HN Explorer

Ask HN: Should I compete with an Interpreter?

7 points by mkice ↗ HN
There are a very unique set of programming languages that are dated. The set of languages I am referring to is "Business Basic". You can get more info here: http://en.wikipedia.org/wiki/Business_Basic

Most of the larger players are either non-existent, or have been bought out. There are only a few major players left, http://pvx.com and http://pvxplus.com, http://www.thoroughbredsoftware.com/ and http://www.basis.com/

I was wondering, if I should create an interpreted business basic to complete, however make it like PHP (allow cgi sessions off of an apache server). There is currently nothing like this now. The majority of people using this, are pretty old (or ignorant), and wouldn't switch to Python, etc.

My biggest concern, is that there are really only about 3-5 million end users of programs that are created in Business Basic, guessing that there are probably 50,000 developers using these languages at most. Should I enter a dieing but specialized market, or should I go back to the drawing board?

3 comments

[ 3.6 ms ] story [ 17.3 ms ] thread
Xurinos
I would take this approach (and it may very well be flawed, or it may work)...

First, I would find people using business BASIC. Next, I would approach them and clearly state the problem you identified. I want to know if they would be willing to pay for me to come up with a superior solution. What other limitations have they experienced? Would they be interested in being part of a pilot program / alpha testing?

As this project has a high uncertainty, my goals would be to establish dedicated customer relationships early and to see if actual money could be committed to it.

In other words, I would strive to reduce the unknowns, to reduce assumptions by gathering real people with problems they are willing to pay to solve. The alpha/beta process gives them extra bonding to the project, and dedicated customers generate leads.

russell
Another approach is to create a BASIC to X translator, where X is some language with a future. I did a Pascal to C translator a long time ago and it was not a huge task, certainly less than a full blown language implementation. I would pick Python as a target language, because of the large libraries and the development activity in the language. Alternatively you could pick Java or the JVM.
CyberFonic
Many years ago I wrote a Data General (now owned by EMC and no longer making computers) Business Basic to C translator. Used it to port a suite of programs to Unix system. So the technique can work. But I agree with Russell, that Python would be a better target these days. You would need to take care to ensure that the generated source code is easy to read and modify by those who are familiar with the BB original.