This is None in view mode; [Up]
Date: Sun 30-Sep-1991 20:50:23 From: waltrip@capd.jhuapl.edu Subject: <None> In article <1991Sep27.004249.12744@microsoft.com>, edwardj@microsoft.com (Edward JUNG) writes: [...lots of interesting stuff deleted...thanks for the info...] > Programming a trivial or prototype application is far > easier using OOP than traditional methods. But OOP does > NOT make it substantially easier to deliver a typical > commercial application. I hesitate to point out that > Windows applications are not the only ones that slip > ship dates... nor is this true of the OSes themselves. > Anyhow, there are very few people who have architected > and shipped large commercial applications. Interesting points. I don't think I agree that OOP makes trivial apps easier. I think interpretive languages such as BASIC or Windows BASIC is probably easier for trivial apps than OOP. And OOP doesn't benefit you much if the appropriate object libraries aren't available. The NeXT, of course, has the objects for GUI (NeXTstep) and some other things but, when you go beyond that, useful, pre-made objects can become scarce. The availability of NeXTstep and Interface Builder rather than OOP itself make prototyping and trivial applications on the NeXT easier than in most traditional programming environments. But I doubt if trivial apps are significantly easier (may even be harder) than with Windows BASIC (which looks like a neat product BTW). The great thing that NeXT is doing is making (and encouraging the making of) a large body of pre-made objects for a lot of purposes. An example of how this is expanding in very useful directions is the DBKit that will become available fairly soon. I believe that OOP in a suitably rich environment will always have very large advantages vs. traditional programming for any non-trivial project. (Even without this, a pretty good case can be made for the OOP paradigm for a large body of problems whose characteristics I won't try to define here. But saying that, of course, implies that there is also a body of problems that won't benefit from OOP. But I don't believe size is one of the relevant characteristics.) Main point here is that NeXT is significant not because it is a platform on which you can do OOP (you can do OOP on anything), but it's significant because it surrounds the OOP applications developer with a rich OOP environment that is widespread and well thought out. I believe it has an advantage over the current incarnation of Windows (or anything else I know of that's currently available) but I also know that MicroSoft and other Windows CASE vendors are working hard on providing more productive environments as well. Another point worth making with respect to environments, however, is that much of the impetus behind the strategic alliances (ACE, Apple-IBM) is the realization that a rich environment starts from the ground up with concepts that are rich enough to support a vast range of programming needs. The smart guys at NeXT built on the ideas of a lot of smart people. Now there are a lot of smart people who are also looking at NeXT and will certainly come out with something better than NeXT is today. The real test of how well thought out NeXTstep is will become evident as new features are added and as more developers bang away on it. Personally, I feel that NeXT's weak point at this time is that they have provided a rich set of interprocess communications (IPC) mechanisms but could give more thought to producing standards for defining some standard sets of objects that can be passed along these IPC mechanisms. I believe MicroSoft and probably even Apple are ahead on this (though not far ahead; I just believe that NeXT needs to be well ahead of the pack themselves on this front). [...] c.f.waltrip Opinions expressed are my own.
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.