the binary canary testing pattern

I think I just invented a new testing pattern - The Binary Canary.

Basically, I was grouping my PHPUnit tests into a test suite and I realized that my TestCase super-classes were "failing" because they had no tests in them. Obviously this is intentional - only the specific sub-classes would have tests.

I guess I could have made the TestCase super-classes abstract, but instead I added this to the highest-level TestCase class:

/*
 * global test plumbing here
 */
class Sfx_TestCase extends PHPUnit_Framework_TestCase
{
 public function setUp()
 {
  // more global test plumbing here
 }
 public function test_Binary_Canary()
 {
  $this->assertEquals(
         "Binary Canary says test plumbing is working.", 
         "Binary Canary says test plumbing is working."
        );
 }
}

My little binary canary serves two purposes:

1. It adds an "always-pass" test to each of my TestCase classes so they don't throw up any more PHPUnit warnings.


2. Because my TestCase classes set up context-specific test plumbing, the binary canary test inherited by each of them now alerts me if I screw up any of my test plumbing - and tells me the specific area.

For example:

class Sfx_Db_TestCase extends Sfx_TestCase
{
 public function setUp()
 {
  parent::setUp();
  // Db-specific test plumbing
 }
}

And:

class Sfx_Controller_TestCase extends Sfx_TestCase
{
 public function setUp()
 {
  parent::setUp();
  // Controller-specific test plumbing
 }
}

Just like the coal-miner canaries of old, this mechanism gives me a simple yes/no signal as to whether or not my test plumbing will soon kill me, and which plumbing code is the culprit.

bailouts

I usually try to avoid preachy blog postings, but I can't help it today. Blame it on "casual Friday" or something.

Like lots of people, I'm pretty angry about the government using our taxes to bail out Wall Street incompetence. But now I've gone from angry to pissed-off. Here's why ...

The Dot-com bubble crash wiped out $5 trillion in market value of technology companies from March 2000 to October 2002.[11]

Recent research suggests, however, that as many as 50% of the dot-coms survived through 2004, reflecting two facts: the destruction of public market wealth did not necessarily correspond to firm closings, and second, that most of the dot-coms were small players who were able to weather the financial markets storm.[12]

--http://en.wikipedia.org/wiki/Dot-com_bubble#Aftermath

Not only do I work for a dot-com survivor, I work for one of the most bruised, battered, and beaten-up dot-com survivors. I mean, just check out our stock chart! I haven't asked around the office for official corporate history or anything, but I'm pretty sure no-one remembers getting a couple billion dollars from the government while the company lost close to 99% of its market value, but I have been told there were weekly, if not daily, downsizing announcements.

Now I'm not jealous or vindictive on the part of my company - I wasn't even around for the toughest times. I'm just making a point here that market "catastrophes" and "crises" have happened, happen, and will always happen. Sadly, people will lose their jobs, their wealth, and their houses. In the case of the dot-com fallout, $5 trillion worth on Wall Street alone plus whatever subsequent losses are tallied.

6 weeks ago, the Congressional Budget Office said taxpayers would need to spend $25 billion to bail out Fannie Mae and Freddie Mac. Now we are seeing figures up to $700 billion. The figure could very well balloon to beyond the $5 trillion lost in the dot-com burst, no-one really knows - not you, not me, not Wall Street, and not Washington.

So why are we even considering footing the bill for this when no-one even knows what the total is, how it's going to be paid, or to whom we pay it? Obviously there are some very powerful financial market players gaming the system. They are not the "small players who [are] able to weather the financial markets storm."

It would be interesting to see if the Financial/Credit/Mortgage-Lending industry follows a Long Tail distribution. Might be a case of the tall head exploiting the public to avoid their necessary chop down to obscurity ...

unit tests and just-got-it-working inertia

I've been reading and enjoying The Productive Programmer by Neal Ford. It has re-ignited some of my passion for Test-Driven Development.

This morning I finished a first phase of "refactoring" some code architecture and found myself extremely hesitant to dive straight into the next phase. I think it's because the extent of my "testing" was to tab over to the fully-functioning web page and refresh after each code change. That's pretty much an "all-or-nothing" scenario.

