home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.object:3370 comp.lang.eiffel:1103 comp.lang.smalltalk:1753
- Path: sparky!uunet!psgrain!percy!data!kend
- From: kend@data.rain.com (Ken Dickey)
- Newsgroups: comp.object,comp.lang.eiffel,comp.lang.smalltalk
- Subject: Re: Future Issues of Object Orientation
- Message-ID: <708@data.rain.com>
- Date: 1 Sep 92 21:37:16 GMT
- References: <923@ast.dsd.northrop.com>
- Distribution: usa
- Organization: Microtek DSD, Hillsboro, OR
- Lines: 36
-
- dvorak@ast.dsd.northrop.com (dvorak joseph l.) writes:
-
- > I would like to start a discussion about what people believe
- > will be the critical issues in object technology in the next
- > 5 - 8 years. What solutions to these issues do you see and
- > how will the solutions affect the object model?
-
- I just have time for a few...
-
- There are a large number of issues surrounding persistent computations
- (ones which span a number of power-downs, migrate to newer/different
- HW) and persistent distributed objects in general. Doing persistence
- without global tables will be more important.
-
- My guess is that static typing and direct inheritance lose out over
- typed values and delegation style inheritance as mutability and the
- ability to "proxy" over a network become more important. It is
- horribly bad to export complex type signatures when you want to export
- some object interface. I expect interfaces to become more
- encapsulated.
-
- Another area of interest is the ability to mutate distributed objects
- over time. For example, you have a running connection server with live
- data used over a network and you want to "install" a new version
- without bringing down the server or interrupting its sessions. This
- essentially requires a transformation algebra which deals with both
- the live data and live connection issues.
-
- Robust systems and recovery will continue to grow in importance. When
- a bridge falls down, people go after the (civil) engineer(s) which
- designed it. I expect that people will start building software
- systems differently as we "software engineers" are held more
- accountable for software system designs.
-
-
- -Ken Dickey
-