>> mohrin@sharmahd.com (Torsten Mohrin):
TM> In SC UniPad we use a compressed name table. [...]
>> Why not use Huffman encoding? [...]
MC> I had a look to the Sharmahd's UniPad editor: the Unicode name and
MC> other properties for the character at the cursor are dynamically
MC> shown in the status bar as you move the cursor. So I guess that
MC> the extra expense of using more aggressive compression schemes
MC> would not be so little. Their algorithm looks like a correct
MC> economic balance between memory and computation.
Marco, we're speaking tens of microseconds per character lookup here.
It doesn't make any difference at all for interactive use (you've got
three orders of magnitude to play with, and at any rate the execution
time is dominated by the display).
J.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:02 EDT