>>It's not quite that bad, we just can't test every single graphics and video card.<<
Agreed. But there are some things that could have been done that weren't.
One is to write code strictly to specifications, following all the rules laid out in the documentation. In contrast to what I believe has been done, which is to write code based on experimentation, doing whatever seems to work.
And while testing every single card is impractical, it is unreasonable to completely ignore whole major families, such as the S3 (may not be relevant now, but certainly were when I started with MC2001i), simply writing the off as incompatible.
There are not that many lines of video adapters, it is not unreasonable to test a dozen or so cards representing the most common cards with representatives of all lines.
Nor is it unreasonable, after so many years, for somebody to track down what Mathcad is doing wrong that it so commonly fails with standard acceleration features.
Vista and 64 bit drivers are new enough that one can expect some problems with some configurations. But what is PTC's official response to these problems? Will they continue the tradition of simply saying that it is imcompatible, and telling the user to change his settings, his OS configuration, and perhaps even his hardware? Or will they accept responsibility for their product and look into the problems and provide fixes for them?
__________________
� � � � Tom Gutman
