rogers@ofc.Columbia.NCR.COM (H. L. Rogers) (08/08/88)
In article <1275@sfmag.UUCP> der@sfmag.UUCP (D.Rorke) writes: > Applications written >to issue n of the interface [SVID] will continue work properly on >a system which conforms to issue n + 1 (or any subsequent issue) >subject to a specific evolution mechanism. Does this not, *by definition*, limit technical advancement by constraining new technology with *all* *old* technology? You can do things like dual or triple "universes," but one can only carry so much baggage while successfully delivering new technology. I guess most of us do this anyway for the sake of migration, saving previous investments, etc.; just trying to find out if others see this as a technical handcuff. -- HL Rogers (hl.rogers@ncrcae.Columbia.NCR.COM)
allbery@ncoast.UUCP (Brandon S. Allbery) (08/10/88)
As quoted from <226@ofc.Columbia.NCR.COM> by rogers@ofc.Columbia.NCR.COM (H. L. Rogers): +--------------- | In article <1275@sfmag.UUCP> der@sfmag.UUCP (D.Rorke) writes: | > Applications written | >to issue n of the interface [SVID] will continue work properly on | >a system which conforms to issue n + 1 (or any subsequent issue) | >subject to a specific evolution mechanism. >--^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | | Does this not, *by definition*, limit technical advancement by | constraining new technology with *all* *old* technology? You +--------------- Note the underscored phrase above. The "specific mechanism" in the SVID allows features to be moved from the "permanent" category to the "may disappear" category in a subsequent release of the SVID, and from there to a special category in some future SVID which will phase the feature out in something like 3 years, thus giving developers plenty of time to prepare for such changes. ++Brandon -- Brandon S. Allbery, uunet!marque!ncoast!allbery DELPHI: ALLBERY For comp.sources.misc send mail to ncoast!sources-misc
fangli@ihlpl.ATT.COM (Chang) (08/10/88)
In article <226@ofc.Columbia.NCR.COM>, rogers@ofc.Columbia.NCR.COM (H. L. Rogers) writes: > In article <1275@sfmag.UUCP> der@sfmag.UUCP (D.Rorke) writes: > > Applications written > >to issue n of the interface [SVID] will continue work properly on > >a system which conforms to issue n + 1 (or any subsequent issue) > >subject to a specific evolution mechanism. > Does this not, *by definition*, limit technical advancement by > constraining new technology with *all* *old* technology? You .... > saving previous investments, etc.; just trying to find out if others > see this as a technical handcuff. > -- > HL Rogers (hl.rogers@ncrcae.Columbia.NCR.COM) By your definition, yes, it is a technical handcuff, if you think restrain from changing user interface is a technical handcuff. In the software business one can adapt any new technology INSIDE one's software but you never ever change your user interface, you just add new parts to it but by any means avoid modifying the current interface. Think about it, SVID is the user interface of UNIX(R) System V by definition. Fangli Chang If every new issue of SVID is like MVS to DOS, it is not technical wise, it is suicidal.