On this day in 1469, the Italian philosopher and writer Niccolo Machiavelli is born. A lifelong patriot and diehard proponent of a unified Italy, Machiavelli became one of the fathers of modern political theory. The term “Machiavellian” is used to describe an action undertaken for gain without regard for right or wrong. Though Machiavelli has long been associated with the practice of diabolical expediency in the realm of politics that was made famous in The Prince, his actual views were not so extreme.

How does that apply to performance testing web applications? Well, I have encountered several QA professionals that use the title “Performance Engineer” that base their testing decisions more on their self-serving career goals than what is best for their employer or customers. Specifically, there are many testers out there that will recommend LoadRunner as the only tool that could possibly be good enough for performance testing their clients’ web applications. The true reason for such Machiavellian behavior is so that the “expert tester” can change higher consulting fees. If the client will spend several hundred thousand dollars for the tool, then the consultant can breath easy – his part of the project cost will be minor in comparison (under the radar of scrutiny). Additionally, by forcing the most expensive tool solution onto their clients, performance consultants justify the extensive training and years of experience with the tool. It eliminates many of their competitors. The decision is more about sales than solving the customer’s problem.

On to other tidbits.

Why You Should Do Your Own Performance Testing

Ecommerce experts GetElastic wrote a good article explaining why no one should just accept the performance claims of software vendors entitled: Why You Should Do Your Own Performance Testing. The post describes ways in which “performance characteristics of your ecommerce implementation can be different from standard vendor’s implementation:”

  1. Structure and Size of Data
  2. Store Usage Patterns
  3. Caching Strategies
  4. Storefront and Backend Integrations

Good points, and it seems to me that these apply to any software that you implement.

State of the Web 2010

Here are a few of the more interesting tidbits from a survey of web developers and designers brought to you by Web Directions con­fer­ences, and Scroll Magazine. It covers 50 questions cov­er­ing tech­nolo­gies, tech­niques, philoso­phies and prac­tices that today’s web pro­fes­sion­als employ.

Seeing as how I bought a Mac in 2008, I must agree with the choice of OS X. I have had probably 2 or 3 times where the beach ball of death killed my mojo in the past 18 months. Whereas, my Vista machine was requiring 2-3 Bill Gates’ fixes (reboot) per DAY! No kidding. It was horrible for productivity – and my attitude. That’s not even counting the number of times I had to completely re-install software packages when they quit working in Windows.

The hardware will have impact on the frequency of crashing, and my Dell wasn’t the highest quality machine available; however I am certain that the OS deserves more of the blame. OSX isn’t perfect, but it’s been much more stable than Windows for me. My only regret is that it took me so long to switch.

When I had friends that were Apple fanatics back in the 80s and 90s implore me to dump my DOS or Windows machines, my response was, “If I was a graphic artist, I’d get a Mac. Since I’m a programmer, I’ll stick with what everybody else is using because it is the machine of the business world.” You could fairly say I was a Microsoft zealot back then. These numbers prove that the trend in software development is to work on a Mac. It’s simply a better platform.

Yep, I agree with Firefox. It’s surprising to me how only 2% of developers choose IE. Not that I would expect many to say it is the best technology, but there are a huge number of Microsoft bigots that think any code produced in Redmond should be placed on a pedestal. Is IE really THAT bad? I can see that argument for IE 6, but from what I’ve seen, IE 8 is a quantum leap forward in quality.

I just started using Google Chrome this year, and its usage has rocketed onto the scene with impressive numbers within the devel­op­er community. Barely behind Safari (only 3%), it proves how truly influential Google is in all things web related. I like it because of its performance – it is faster than Firefox. Still, I like some of the features in Firefox (auto save my tabs) and the plug-ins available such as Firebug.

Does the use of test browsers strike you as odd or unbalanced? I suppose that most developers don’t use IE themselves, but they feel compelled to test IE more than any other because of the bugs and lack of standards compliance of the MS browser.

An alternative and additional reason is likely that developers are aware of the continued large market share of IE in the user community. We may not like it, yet we must keep the users happy.

As a tester, I will test my stuff in IE more because of the bugs. Ok, I admit that I don’t like testing in IE so I probably don’t give it the attention that I should. This data is a good reminder why I should.

Lastly, the overwhelming choice for Javascript libraries is unsurprising. JQuery has been the leader for some time now, and these numbers confirm that it’s dominance is increasing.

Although JQuery may be the most widely adopted library amongst the developers surveyed, there is an interesting performance comparison of JQuery against Dojo, MooTools, and Prototype that shows that other libraries can be much faster. Adoption or speed? I would take speed. Wouldn’t you expect that from a performance testing guy?

Other Key Findings in the Survey

  • CSS3 is being used by 45% of the developers responding, double from last year
  • Font linking with @font-face is up to 23%, compared to 4% last year
  • Last year, less than 10% used HTML-5 and this year about 30% are using it in some way
  • Mobile browsing is dominated by Safari, usage doubled from last year to nearly a third
  • Android use is growing, but still at a small 4%
  • Desktop-​​like appli­ca­tion frame­works (e.g. Cappuccino) don’t seem to be catching on – it’s just a bad idea to make a web app mimic a desktop app.

I hope you find this useful. Our big thanks to all of those involved in conducting the survey and compiling the results! Just in case you didn’t get your high school diploma, find an alternative to the GED test at Franklin Virtual Schools. They are a friend of ours.

Similar Posts