and (4) I can do a full multi-module build to evaluate parts of implementation as I go.

I do prefer my own implementation for a few reasons.  The JUnit @Ignore annotation has an optional value attribute which can hold a reason for the ignore.  In etl-unit, the @Ignore annotation has a mandatory reason attribute which should contain a description of what is broken, etc (reason seems so much better to me - value is too vague).  Etlunit goes a step farther, however, and adds an optional reactivate-on attribute which holds a date after which time the test is no longer ignored, but rather than executing will throw an ERR_IGNORE_EXPIRED error, or a user-specified error or failure can be substituted, the latter being used when you would prefer for the test to fail rather than an error being generated.

I really like the reactivate date since these kinds of tests can easily be forgotten about, and when your continuous integration server builds start failing you will have a not-so-gentle reminder.

Life without Caffeine . . .

I had to give up caffeine, fizzy drinks, and a few other pleasant things a few months back because a Doctor told me I am getting old (I had no idea!), and that has revealed some things about myself I wasn't entirely ready for.  For one, I rely WAY to much on caffeine, and it wasn't easy to get it out of my system.  I was irritable for a few days - like I usually feel when I take in too much caffeine.  Secondly, without being able to just drink a cup of coffee (Bleh!!) or, better yet, a Doctor Pepper, to get me past the afternoon, I now have to manage my biology in an entirely different way.  For one, I have to sleep (What a concept!), and when I hit a dip in the afternoon it's a reminder that I am not getting enough sleep, and correcting that is no small task.  Gone are the late nights working after the kids go to bed in the solitude and serenity of the living room - a time which formerly was my most productive for working on open source projects and other things I didn't have to actively account for my time to anyone.

In my case, just stopping the caffeine was an unworkable solution - unless I am okay with my evening commute inevitably ending in a fatal accident due to exhaustion.  I feel so much better not ingesting all the sugar and caffeine and other toxic crap that soft drinks and coffee bring (I've never cared for energy drinks), but it is more than just a change in drinking habits - it's really a change in how you manage your body.

Be Systematic

Years ago when my children had cousins who were into such things, I heard a cheer (ad nauseam) which went like this:

Be, Aggressive!B-E, Aggressive!B-E AGG-R-E-SS-IVE Aggressive!

While being impressed that you could effectively get a 7-year old to spell 'Aggressive' by putting it into a catchy cheer, a parallel occurred to me from my own experiences.

Be, Systematic!
B-E, Systematic!
B-E Sys-t-e-matic Systematic!

Some background.  In my career, one of the most frustrating things I have encountered  happens when trying to brainstorm a problem.  It usually goes something like this (in a group setting):

Facilitator:  So, the widget is failing every day.  Any ideas?
Suspicious Engineer #1: I suspect it is the hoozit of the whatzit.
Overconfident Engineer #1: Naw - it can't be that - the hoozit is finely-tuned.
Facilitator: Any other ideas?
Suspicious Engineer #2: What about the Ides of March?  It has been known to cause problems before?
Overconfident Engineer #2: Not anything I have seen.
Self-proclaimed SME and overall dominant personality:  I think it has to be muppets in the server room.
Facilitator:  I'll bet you are right!
All:  Let's go get those muppets!

The big problem here is the complete lack of discipline in testing conclusions and challenging assumptions.  The facilitator did the whole team a disservice by allowing dominant personalities (I.E., those with who make the strongest assertions) to squelch good ideas just by saying it isn't likely to be true.  A good facilitation of this sort of meeting should collect both pieces of information:  the identification of the problem area, and the likelihood (priority, or rank) that it is the culprit.

Instead of just throwing ideas around like clay pigeons and letting people take shots at them, the facilitator should collect all ideas and rank them according to the order the team agrees on, then test each idea in order.  This works not only on teams, but especially in individual work.  I can't count how many times I have been told by an engineer that the problem was identified, as well as a fix, but once the fix was done they discovered that either (1) there was an additional problem that hadn't been identified, or (2) the analysis was wrong and the fix didn't work.  This is terribly frustrating for a team - thinking a problem is fixed only to have it pop up again.  It is even worse for the team when the fix is reported and deployed to production only to have another team (usually production support, or worse the business users) report the lack of fix back to the team.

In conclusion:  follow the scientific method.  Observe and analyze the situation, make a hypothesis, test that hypothesis, then commit to work.  Challenge yourself.  Don't believe your own conclusions (let alone others) until you can see it (as close as possible - some problems in computer science just can't be observed - such as a rare race condition).

Why you should do assertions the way I don't want to do them . . .

I have heard people say things like "In unit tests, each test should pass or fail for exactly one reason."  Practically, this means there should only be one assertion per method.  Yeah, right.  I've done the math.  That's a friggin' big pile of tests when you could have 20 assertions in one method!  And who wants to think up that many good test names (because runtest0 will only go so far)?  Am I right?

So, I coded a test like this, because there were multiple assertion validations with one condition, E.G.,

void testSomething()
{
 stage(data);
 assert(reason1);
 assert(reason2);
 ...
}

That worked fine, and met my goal of being green and saving the planet by reducing the number of carbon-footprint bearing test methods.  Then I refactored something that I knew would effect this code.  Unfortunately, I had to do a lot of creative commenting just to get the crap fixed because the entire test wouldn't pass until all the assertions passed.  What a pain.

So, I guess there is something to that after all . . .

Test Driven ETL Development

I did a screencast showing how my team uses Test Driven Development to develop ETL code.  Here is the link:

Test Driven ETL Development

The presentation is a bit long because I discuss some of the theory of what we do and demonstrate the entire end-to-end process.

Siri, Punctuated

I have been a little tired of dictating text messages to Siri and having them come across bland - having no punctuation to speak of.  I would ask "How are you?", and she would transcribe "How are you."  On a hunch I tried speaking the appropriate punctuation and she picked it right up.

I tried:

Me: "Where are you.  Where are you question mark.  Where are you exclamation point."
Siri: "Where are you.  Where are you?  Where are you!"

Just for fun, I also did these:

Me: "Are you quote happy unquote question mark."
Siri: 'Are you "happy"?'

Me: "I picked up the kids semicolon they are very whiny."
Siri: "I picked up the kids; they are very whiny."

Me: "Are you colon happy comma unhappy comma or sad question mark."
Siri: "Are you: happy, unhappy, or sad?"

Anyway - pretty cool stuff.

This is on iOS 7 - not sure if it works elsewhere.

Maven assemblies with file permissions (modes)

I have this maven project which packages up some shell scripts and batch files in a plugin.  On the client side, they get extracted into a local directory where they can be used for various functions.  Long story short, I was a bit irritated that on windows the batch files are treated as executables, but on Mac OS X and Linux, they were simply text files, and had to be launched with:

sh etlunit

Granted this is not a big deal, but I was annoyed anyway.  So, I Googled and found a directive to the maven assembly plugin which supports setting the file mode.

<fileSets>
<fileSet>
<directory>src/bin/</directory>
<outputDirectory>bin</outputDirectory>
<fileMode>0755</fileMode>
</fileSet>

  </fileSets>

I tried that, and for some reason now my bin directory (created by the plugin) was inaccessible - it's mode had changed inadvertently - to 0000.  On a hunch I added the directory mode as well:

<fileSets>
<fileSet>
<directory>src/bin/</directory>
<outputDirectory>bin</outputDirectory>
<fileMode>0755</fileMode>
<directoryMode>0755</directoryMode>
</fileSet>

  </fileSets>

And now I feel much better about the universe.  I can add the scripts to the path or use:

./etlunit

It's little victories like these that make the Open Source Software world so much fun . . .