And the thing about all-or-nothing scenarios is that once you've achieved the "all" state, you're very hesitant to go back to the "nothing" state. Maybe I'm starting to understand one of the benefits of unit tests as opposed to whole-sale acceptance tests. With smaller unit tests, you can move more concretely from nothing to something, then from something to something a little more, then finally to all done.

Brasil

One of the great things about working for a big-name web company is that you get big opportunities. I'll be speaking at PHP Conference Brasil '08 about how we use PHP at SourceForge.net. Needless to say, I'm very excited and planning some vacation time around the conference.

Open source is big in Brazil. And even more pertinent, Brazil is our 3rd-highest nation in terms of site traffic - after U.S.A. and Germany.

Hopefully I can try to overcome the language gap and present some informative material for everyone. One of the things that struck me when I joined SourceForge.net was that the site code isn't super-magic - it's really quite ordinary PHP, it's just very highly used.

Ubuntu FTW; Boot Camp FTL?

quick update on my ubuntu experience ...

webex went fine. linux client seamlessly downloaded and allowed me to join a webex conference. joined the teleconference via skype with just a small hiccup - took me a couple minutes of tweaking with various sound levels to get my mac's built-in mic going. but iSight camera was working with Skype straight away, so that was nice.

but, near the start of my first full day, I needed to investigate an IE bug, so I tried booting up into my Windows partition. epic fail. blue screen immediately upon boot. :( tried fixing my mbr via the Windows XP install CD, but no luck with that either.

I remembered that the ubuntu instructions for triple-booting suggested installing GRUB on the partition with ubuntu, and I also remembered that I had missed that step. so, I decided to reset and start all over again. I booted into Mac OS (no probs there - in all of this experimentation I never once had a problem booting into Mac OS), and used disk utility to destroy both my ubuntu and windows partitions.

this time, I tried the other (easy) approach and after installing Windows, I ran ubuntu installer. this time I remembered to install GRUB on the ubuntu partition, rather than defaulting to the mbr. however, when I booted up again, I had the same experience - Mac and Ubuntu would start up fine, but Windows failed to start again. :(

because I constantly need to test in IE every day, I sadly decided that I needed to stop with all the experimentation and just resign to using Ubuntu on my secondary computers. :(

it's too bad - I would love to have a triple-booting MacBook Pro, especially if I could fire up both my Windows and my Ubuntu systems inside a parallels or fusion vm.

Ubuntu FTW?

I picked up an Ubuntu CD at OSCON and have now installed it on my macbook pro. I have no qualms saying this version (8.04) is easily the best Linux experience I've ever had...

I had to connect to ethernet to get first batch of updates which also let me get ndiswrapper and appropriate driver for my wifi card.

once I had that, wifi connected and it was a single apt-get command to get the proper bluetooth module so my mouse would work. then I started downloading and playing with the new compiz stuff. I have to say, compiz effects blow away Mac OS X effects, though they're not quite as pragmatically integrated into everyday uses.

I fired up pidgin and got connected to our company jabber and my google talk account. similar simplicity and ease with Evolution for company email (though I also started trying Zembra Desktop since OSCON).

our setup at sf.net is kinda unique in that we have our own sandbox sites that we can access and edit via webdav, so I did a simple apt-get for davfs2, made the necessary mods to /etc/fstab and I was able to vim edit some code. but I decided to go looking for linux php editors. I tried bluefish and jedit and liked jedit much more - its performance with the webdav-mounted dir was much better.

I also exported my del.icio.us bookmarks and imported them over into Ubuntu firefox. and I installed GnomeDo because I'm a quicksilver junkie.

total time was probably 2 hours or so - much better than any of my other previous jaunts into Linux. large credit to the high-quality ubuntu wiki.

so I'm thinking to try out a full day of ubuntu tomorrow. my only concern is webex, though there is a native Linux client.

interesting mix of ideas

the two reading pieces I have sitting on my desk are this Origin of Wealth and the latest Wired magazine. interesting to be reading them simultaneously ...

From Wired:

"Scientists are trained to recognize that correlation is not causation, that no conclusions should be drawn simply on the basis of correlation between X and Y (it could just be coincidence). Instead, you must understand the underlying mechanisms that connect the two. Once you have a model, you can connect the data sets with confidence. Data without a model is just noise.
But faced with massive data, this approach to science - hypothesize, model, test - is becoming obsolete.
...
There is now a better way. Petabytes allow us to say: 'Correlation is enough.' We can stop looking for models. We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot."

