#R3 vs. Silicon Garden
29-Oct-93 10:13:50
Sb: #63985-#R3 vs. Silicon Garden
Fm: Syndesis 76004,1763
To: Roger Danielsen 70323,3034
Autodesk multimedia developers can easily secure beta versions of 3DS, of
course… and in this case, the Garden routines are developed and/or
supervised by the same people who were working on the next revision. They
certainly had access to the betas.
As a fellow software developer, I'm completely sympathetic to an unspotted bug.
It's tough. Yes, there are complete disciplines and careers devoted to finding
bugs and "quality assurance", but few software companies can devote as much
effort and personnel to this task as many users might demand.
There are a number of similar scenerios that might crop up in the future…
Hordes of users might complain that R4/5/N, etc. stopped working with a
half-dozen company's IPAS routines. Of course it's a "good thing" that IPAS is
improved, but that necessarily will break old software. I don't think it's
Autodesk's responsibility to test and debug everyone else's IPAS routines, yet
with only beta-quality pre-release versions of R4/5/N, how can a third party
reliably debug the interface? It isn't "final" until it ships, and then some.
Someday, I'm sure that third-party developers will complain that it's "unfair"
that any IPAS interface enhancements to R4, R5, etc. will be documented and
exploited by the Yost Group before anyone else, simply because the new spec
remains undocumented until well past the release. On the other hand, it's good
that a company with a real stake in IPAS is developing real-world routines with
any new spec, as opposed to a company handing down a relatively untested
interface.
Allowing add-ons is a two-edged sword! Look at AutoCAD. Someday, Autodesk
might start bundling IPAS routines that compete directly with the hard-won
efforts of loyal third-party developers, and more toes will be stepped on.