Gaussian 98 makefile policy

[frisch at.at gaussian.com (Mike Frisch)]

C. Klein writes:
 >
 > Richard Walsh's suggestion sounds very reasonable to me. Gaussian98 may
 > be a nice program (once you have it running) but the company's policy,
 > translated to the automobile business, appears to be:
 >
 > "OK, we'll sell you the car (program), but you have to produce the
 > proper key (makefile) yourself.. if you copy the key from someone, we'll
 > sue you ... maybe we can give you the key for the trunk.". Rather
 > strange way of doing business.
 >
 This is not a fair statement of our policy.  We provide source code, which
 few other vendors do, and makefiles for supported platforms and compilers.
 We have a careful testing procedure to help ensure that the program is
 fully and correctly functional with the supported configurations.
 People can do what they want locally to make things run on on other systems,
 but passing around makefiles for unsupported platforms which have not been
 subject to careful testing is not helpful to users and also technical support
 more difficult for us.  We originally started to forbid this after a number
 of problems occurred with people unknowingly using unreliable and incorrect
 versions.  In this particular case, where there is a carefully tested and
 supported version (makefile and compiler), I think most users are much better
 off using it.
 To pursue your analogy, we're not saying you can't use the car, we're saying
 that we'll show you the blueprints to help you fix your car, but you can't
 start building your own inferior copies and selling or giving them to people.
 I do agree that this is a strange way of doing business.  The normal way
 of doing business, which is what most of our competitors do, would be to not
 license the source code at all and hence not be subject to criticism of the
 terms of the source license.
 Michael Frisch
 Gaussian, Inc.