From Origin of Wealth:

"... not only is there a problem with data that contradicts Traditional [Economic] theories, but many theories have simply never been properly tested. One branch of economics, called econometrics, deals with data analysis. Rather than testing theoretical models, however, much econometric work is devoted to finding statistical relationships between variables (often for public policy or other applied purposes). Unfortunately, statistical correlations don't provide a causal explanation of the phenomena. Furthermore, as many economists would point out, there is often a lack of readily available data to test theories with, and even data that is available is frequently noisy or otherwise problematic."

should be an interesting week as these seemingly conflicting ideas bounce around in my head.

Support Tech Opening @ SourceForge

SourceForge is hiring for a Level 2 Support Tech!

That is all.

boycott boycottnovell.com

okay, I'll be the first to say that the Novell-Microsoft deal was bad - way bad. and I am boycotting Novell myself. to summarize my perspective on it, I'll simply present Moglen's analysis on the subject, which I consider to be one of the top 5 extemporaneous monologues of all time:



but I also take issue with the opposite extreme that, because Microsoft has done bad things with the community, they are comprehensively unable to do any good things with the community. I know that boycottnovell.com isn't the only crier of this fallacy, but they're the ones who are recently taking pot-shots at my employer - a company that I think has always valued, and continues to value, the success of the community, or else I would not have taken a job with them.

from the inside, I can tell it plainly that Microsoft is simply a sponsor and participant in our 2008 CCA awards. technically, I also happen to know that we did nothing more than pre-load a big list of Codeplex project names into our CCA 08 database. there's no conspiracy to go thru all of our projects and attach Codeplex EULA's to them, or to use our CCA awards to scare the OSS community, as Roy Schestowitz seems to imagine.

Roy's pieces at boycottnovell.com seem to flow from his presumed indisputable inference that the motivations of not only Microsoft, but also anyone who collaborates with Microsoft, are sinister in nature. and he's repeatedly making these near-libelous statements, which boycottnovell.com is enabling, and this is the reason I'm boycotting them in addition to boycotting Novell.

as I preview this post myself, it is comical to contrast Roy's proof-by-verbosity case with Moglen's eloquent and exemplary case against a specific abuse.

Austrians want to be FREE yo

wow - I must be pretty stirred to actually write a blog post again, but here it goes.

I'd call myself an Austrian-leaning student of economics. I got a minor in ECON, but that's pretty much just enough to know that I really don't know very much at all. I hit up RSS feeds from the likes of The Mises Institute to keep myself in Austrian shape.

I've also been a big fan of The Long Tail, both the book and the blog. and Chris Anderson is on the advisory board of my employer, so I respect and subscribe to pretty much all of his ideas.

so when a friend shared a Mises article discussing Anderson's upcoming book - FREE with me, my interest was most assuredly sparked. but as I read, I was disappointed to find Fernando dismissing, whole-sale, Chris's entire analysis.

I actually agree with Fernando's closing thought - "With time rightly identified as a scarce resource, economic theory is needed to understand the interchange process." and I'd be willing to bet Chris agrees as well, since Chris's article plainly states: "There is, presumably, a limited supply of reputation and attention [i.e. - time] in the world at any point in time. These are the new scarcities — and the world of free exists mostly to acquire these valuable assets for the sake of a business model to be identified later."

so really, I don't think Chris's latest thesis is contradictory to the "laws" of economics, as Fernando apparently perceives. my conclusion would rather be that new and innovative business models will live and die by how well they apply of the laws of economics to actually-scarce goods in a new "freeconomic" culture.

I think we just have two different-but-overlapping spheres of study - economics and business. Fernando cites Buchanan's explanation of why marginal costs don't determine prices - with which I agree. having not read the cited book, I poked thru it with Google Books for "marginal cost" and came onto a few interesting blurbs:

Instead he [welfare economists] would introduce, as Knight did, the possibility that hunters, generally, may have some non-pecuniary or noneconomic arguments in their utility functions.

emphasis mine. so Buchanan points out that price-marginal cost scenarios tend to rely on non-pecuniary circumstances. does he further go on to refute that those kinds of circumstances occur? nope, not really - it seems he merely elaborates on what kind of analysis is produced by their inclusion.

In resorting to noneconomic arguments in the utility function ... the economist has shifted the whole analysis from a predictive to a nonpredictive and purely logical theory.

