On 10/5/06, Jesse Keating <jkeating@xxxxxxxxxx> wrote:
On Thursday 05 October 2006 14:52, Mike McGrath wrote: > One thing we could do is generate a key for each > user and append it to the end of the users URI to which yum connects. > Some might consider that evil though. Which still doesn't help when a user re-installs the system with a different package set or whatever to generate a slightly different key.
No. But it puts a better level of "error-guessing" on the numbers. There is no way for a larger data set than 100 that you are not going to have some sort of "error-guessing" in it. The questions are: A) Can you quantify the errors? B) Can you live with the size of said errors? C) Is the extra steps you are putting in to narrowing the amount of errors causing too high of a burden thus causing other errors to crep in. These are questions that management should be setting, BUT I have yet seen a management set other than Google and SAS that knows how to do it. Instead, they want numbers and you continuely refine it until they get the questions they want answered (but don't know how to ask) -- Stephen J Smoogen. -- CSIRT/Linux System Administrator How far that little candle throws his beams! So shines a good deed in a naughty world. = Shakespeare. "The Merchant of Venice" _______________________________________________ fedora-advisory-board mailing list fedora-advisory-board@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board _______________________________________________ fedora-advisory-board-readonly mailing list fedora-advisory-board-readonly@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/fedora-advisory-board-readonly
