Re: 4GL's (was: Application modernization & re-engineering)

david gibbs wrote:

You're 80% of the way there.
You are correct ... I don't trust generated code any further than I can spit it. I far prefer to use code that I wrote or at least can reasonably maintain myself (which is why I'm not adverse to using a code Wizard to create a starting point).

I understand this point. But I hope you agree this is purely a personal stand - David Gibbs prefers rolling his own, although he will accept help from the friendly Wizard <grin>. I understand that position, and even agree wherever practical. However, I hope you'll also agree that lots of shops don't have the time to address their application backlog, much less rewrite existing plumbing code, and that there are valid business situations to use proprietary and/or generated code. We all rely on code we cannot maintain to one degree or another, especially those of us who use Windows (much less those who program for it).

So, the idea that the code is generated shouldn't be a showstopper for everyone, even if it is for you. For people who don't have the time to reinvent or even reimplement the wheel, generated code can actually be a positive thing because it allows them to concentrate on the business logic rather than the plumbing.

The rest of my dislike is half the quality of the code they generate ... and half being tied to a single companies product to generate the code. Heaven help the company who used a code generator to develop a major application ... have the generator's author go out of business ... and then have something go wrong with the generator due to an OS upgrade. Yes, in the case of EGL, I know the likelihood of this happening is vanishingly slim ... but not with other products.

So you agree that your other issue is one that really doesn't apply as much to EGL.

Let me pose a question: do you think there's a larger contingency within IBM backing EGL or RPG? I know it's not quite an apples-to-apples comparison, since RPG is likely to be around as long as the platform, but I'm just pointing out that EGL has as large a commitment from IBM as anything. Rational has its own conference, and it's a big one. EGL is prominent there, and getting moreso.

I guess I'm saying that while the idea of EGL (or indeed any technology) being orphaned is not a completely unjustifiable concern, it's harder to justify when your primary development choice is RPG and the i.

Joe