I don't think Chris would have any qualms about admitting his idea is a "purely logical theory" rather than a "predictive economic theory", and that's how I look at it as well.

and from the perspective of an entrepreneur hoping to enter the market, do I really care which it is? isn't it enough to observe that prices are converging to marginal cost, that indeed I am able to buy marginal units of storage and process capacity, and that technological advance and competition are driving each other in a cycle?

all this stuff is pretty new - we're not re-hashing scenarios that have been recorded in dusty economics tomes for decades. sure there have always been such things as cross-subsidies and non-pecuniary psychic revenue driving free economies; Chris's theory should at least be respected because it indicates these underlying economic forces emerging in a noticeable change of our culture.

this theory is like an elephant, and we're all a bunch of blind folks getting a feel for different parts of it. some of us might be observing only this or that piece of it and get the wrong impression of what it really is, but it's certainly something - we shouldn't touch a single piece of it and dismiss it altogether.

Re: An Open Letter to the Ron Paul Faithful

Mr. Wastler,

I don't want to simply rant or YELL at you, but I'd like to make a few points in response to your pseudo-accusatory remarks directed at Ron Paul supporters.

First, when you say that, " ... these Internet polls are admittedly unscientific and subject to hacking," you are incorrectly implying that other "legit" polls are not so. As I understand it, a "traditional" poll method leading up to primaries is to call 400 land-line numbers of previous primary voters. This alone introduces selection bias which detracts from the "scientific" accuracy of results.

Then, you go on to conclude that, "Our poll was either hacked or the target of a campaign." Now, whether or not your poll was hacked can only be known by CNBC and your web administrators. But since your core business is to deliver accurate financial information to the public, one would hope the security and reliability of your information isn't compromised. To give you the benefit of the doubt, we'll assume that your website is reliable and, also assuming very little "legit" Ron Paul support exists, that the poll was "the target of a campaign." Is this not, for better or worse, the way our political system works? Isn't the primary vote, the presidential vote, and the votes for every other elected position in our government the "target of a campaign?" So clearly, this isn't a sufficient reason to remove the poll, or else you would also have to remove coverage of every vote ever conducted. Your reasoning doesn't add up, so to remove *this* poll is to censor and edit facts based on some other judgment. If the actions of a well-organized campaign of individuals, acting within the rules *you* establish, makes a poll un-scientific, then how much more-so does your action of censorship?

Though your "show of hands" analogy is an admiral objective, you either intentionally refuse or ignorantly fail to see where it breaks down. It seems the poll was meant to be a "show of hands" of a certain or select few people in a certain or select space you consider to be "scientifically" representative. What actually happened is that masses of people from outside your pre-supposed space poured in to demonstrate not just their support they have for one of the options you were offering, but also to indirectly demonstrate the clumsiness and inappropriateness of your selection of individuals and space. Admittedly, this is more a criticism of all of "old media" than your poll specifically, but your poll is just another example of new forces at work in our country, and in your industry specifically. I.e., "old media" used to be the ' ... well-organized and committed "few" ... ' guiding ' ... a system meant to reflect the sentiments of "the many"' but that is changing. You should be worried.

-L

P.S. There are others who want to talk to you.

Ron Paul Excluded in Iowa, pt. 2 electric Bugaloo

Original digg post

I try not to post about political subjects, but this debacle is just too ripe and Ron Paul is just too cool for me to follow my normal political blog silence. Give an idiot a microphone, and I'll use it...

Ron Paul is being excluded from an upcoming presidential candidates' forum sponsored by Iowans for Tax Relief and Iowa Christian Alliance. Now, they have every right to exclude anyone they want, and no-one is arguing about that. But, if you know anything about Ron Paul, you're shocked to think that these two organization would exclude Dr. Paul, of all people. Ron Paul's campaign posted the public contact information (taken from their websites) for these two organizations and encouraged supporters to call and question the action. Apparently, the response is overwhelming the organizations, and a local radio show picked up the story and interviewed parties from both Ron Paul's campaign (Kent Snyder), and Ed Failor from Iowans for Tax Relief. The result is a pretty comical display of Failor's total ineptitude in explaining any kind of rational argument.

