Congratulations on being one of the first magazines to publish some straight scoop on OOP (object-oriented programming) in the May cover story called "Componentware." This can only mean one thing: OOP has lost its luster as a marketing gimmick. Will we ever stop being swayed by false promises? You can't turn chicken soup into a cure for cancer by simply stirring the pot a little.
Ernie Deel
Marietta, GA
Your May cover copy "Object-oriented computing has failed. But componentware...is succeeding" is causing many headaches. Why the sensationalist title? Why is it that in the cover story, everything other than Microsoft's VBX/OCX is an alternative? Why did you ignore companies such as ParcPlace Systems, Oberon Software, Inference, Easel, and at least a dozen others? I don't know, maybe I'm just overreacting to
your title page, which should have been called "Component Objects: A Look at VBX and Everything Else."
Chris Stone
Framingham, MA
Did you get your headaches before or after you read our cover story?--Eds.
I want to congratulate you on popularizing a new buzzword, component software, in the May cover story. Unfortunately, this is not what we software engineers need. We already have a reputation for jumping from one technological fad to the next and having more buzzwords than nearly any other profession. But hey, it sells magazines. And it sells software.
VBXes are cute and great for prototyping; however, I can't extend them as I can with C++ objects. As Visual C++ incorporates templates, my bet is that any leftover hype for OLE will continue to wane. Borland has a wonderful object-oriented library called OWL (Object Windows Library) that is based on templates. And between inheritance and templates, C++ has a rich set of features for creating reusable code. VBXes are not the epitome
of reusability.
David Cox
Rochester, NY
No, I don't think "VBXes are the epitome of reusability," and I didn't say that; however, despite many flaws, they are surprisingly useful for exchanging chunks of reusable code in binary form. I did, however, describe NextStep as having the same advantage that VBXes have but without VBXes' limitations, and I aired some issues that are being debated in a very lively way in the object community--notably, the COM/SOM controversy. I've heard of OWL, and in the cover story, I discussed its new relationship with Novell's AppWare and what implications follow from that. If your impression of the article was that VBXes are the be-all and end-all or that C++ is dead, then please reread it--that is not at all the message.
--Jon Udell