Posted on Jan 26, 2014
Clients are always asking interesting questions. They’re interesting, not always because of the depth of the question, but because sometimes they force us to reevaluate our assumptions.
Recently I had a call with a company seeking a new maintainer for an existing (i.e. legacy) Rails application. As I spoke with them and looked over the application, it became evident that it was several versions out of date. As part of the work they wanted, I recommended they also upgrade the application to a more current version of Rails, to which they asked, “Why do we need to upgrade?“.
“Why do we need to upgrade?” is an interesting question. As a developer, I take for granted the need to keep applications current, but for business people, especially less technical ones, the need is less apparent. “After all”, they argue, “the application still works and does what we need it to do.”
There are three big reasons to upgrade: security, performance, and maintainability. Let’s address these reasons and also look at the potential costs of upgrading.
The number one reason - and this is a big number one - you need to keep your web application up to date is for security. There are a lot of very clever, but malicious, people out there looking for new and interesting ways to take advantage of even the smallest chink in a system’s armor. If your application isn’t kept up to date, it’s vulnerable to their attacks.
“But my site’s small, no one even knows about us.” This would be a great argument if the bad guys cared about what site they hit, but they don’t. They have programs which scan the Internet looking for any site which might be vulnerable. When they find one, like a tick, they dig in. Things can get really bad at this point.
If this happens to you, these are a just a few things you can look forward to:
The cheapest solution is to be proactive and make sure your application is kept up to date with all the latest patches. If it’s not, the bad guys’ll use the vulnerabilities you leave open. As they say, “It’s not a matter of ‘if’, it’s a matter of ‘when’.”
Most security patches are super easy to install and may take as little as 15 minutes to download, install, and deploy. There’s absolutely no reason not to do this.
If “security” is the number one reason to upgrade, then the second big reason is “performance”. Of course, this is a little tough to argue since there is evidence to show upgrading from Rails 2.x to 3.0 incurred a significant performance hit. Since then, however, effort has been put forth to improve performance, and we’ve seen that with the Asset Pipeline, the Arel SQL engine, “Russian Doll” caching, and more.
But performance isn’t merely limited to the applicaiton. As improvements are made to the framework , developers reap the rewards with expanded and improved features; features which reduce our development time (I can’t even guess how much time I’ve saved by using Arel, CoffeeScript, SASS, and the Asset Pipeline), expand what we can do, and simplify our workflow.
Some time ago, Google showed us that speed matters and directly affects your bottom line. It also been shown that a site’s performance directly affects its SEO. If that’s not compelling enough, then what about the savings you’ll reap in increased developer productivity? Is getting a few more hours of development time a month worth it?
When looking for reasons why you should upgrade, it is sometimes argued that you should upgrade your application so that it’s easier to upgrade later. While this argument isn’t wrong, it is a tautology.
Unfortunately, the simple truth is that as an app’s Rails version gets further away from what is current, there are fewer and fewer developers who are able or even interested in maintaining it. Furthermore, libraries and plugins used by an application and which add functionality and features to the application also fall out of maintenance - and do so much faster - as the creators focus instead on newer versions of the framework.
By keeping your application up to date, you ensure that you’ll have a larger pool of developers who can maintenance your application. You also reduce the numbef of hurdles needed keep the site running.
So here’s the big question: How much is this going to cost?
The cost to upgrading your application from its current state to the current stable release depends on a number of factors:
Most upgrades I’ve run across average between 6 and 10 hours of effort per “minor” version jump. Of course, that time is affected by the factors mentioned above, but these are general numbers.
With the added security, improved performance and faster development times, and simplified maintenance, the real cost is not in the upgrade, but in remaining with the status quo.
By putting off upgrades you run the risk of a security breach, you allow your application to be less performant, you hinder the ability of your developers to work quickly and effectively, and you add to your application’s future maintenance cost.
So how much does an upgrade cost? Not nearly as much as not upgrading.
I live in the greater Kansas City area with my beautiful wife, our two great kids, and our dog. I've been programming using Open Source technologies since '97 and I'm currently an independent software developer specializing in Ruby on Rails and iOS. I am for hire.