While working on a project in 2001, I realized I was doing an awful lot of copying and pasting ... both code and repetitious comments. While trying to find some code in a previous project that was also relevant to the current application, I found myself saying, "Self, there has to be a better way." I was spending too much time looking for stuff, copying that stuff, then modifying that stuff to make it work in my new project. All that rummaging and fixing was slowing down my progress on the new project. It occurred to me that having re-usable code and comments readily available (as in one mouse click) whenever I needed it would be a lot easier, smarter and a whole lot more productive then what I was doing. Duh! Thus was born the kernel of an idea that has evolved into Double Text.

As is my usual approach to such epiphanies, I jumped on the Internet to try and find an existing tool that would do what I had in mind. I was disappointed to find nothing. There are, as you are well aware, programs that can be used to store and regurgitate code snippets. But, none that I found had the flexibility I had in mind. Ergo, I began to create my own solution. It has been a journey of discovery that has, thus far, taken four years and has lead to an entirely new approach to the age old issue of re-usable code.

Why Repeat Text Writer?

As part of refining this new approach to re-usable code, I've been trying to figure out how to refer to Double Text. It certainly didn't fit the mold for any software that I'm familiar with. During the development of version 3.0, I finally hit upon the term "repeat text writer." I don't call it a repeat code writer because Double Text can be used for any text that gets re-used. The application has proven to be even more versatile than I originally envisioned. For example, I use Double Text to store my logons and passwords. They are very easy to find and re-use whenever I can't remember how to log into a web site. I've even got a library that I use to generate an email to my Shaklee distributor for our monthly order of food supplements. It takes me just a few seconds to select the products I'm ordering and then enter the quantities for those selected products (most of which default to 1). Double Text then generates my entire email that I paste into Eudora. It's a simple process, but, I don't have to re-invent that particular wheel every month.

Of course, I have also accumulated a considerable collection of code libraries. Since Double Text has virtually no language restrictions, my collection is eclectic. In addition to repeatable code for VB 6 and VB .NET (my preferred languages), I also have useful routines in Java, JavaScript, C, HTML and Crystal Reports to name a few. A library I set up recently will be used to generate RSS feed files. I enter the titles, descriptions and links and Double Text generates the entire XML file. I've even got the full set of medical ICD-9 codes which, as you can imagine, is huge. However, finding a medical code is easy and quick using Double Text. Why in the world, you may ask, do I have the ICD-9 codes in a Double Text library? The answer is typical for a programmer, I think. I was curious if it would work. So I downloaded the ICD-9 files and wrote a simple VB 6 program to convert those files to Double Text source files. It was an interesting diversion.

I've begun making a distinction between code that can be re-used, and code that is stored and documented in Double Text. I call the latter repeatable code.

Re-usable vs Repeatable

I think a generally acceptable definition of re-usable code could be "code or functionality that will probably be useful in one or more future projects." It's a category of code that few managers or developers seem to pay much attention to ... until they need it. This is unfortunate since, in my opinion, re-usable code, when upgraded to repeatable code, has the potential to save big on development costs by truly putting the "R" in RAD (Rapid Application Development).

From what I can gather, there are currently five primary sources for re-usable code.

• Components (objects) that expose their functionality through an interface
• User controls
• Repositories that use a database and accept a limited number of languages
• Web sites where code snippets and sample projects are freely exchanged
• Other projects you've worked on

Code repositories address this type of re-usable code. Unfortunately, code repositories tend to be limited. They can usually only be used with a few languages (normally one). In addition, repository programs typically use a database which makes the sharing of stored code less than easy. While these programs are typically very intelligent with the langauge they are designed to store, they are usually not known for their flexibility when preparing code for insertion into a new project. The ability to easily adapt re-usable code to new situations is essential, in my option, if code is to be truly re-usable.

Exchange sites, like The Code Project, make chunks of re-usable code, with instructions, readily available to anyone with access to the Internet. This free exchange of knowledge and specific coding solutions is ideal in concept. The usual implementation, however, has, in my opinion, room for improvement. You'll see what I'm referring to later.

Similar to exchange sites, the store of projects that a developer has worked on provide a gold mine of re-usable code. However, this type of re-usable code, like gold, must be found and then mined. And the search for, and mining of re-usable code in other projects is normally done only by the developer of the project because that person is usually the only one who has some idea of what nuggets are to be found. This "one person show" type of re-usable code makes this buried code of little use to anyone else.

Repeatable Code Defined

Now that you have an idea of what I mean by repeatable code, let's look at what typically happens when we take advantage of free code from exchange web sites.

Exchanging Code

The free exchange of knowledge and code is a long standing practice among developers. You know as well as I do that, especially with complex development environments like Microsoft's Visual Studio, getting some help from your friends can be a good thing, if not sometimes downright life saving. Contrary to popular opinion, Mountain Dew, Skittles and pizza can keep you going only so long.

You quick like a bunny copy the needed code from your browser or download either source code or a sample project. Ok, now what? If your experiences using free code snippets are anything like mine, the sample code you copied or downloaded will not just plug into your project. First you have to dig your way through the code to try and get a glimmer of what the code does and how it does it. Then you have to make the exchange code compatible with your code. Cold bananas! This may not be as easy as it sounds. When I go looking for code on the Internet, I typically have to find several samples for the same function before I can find code that does everything I need it to do. It is not unusual for me to have to cobble something together by taking snippets from several exchange sites. It is also not unusual to end up with code that flat doesn't work in my project. What then? This is, after all, public domain and tech support can be thin.

Once you do manage to wrestle the free code sample into submission ("It's how late?!?"), and the gloating for a job well done is over, what happens later? What happens when you need this code again? Fast forward a few months or years. You're writing a project that also needs the procedure that you struggled so hard to make work in your last project. Or was it several projects ago? You know you have it somewhere. So, instead of searching the Internet for code samples, you start searching your own projects. ("Where is that code?!?") And, once you find it, you are again faced with ferreting out all the pieces you need and making them work in your new project. The only advantage you have this time is that, hopefully, you don't have to try and understand what the code author (you) wrote. I'm sure that's not a problem since you added lots of comments ... right?

Does all this sound familiar? How much time do you think is spent by coders like you and me world wide searching for sample code, and then trying to make that code work? (I think someone should do a study on how much this code search and rescue time has cost businesses. We could be talking gazillions!)

Conclusion

Double Text has had a profound effect on the way I think about, and do my coding. I honestly don't think I would enjoy developing software nearly as much as I do if I didn't have this tool for repeating code so easily. Code that I use often, or code that is complex to integrate into projects are made readily available and simple to use. When I finish writing code that I've never done before, I immediately think about storing it in a Double Text library; both for my own use and to share with other developers. These are subtle shifts in my thinking that have enriched my software development experience.