mo@seismo.UUCP (Mike O'Dell) (04/02/85)
A quick rejoinder to Guy Harris's reply to the person placing Faith in VMS RMS: I have personally seen RMS write files which it cannot later read. It somehow corrupted the length count at the front of a text record, causing all hell to break loose when you tried to copy the file. With the aid of 2 very powerful VMS wizards, we dumped the file, found the damaged count, and superzapped it to the correct value. Afterwards, everything worked fine. Incidently, the file was created with standard VMS utilities in the first place. The point is that VMS RMS is no more perfect than any other piece of software, and faith in the "security" afforded by untouchable kernel code requires a leap of faith which is beyond my subscription. "If it works, you're lucky." -Mike O'Dell