In response to the simple question: "Why was Ron Paul excluded?" he offers the following answer a number of times: "Because we drew a line of exclusion months ago." But this is nothing more than saying Ron Paul was excluded because he was excluded. It's circular non-sense.

Failor also offers up a dog's breakfast of off-the-cuff trite arguments in a scattershot manner.

Failor also tries to explain the exclusion by saying that they have an obligation to be educational and can't slant or bias their invitations towards people who favor any certain political agenda. But cutting taxes is already a political agenda, and the other candidates invited have pro-tax-cuts agendas, just as Ron Paul does - though none of them have the 100% record he has.

Failor then tries to explain that the exclusion of Ron Paul is done on a basis of credibility and that Ron Paul has less than 1% support in popular polls. However, Tom Tancredo consistently scores lower than Ron Paul and he was invited. In addition, Ron Paul's rise in internet popularity is nothing less than meteoric. http://www.pollingreport.com/wh08rep.htm

Failor says that they can't alter their lineup for logistical reasons. But he also says they are non-partisan and so they have to invite all candidates - not just Republicans - and they have to account for the possibility that ALL invitees could attend. If that's the case, then clearly, the event could have accommodated many more speakers - the Democrats who were invited but refused.

Failor tries to say that because they have to put on events that are reflective of what their members want to see, and because only a tiny portion of the response they've gotten has been from Iowans, that they were correct in their exclusion of Ron Paul. But Ron Paul's national popularity does not imply un-popularity in Iowa. It's a logical fallacy, and it doesn't prove anything.

Failor also tries to say that Ron Paul's supporters exhibiting "fringe-type" behavior shows that Ron Paul is not a serious candidate. He says supporters have called his home phone at all hours and said rude things to his family. But Snyder explains that the campaign only published the already-public contact information from the two groups' websites, and nothing more. Ron Paul's campaign is every bit as credible, based on its behavior, as any other candidate.

It seems to me this organization should have invited Dr. Paul first and foremost among all candidates, or at least that they should have jumped on the opportunity to get such a great speaker added to their forum. Instead, they've decided to be completely inflexible for a bunch of poor reasons. That's their prerogative, I suppose.



read more | digg story

Continuous Integration

This is another Agile/XP practice with which I'm fairly happy; although I haven't yet seen it live up to its full potential, that potential is great enough to make me a believer.

Continuous Integration is a process that completely builds and tests code frequently. The "process" usually takes the form of a dedicated server running special software that continuously performs a series of tasks similar to the following:
(though apparently this process can be un-automated by using a rubber chicken)

1. Perform an update from the code repository
2. If changes are found, run a build (compile, test) of the latest code
3. a) If successful, package the latest code for deployment or b) If failure, report failure

Although I'm a fan of CI, it seems to be a more complicated practice than TDD. Though my experience may be tainted by bad hardware + software on which our CI depends.

CI requires that you maintain an automated build script. This isn't a tall order amongst Java and other compiled-language developers since projects of any moderate size need an automated build to simply compile and to separate source code from compiled code.

Interpreted languages are a bit different, though, in that they can usually be tested immediately upon edit. As such, automated build scripts are a bit more un-common for software written in interpreted languages. But most interpreted language software projects of any moderate size do have a consistent process for deployment, even if it's as simple as: make db changes, move files - and most interpreted languages have builders to automated this consistent process. In PHP, I've been looking at Phing.

CI is really most helpful when the build process includes a solid test suite. (Defining "solid test suite" is an exercise left to the reader.) With a solid test suite, CI can help you catch bugs earlier than usual because it typically re-runs all those tests after every check-in.

In addition, CI gives creates a vast sequence of clean builds similar to the "nightly builds" you hear about in open-source projects - a finalized packaged release of the project ready for deployment.

Finally, if you ensure that your CI platform replicates your target production platform(s), you can use it as a reliable measure of your project's production-platform readiness. This can be a double-edged sword, however - if your CI platform is different than your target production platform, it can give you false confidence of production-readiness, and even cause problems that aren't caught until later in QA or worse, actual production.

As with TDD, the benefits are not without drawbacks and you should weigh them for your own project before deciding if/which/how Agile practices are adopted. CI has the above benefits, but it is also a fairly complicated development platform for which engineers will be primarily responsible - it sometimes requires a good deal of time and attention to keep going. You still have to judge for yourself if it fits into your project, goals, and style.

