Kat's 8bit sequences
** The Idea **
Some sequences store their values using 64-bits each value others 32 bits. Yet
it is transparent to the user. See the performance note under strings. What
was done for 32 bit values could also be done for 8 bit byte values. Call them
Kat sequences. The EUPHORIA syntax wouldn't change, yet under the hood the
amount of memory used for some sequences is 1 byte per value plus sequence
overhead.
** The Need? **
When EUPHORIA was released 15 years ago when computers were lucky to have a
500 MB HARD DRIVE this wasn't a problem. It would seem less important these
days
with so much RAM. Yet, Robert Craig didn't think it was an issue then. Is it
an issue today?
War and Peace ( 3.13 MB )
The Bible ( 4.57 MB )
Don Quixote ( 2.24 MB )
[ From project Gutenburg ]
You could multiply these sizes by 10 without significantly taxing RAM. If
you are working with ASCII records. It would seem to me you only need two
in RAM at the same time, if you are doing a comparison operation of some kind.
Perhaps Kat is working for Big Brother and has everybody's DNA finger print in
Alabama on her computer.
When you move to HTML even, you still are shy of this 500 MB limit for the
longest books of all time.
Although it seems ridiculous to most of us. It may become reasonable as RAM
gets cheaper to load an entire video file into memory for editing, for example.
And It might be more convenient to edit as a stream of bytes rather than quads.
Shawn Pringle
|
Not Categorized, Please Help
|
|
