[
Lists Home |
Date Index |
Thread Index
]
David Lyon wrote:
> True. The old parts of Excel are written in assembly language
> by true masters. They are efficient. The CSV era was at the
> same time as the assembly language coding.
But, there were others in those days that were writing tight-code
but not using CSV... At Digital, we were using binary ASN.1 based encodings
to store spreadsheets in the 1980's since the expressiveness of CSV files
was simply too limited to allow for efficient encoding of what we wanted to
support in a spreadsheet. While there are tricks to build "expressiveness"
into a CSV, they can get pretty ugly, large and expensive to process. The
same "expressiveness" arguments that were used to justify ASN.1 encodings of
spreadsheets in the mid-1980's would make good arguments for using XML today
instead of CSV. For instance, the structure of ASN.1 (or XML) makes it easy
to do things like distinguish sheets from workbooks or to maintain multiple
versions of the data in a single file (resulting in great compression if
encoded as deltas at the cell or region level!). Also, either ASN.1 or XML
allow you to vary the types and formatting of rows in your spreadsheet in
non-uniform ways (i.e. have every cell be a different type/format
combination). That is hard to do with CSV.
Another nice thing about XML is that since its model is so similar
to ASN.1, it's real easy to go one step farther and use a binary encoding to
get efficient storage. Expressiveness when needed, compactness when needed,
a consistent data-model either way. Today, you can have it your way. Or,
tomorrow maybe...
bob wyman
|
