Ok, I agree, in technical terms, the data is spread evenly on the drives. I was giving an overly simplistic view of RAID 5, and I failed in my example.
The idea, though, that the data can be regenerated is one drive goes out, is the main point of what I said, and that IS what RAID 5 is good for. RAID 4, you have to rebuild your data before the drive array is usable again, I BELIEVE, but RAID 5 is can 'limp along' by computing the missing data via CRC computations until the drive is replaced, THEN it does the rebuild, with the drives still available during rebuild.
Also, you can't say that RAID5 is more reliable than RAID1 (mirror).
But I did, and I do. RAID 1, one drive goes out, you have to switch to the mirror drive. This is not automatic. RAID 5 can continue processing data, as I mentioned before, by limping along. Reliability is operational up-time in this case.
But thanks for pointing out, with your kindly worded and through initial explanation, the glaringly huge whole in my example that totally invalidated the whole comment.
Oh wait, you didn't, and it didn't, and it wasn't.
Like I said, I was wrong on the technical detail of the RAID 5 example, but the whole comment is still valid as a discussing of backup. Your comment, although correct, was stated in such a way that it was saying "you spelled a word wrong, but I am not telling you which one."
Thanks again, though.