All: This is my first post on this thread, and I'm here in response to an email I received asking for my input on Sapiens. I have no interest in Sapiens, personally. A few weeks ago, I did some cursory examination into their Web site in response to some comments on another thread.
Having been intimately involved with software development for over 25 years, I know that it's much harder to produce fine quality software than most people would imagine. And, you can't just review a Web site and be an "expert" on any subject, any company, or any of their products. However, I can use my experience in dealing with the various software challenges to form an opinion on whether or not a given tool is viable in its marketplace. Note that everything in this post is my personal opinion.
That said, allow me to respond to Addi's post ...
Generally, language conversion tools are powerful utilities that will assist in migrating from one version of a compiler to another, and/or to migrate (or, share code) between/among multiple hardware platforms. They are not "silver bullets", as typically there is a large amount of manual programmer intervention required. If done properly, these tools will go a long way towards getting you to the desired objective.
Version upgrades within the same language are necessary evils, and these types of tools do a great job of performing those types of conversions. However, I question the desirability of converting from one language to another. Sapiens' Falcon tool reportedly converts Assembler code to 'C'. I saw the examples on their Web page.
This type of conversion might be useful for converting a few assembler utility programs or subroutines, but to convert an entire application seems out of the ordinary. It's a much larger task, and has many more environmental gotcha's that need to be considered. It would seem more beneficial to me for the same amount of effort to be placed into developing a new application with the more powerful application generators available today.
The comments Addi made about taking ugly COBOL applications and converting them to 'C' to make them Web aware is a valid point ... as long as you restrict it to common subroutines. Again, an entire application conversion seems abnormal. By the way, anybody who has ever tried to maintain a large program that is the result of an automated language conversion program would most likely be thrilled to be able to maintain that same program in its original form. Talk about ugly ... !!!
In summary, I think that language conversion tools have their place in the market. I don't believe, however, that the overall appeal and usefulness is as great as the people who are promoting them would like to think ... or more correctly stated, would like you to think. If that's true, then the valuations of the stock will be lower than what has been near-hyped on this thread. Caution is advised.
Regards,
TED |