Yet Another Tutorial

Not sure what's come over me. Must be all the Coke ZERO. In any case, this tutorial is a continuation from my previous tutorials to setting up an automated build using Phing.

Test Driven Development

I've read a few anti- and pro-Agile rants in the past couple days. Because I'm somewhere in between, I can't really rant in either direction. Instead what I might try to do is give my opinion on the actual effects I've noticed from some Agile/XP practices on my own coding. Note that I'm probably only picking out the ones I like, so my posts on the subject will betray a pro-Agile bias. But, the sparse number of posts will hopefully balance the scales in demonstrating that there are only so many Agile/XP practices about which I actually care enough to write.

I'm a fan of Test Driven Development. I don't do TDD 100% of the time, and there are quite a few things I'm not sure how to automatically test (CSS tweaks, anyone?). But I agree with just about all the benefits I read in this pro-TDD article, though I'll re-arrange their listing by my personal opinion of their importance:

When you follow ... TDD, all your code will be testable by definition! And another word for "testable" is "decoupled". In order to test a module in isolation, you must decouple it.

When I use TDD, it forces me to write "decoupled" code. "Decoupled" is one of those magic words programmer types say to each other in intellectual flexing competitions. But TDD shows what it really means - code that can be isolated. The benefits of isolated/decoupled code are numerous - re-usability, less duplication, more concise, and I'd also say "testability" is a benefit.

This is not to say that you have to use TDD to write decoupled code. There are much smarter and more disciplined developers than myself all over who write excellent code without using TDD. But for me personally, TDD forces me into just enough structure and pre-code analysis to keep me from writing messy code that will need cleaning later. Speaking of which ...

Why don't we clean up code that we know is messy? We're afraid we'll break it. But if we have the tests, we can be reasonably sure that the code is not broken, or that we'll detect the breakage immediately. If we have the tests we become fearless about making changes. If we see messy code, or an unclean structure, we can clean it without fear.

The "safety net" feeling you get from gradually building up a big suite of automated tests is, like all "feelings", impossible to describe, but I'll try to relay an anecdote which might help.

I still don't consider myself a Java programmer, although I spend at least 50% of my time programming in Java. I don't "feel" comfortable in Java. But, in the course of our development, I made at least one deep and far-stretching re-factoring (more fancy talk for "change the guts of the code without changing its behavior") to maybe 30 different Java source files all over our code-base, with no hesitation before committing. My uncomfortable Java feelings were superseded by my comfort in the fact that all 800+ tests passed after I made the change. So I'm wasn't afraid of making the needed changes, even in a language in which I'm uncomfortable, because all the code was covered.

Again, this doesn't mean someone can't make sweeping changes unless they have tests covering all their code. I've simply noticed myself indeed becoming more fearless when I myself have to make those kinds of sweeping changes.

Have you ever integrated a third party library into your project? You got a big manual full of nice documentation. At the end there was a thin appendix of examples. Which of the two did you read? The examples of course! That's what the unit tests are!

This particular benefit is quite a distance behind the previous two. Mostly because the kind of example code you find attached to "nice documentation" is a better reference than unit test code. Unit test code is oftentimes performing some superfluous tricks for isolation, and/or hard to understand. It could be argued that if the test code is hard to follow, the tested code's design is to blame. But I think personally I'd rather move extra verbosity into my test code and keep my productive code clean. Could just be a matter of personal preference.

However, unit-tests are useful in understanding the intended use of the tested code. And programmers are more likely to spend their time to write test code that benefits themselves (see above) than they are to "waste" their time writing example code which benefits only others. So, lacking the refined, formal example code, unit tests can act as use-case specifications. (Though not 100% comprehensive design specs, as some might say.)

There are a few anti-TDD points with which I also agree...

For many, TDD is a new way of programming. As such, it has an associated learning curve, requires new thinking patterns, and takes time before it is comfortable to someone. However, I have found TDD easier and more enjoyable to adopt than Java. Some might say that isn't high praise for TDD, but to those people I would say, "Well, at least it's simpler than Java." TDD can be practiced in the language of your choice, and you will probably find that TDD resources from within your preferred language can really help to match your existing programming style with TDD.

