Posts Tagged ‘Code Duplication’
What 108M Lines of Code do not Tell Us
Source: Nemo
Coming on the heels of Gartner’s research note projecting $1 trillion in IT Debt by 2015, CAST’s study provided a more granular view of the debt, estimating an average of over $1 million in technical debt per application in a sample of 288 applications. Between these two studies, the situation examined at the micro-level seems to be quite consistent with the state of affairs estimated and projected at the macro-level.
My hunch is that the gravity of the situation from a software quality and maintenance perspective is actually masked by efforts of IT staffs to compensate for programming problems through operational excellence. For example, carefully staged deployment and quick rollback often enable coping with defects that could/should have been handled through higher test coverage, lesser complexity or a more acceptable level of code duplication.
Part of the reason that the masking effects of IT staffs are not always fully appreciated is that they are embedded in the business design of IT Outsourcing companies. The company to which you outsourced your IT is ‘making a bet’ it can run your IT better than you can. It often succeeds in so doing. The unresolved defects in your old code plus those that evolved over time through software decay have not necessarily been fixed. Rather, the manifestations of these defects are handled operationally in a more efficient manner.
Think again if your visceral reaction to the technical debt situation described in the Gartner research note and the CAST study is of the “This can’t possibly be true” variety. It is what it is – just take a quick look at Nemo to see representative technical debt data with your own eyes. And, as indicated in this post, it might even be worse than what it looks. As Gartner puts it:
The results of such [IT Debt] an assessment will be, at best, unsettling and, at worst, truly shocking.
Technical Debt Meets Continuous Deployment
As you would expect in a conference entitled velocity, and in a follow-on devops day, speeding up things was an overarching theme. In the context of devops, the theme primarily manifested itself in lively discussions about the number of deploys per day. Comments such as the following reply to my post Ops Driven Dev were typical:
Conceptually, I move the whole business application configuration into the source code…
The theme that was missing for me in many of the presentations and discussions on the subject was the striking of a balance between velocity and quality. The classical trade-off in process control is between production rate and product quality (and safety, but that aspect [safety] is beyond the scope of this post). IMHO this trade-off applies to software just as it applies to mechanical or chemical processes.
The heart of the “deploy early and often” strategy hailed by advocates of continuous deployment is known deployment state to known deployment state. You don’t let the deployment evolve from one state to another before it has stabilized to a robust state. The power of this incremental deployment is in dealing with single-piece (or as small number of pieces as possible) flow rather than dealing with the effects of multiple-piece flow. When the deployment increments are small enough, rollback, root cause analysis and recovery are relatively straightforward if a deployment turns sour. It is a similar concept to Agile development, extending continuous integration to continuous deployment.
While I am wholeheartedly behind this devops strategy, I believe it needs to be reinforced through rigorous quality criteria the code must satisfy prior to deployment. The most straightforward way for so doing is through embedding technical debt criteria in the release/deploy process. For example:
- The code will not be deployed unless the overall technical debt per line of code is lower than $2.
- To qualify for deployment, code duplication levels must be kept under 8%.
- Code whose Cyclomatic complexity per Java class is higher than 15 will not be accepted for deployment.
- 50% unit test coverage is the minimal level required for deployment.
- Many others…
I have no doubt whatsoever that code which does not satisfy these criteria might be successfully deployed in a short-term manner. The problem, however, is the accumulative effect over the long haul of successive deployments of code increments of inadequate quality. As Figure 1 demonstrates, a Java file with Cyclomatic complexity of 38 has a probability of 50% to be error-prone. If you do not stop it prior to deployment through technical debt criteria, it is likely to affect your customers and play havoc with your deployment quite a few times in the future. The fact that it did not do so during the first hour of deployment does not guarantee that such a file will be “well-behaved” in the future.
Figure 1: Error-proneness as a Function of Cyclomatic Complexity (Source: http://www.enerjy.com/blog/?p=198)
To attain satisfactory long-term quality and stability, you need both the right process and the right code. Continuous deployment is the “right process” if you have developed the deployment infrastructure to support it. The “right code” in this context is code whose technical debt levels are quantified and governed prior to deployment.
Measuring Agile Success Rate the Right Way
Much has been said recently about the success/failure rate of Agile projects. In particular, a debate arose around the success rate of Scrum vis-a-vis Kanban. For example, in a post entitled Some Day Kanban will fail 75% of the Time, colleague Jurgen Appelo states as follows:
Unfortunately, some people arguing against Scrum include these ScrumBut teams in their evaluations of the “high failure rate” of Scrum. They love quoting that “at least 75 percent of Scrum implementations fail.” And I think “Yes of course, 75% fails when that includes the teams that don’t understand what they’re doing.”
I would like to add one other “dimension” to the discussion: boundary conditions.
Any Agile initiative – Crystal, Scrum, Kanban, etc. – typically starts from a certain state of affairs of the code that has already been developed using a Waterfall method or no method at all. Even brand new projects produce code that invariably interacts with other software components that are already deployed, warts and everything. Pristine environments with no technical debt for the Agile initiative to deal with are rare.
Like it or not, the Agile initiative is saddled from the outset with a certain amount of technical debt. Code has been duplicated, rules violated, complexity ran amuck, etc. A typical enterprise software team starts with hundreds of thousands $$ in technical debt, if not millions. This debt needs to be “paid back.” Probably not over night, but certainly over a period of time. As illustrated by the following figure from Jim Highsmith, things get ugly if the debt is not paid back over an extended period of time.
The evaluation of success or failure of the Agile initiative needs to take technical debt into account. A team of 50 with an accrued technical debt of $100,000 has a much easier job on its hands transitioning to Agile than a similar size team starting with $1M in technical debt on its hands.
Whatever criteria you use to determine whether an Agile initiative has been successful, I would suggest the following boundary condition needs to be satisfied:
Technical debt at the end of the project/initiative must be significantly lower than technical debt at the start of the project.
Use the techniques outlined in Using Credit Limits to Constrain Development on Margin to calculate technical debt before and after. In addition to qualifying your Agile success, quantifying technical debt will do a lot towards improving the quality of your software.
The Agile Executive 