Thursday, May 27, 2004

When and where to methodologize!

Look I am not against methodologies, they are definitely a step in the right direction, but I think there is a mis-understanding of what the priority is in creating/using a methodology.

But we have to be using them for the right reasons, and prioritizing what's important in coding.

First off, the most important thing is:

1. Performance and Scalability, there can be no sacrifices made in this animals name. Simply if your methodology even takes 1 millisecond away, it's not worth it. Really, because it can so easily add up. Because without performance or scalibility, why are you even coding? Unless it's just for your own use?

2. Usability, can this methodology hurt/inhibit/enhance customer experience? Now my personal experiee in fuseboxed sites, is definitely yes! I hate fusebox url's, it may make sense to the coder, but i like having pages with distinct file names, that make it easy to understand where in the site I am.

3. Modularity, this is a big one, used as the catch phrase of methodologies, which method of modularity/code re-use do you use? Cfincludes, custom tags, cfcs, cffunction, udf, cfmodule? Remember priority one???

That's a big conflict, because seriously native cf, or cfincludes are top of the chart when it comes to speed and performance.

I am all for modularity/code reuse, as long as it doesn't hurt even by a nanosecond my performance or stability.

Because in this day and age of high traffic sites, and different new technologies competing for web application heaven, we must compete by only offering highly scalible, high performance, high available web apps/web sites.

Anything else, and we might as well not be coders.

It's drastic, If we want ColdFusion to be top dog, then we must stick to our guns to those top 3 rules, in that order!

Got it!

Go CF!!


  1. Anonymous2:49 PM


    You rant about how important performance is, and how OO technology is a "fad", but then why are you using Cold Fusion? The entire server is built using Object-Oriented technology, and there would be no Cold Fusion without "performance-sacrificing" software development methodologies. The bottom line is this: if you truly believe your arguments against the current "best-practice" ways to develop software (in the name of increased performance), there is absolutely NO reason you should be using CF as a development environment.

  2. The reason we went from C+ foundation to Java was for it's very stability and performance.

    The reason why I stick to ColdFusion is because of it's ease of developing applications rapidly.

    I agree java is powerful, but the learning curve is unrationally too sharp.

    The key to ColdFusion's power is in the speed of developing solid applications.

    I am not here trying to knock other languages down, I am just tired of everyone knocking CF down!

    The whole purpose of my blog, is to help people show by the quality of their work that CF is a top web/intranet solution.

    Too often we follow trends, rather than what is solid, stable and is the right way to the future.

    I know my opinions may not be popular, but that doesn't matter to me. I am not here because I expect you to agree with me.

    I am here to challenge that CF is a wimpy, old-fashioned web app, that java or C# or ASP.Net are the IN thing, and leave CF in the dust.

    Well I am all for learning how to improve the speed and performance of CFMX, but I am unwilling to use an unnecessarily complex language to do so.

    That is my humble opinon.