TDD results in a LOT of code. In the course of adding tests for the ZF Tutorial app, I realized I was adding verbose testing code to already-verbose MVC code. Indeed, the test code for the controller was more code than the controller code itself. This is a simple fact of TDD - more code. There's no getting around it. You simply have to decide if the benefits of TDD outweigh its drawbacks, such as this one.

We have run into problems where our 800+ test suite takes a significantly long time to run (~10 minutes). This can be a real pain if you're working under strict CI rules in that every change you make, even that pesky css tweak, is supposed to be sent thru that test suite before committing to your source repository. Typically, though, once developers have the hang of TDD, they know what kinds of changes really need the entire test suite, and what kinds of changes can be simply checked-in, or can pass thru only a sub-set of tests. But the pain still exists in that you have an extra step of responsibility between writing your code and committing it. Again, weigh the benefits of TDD against this drawback, maybe come up with a compromise of some kind.

More tutorials

I've been neglecting my blog, but I've been working on more tutorials:

1. Tackling the Refactoring challenge from the Getting Started with ZF tutorial
2. Selenium Testing with PHPUnit

Controller Testing in Zend Framework

Ouch, the previous post here was pretty bad. Messy design and it didn't even work correctly. A better guide on the topic is here:

http://tulsaphp.net/?q=node/40

More on design patterns

Design patterns in the context of frameworks is a nice opportunity for me to take a break from my framework posts. I've been reading a bit on design patterns as a result of that previous post, and the poll activity on the Tulsa PHP Users Group.

In my readings, I came across this old post from the Loud Thinking blog re: Patterns & Practices over languages. Good stuff, and I'll recall my related personal experience.

Before joining SourceForge, I was okay in a few languages - Java, ColdFusion, WebMethods (shudder), Javascript, SQL, XSL/XPATH - and I was pretty good in one language - PHP. The first couple months at SourceForge were simply LOTS of PHP with a bit more architecture than I was used to - caching, entity objects, a Smarty view layer - but not much different than the PHP coding I had always done.

I was exposed to a pretty sophisticated Java architecture when I joined the Marketplace engineering team. But, my main task initially was to build a few simple wrapper classes that would allow our PHP scripts to communicate with the Java backend over STOMP to a message queue system (ActiveMQ, specifically). This would let us continue to write simple (display-oriented) PHP scripts which relied on the sophisticated Java backend for most logic. I still like that idea, and think it can work well if given proper time and attention.

But as we faced mounting time constraint pressure, and in the interest of leveraging our available talent pool, we stepped away from using STOMP and took out most of the PHP code - save for a proxy-type controller which speaks to the Java app server and some helper classes. We started using the Web MVC module of the Spring framework (in addition to the Core, Testing, Transaction, and Hibernate modules we were already using), and were therefore implementing everything, including display via JSP, in Java. So I poured over Spring reference material. (Highly recommended book)

It was quite a (beneficial) learning experience for me. As I learned (and accepted) the Spring approach, I found that I started to care less and less about Java - it just happened to be the language in which Spring is written - and I cared more and more about the ways in which Spring worked.

Those "ways" are, of course, design patterns. Inversion of Control, Object-Relational Mapping and Active Record, MVC, Front Controller, Command Controllers ... these are all just "ways" of doing certain things, solving certain problems, that are common to many systems. As I said, in learning Spring, I'm appreciating these patterns more than the Java language.

This is what lessened my resistance to frameworks, so much so that I became a fan of (some of) them! But, still having fondness for PHP, I started looking for PHP frameworks which work in the same ways that Spring does. But, it has really changed my perspective from analyzing languages based on syntax, (or types, or whatever) to analyzing languages based on the availability of design pattern implementations.

In this respect, and from my searching, I still think (though it may still be simple bias) that PHP is the best language available for the web. There are many PHP MVC frameworks (look at all the ones in the comments as well) which implement nearly all the best proven design patterns for the web. There's enough choice out there that PHP engineers can choose not only the design patterns to use, but also between many different styles of each design pattern. And of course, because PHP is so easy, it doesn't take an "architect" to craft their own style implementation of a favorite design pattern. (like the Active Record implementation I referenced in my last post.)

It was a big step for me to stop focusing on certain languages and instead focus on design patterns. I think it's been a big step upwards in my engineering aptitude/skill/whatever. And the more design patterns I learn, the more I realize there are so many others I don't know that could make my programming even easier. Sadly, I'm nowhere near the competence needed to recognize the "informal patterns" (i.e., unnecessary repetition/duplication) in my own code, but hopefully I'll get there.

