#Toaster Software failure
15-Feb-94 22:28:48
Sb: #74662-#Toaster Software failure
Fm: Adib Nasle 72722,2052
To: Arnie Cachelin 71332,3717
We have a very controled duplication procedure and have almost no bad
disks (less than 1%) shipped. The duplicators and controlers we use
take a digital "picture" of the master disk and then copy and verify
the integrity of the data on each disk. Bad disks are automatically
rejected by the duplicator and are thrown out or used for internal
use (putting our own personal junk on them – since the disks are
still useable and are most likely rejected because of a minor flaw).
On top of that, we have a system that randomly checks disks by
actually loading them which makes sure that all the files load
properly. Moreover, our loading procedure allows you to continue
with the installation procedure even if a bad disk is detected (thus
not aborting a 45 minute install at disk #40).
I don't know if shipping two sets of identical disks is the best way
around this as it could cause licensing problems (the release of
Light Rave has shed light on this issue recently), not to mention the
added costs. All of this could be avoided through proper product
release planning and control over the duplication process.
My feeling is that Newtek is so anxious to get a new product
immediately on the market in enough quantity that they are putting
great demands on their duplication house (we need 10,000 disks
tomorrow scenarios), which is forcing them to compromise quality and
meet the required quantity.
But the product is so good that I think people will put up with bad
disks. It's kind of like the Italian car syndrome. They look great,
perform great , but once in a while they can be a real pain in the
neck.