You get a compiler error from doing what? I wasn't giving a code example, I was noting that scalar variables occupy space in memory and have to be able to be referenced, hence they have a dimension.
What's your unmentioned better way of looking at it?
Thank you for contacting me to voice your opposition to the recently released regulations by the Federal Communications Commission (FCC). Having recently been elected to my second term in Congress, Iâ(TM)m honored to be able to represent the people of Ohioâ(TM)s 14th Congressional District...
...which is why I emailed Rep. Latta (co-sponsor) and Rep Joyce (my rep from Ohio) and let them know that I vote, I elected one of them, and I don't support any action to reverse the FCC's recent reclassification.
I know I don't represent big bags of money, but I do directly represent a ballot. I let both of them know that I am a US Army veteran, a long-time IT professional, and a proponent of net neutrality and classifying internet service under Title II.
If you were talking about a spanned volume, I'd understand why you might be frustrated. With RAID, there's a physical relationship between the organization of the blocks of data, and the number and configuration of disks in the set.
If you have 4 disks in a stripe set, then the first 4 blocks of data go to each drive in sequence. The next 4 blocks do the same. Once you're on the billionth block, and you run out of space, you can't just add a new drive into the mix without having to rearrange all but the first 4 blocks.
For RAID levels that include mirroring, doing what you want would be technically possible.
It's certainly possible that someone could write code in the RAID firmware or software to take the mirrored drives out of the array, build a new array using them and 1 of 2 newly-added drives, copy the data from the original stripe set to the new larger one, switch to the new stripe set as the active one, add the second of the 2 newly-added drives to the old stripe set, and rebuild the updated original stripe set via mirroring, but there are more than a few potential disasters waiting to happen in that scenario.
While it might be technically possible to come up with a way to tack a disk on to a RAID 10 array, it would not be safe, and safety is one of the things that the letter R in RAID brings to the table (redundancy). If the process of adding the disks breaks redundancy (which it would have to, while it was happening), then that would be something that most folks would not be looking for as an option.
No man is an island if he's on at least one mailing list.