For now, I'm still just learning my way into the topic and very glad I have an opportunity to do so.

Frameworks - 10 pounds of Design Patterns in a 5 pound bag (hopefully)

I was inspired to write this framework-related post when I read Dirk Merkel's "Practical Active Record in PHP" article from the latest issue of PHP Architect. Ironically, I was inspired because the PHP framework I'm most excited about (Zend Framework) has, AFAIK, chosen to use a Table Data Gateway pattern instead of an Active Record pattern, which I think is not quite as intuitive or easy to use. But a comparison of those two patterns is not only off-topic for this post, but well beyond my analytical qualifications.

Instead what I'd like to point out is the fact that all good frameworks should help programmers implement good design patterns. Programmers uncomfortable or unclear about frameworks might also be uncomfortable or unclear about design patterns, but design patterns are just ways of expressing common solutions to common problems, and frameworks can help teach good patterns. The web-proven ones that come to my mind first are MVC, Front-Controller, and Active Record, but there are many many more.

If you want to fully grok the Active Record pattern, you should of course read code & articles - like the article I mentioned above, or perhaps this article, or any other resource on the subject. But, since design patterns are common solutions to common problems, you don't want to re-write Active Record code every time you use a database in your applications. Instead, once you know how the Active Record pattern works, you can look for a framework which gives you an implementation of the pattern with an interface you like.

This is a much better approach to learning design patterns, IMO, than to learn design patterns outside of a framework. I find that resources covering design patterns without a framework tend to be overly abstract. Pattern implementations inside frameworks tend to constrain their context to the framework's intended use. I.e., you can learn much more about MVC theory by first diving into setting up an example Spring Web MVC application than you can by reading an abstract MVC blueprints paper.

When writing an application with a framework, you might get the feeling that you're dumbing yourself down - that you're relying on the framework too much and therefore becoming dependent on the way it works, or locking yourself into an architecture you don't fully grok and therefore can't fully debug. While that's entirely possible, it's more likely that, given the framework is reputable and has a solid architecture, you're actually relying upon good implementations of well-proven design patterns, and learning all about them in the process. This is not only good for your application, it's good for you as a programmer.

frameworks, licenses, etc.

I came across this interesting and relevant blurb from a Dave Rosenberg post on not using GPL for open source projects:

Spring as GPL wouldn't make a lot of sense, just like SugarCRM as Apache wouldn't make sense. (A very dumbed-down explanation being Apache is for ubiquity whereas GPL is for commercial.)

I'll take a stab at the fully-fleshed explanation.

The reason Apache is for ubiquity is because, as a permissive license, it allows the licensee to do more with the software. Most relevant, it allows the licensee to build new software on top of the licensed software - a derivative work - but close-source the new software. This makes Apache-covered (or more generally, permissively-covered) software a suitable selection for a wider audience - those that won't (or can't) return their own contributions to the community AND those that will.

The reason GPL can be said to be a more commercial-friendly license (from a project admin perspective) is because, as a viral license, it requires that any modifications made by anyone must be returned as open-source software as well. So a potential competitor cannot "steal" your code (it is free for use anyway), but even furthermore, they cannot improve upon your offering without returning those improvements to you as well - competitors will actually help you.

Now, back to frameworks. Spring, Struts, Tapestry, Google Web Toolkit, Zend Framework, CakePHP, Symfony, Ruby on Rails - all of these use permissive licenses. The reason, of course, is that frameworks are intended to build other software. And that other software won't always be open-source. But it's important for any framework to have a strong and active community in order to enhance the framework itself, so you want ubiquity and popularity over "purity".

If Spring, or any other framework, were licensed under the GPL, then all software built on top of the framework would also have to be GPL. I won't say that this never makes sense - it could be the framework author's intention that the framework only be used to create more free software. It wouldn't make sense if the author intends for the framework to be used in as many software projects as possible.

However, most of the frameworks I mentioned are licensed under Apache License, Version 2.0 (or other similar licenses), and apparently incompatible with GNU GPL. This is just another GPL annoyance to me. These frameworks are clearly good free software, and have adopted a widely-used permissive license in order to foster popularity and community participation. I'm all for free software, but when I use any of these frameworks in a free software project, I'll be using a BSD license.