- From: David Woolley <david@djwhome.demon.co.uk>
- Date: Mon, 5 Jan 2004 07:34:18 +0000 (GMT)
- To: w3c-wai-ig@w3.org
> cognitive perspective, might it be more useful to have a character count of > how many characters have been typed, with an appropriate message, rather Typically the engineering limit is on the number of bytes, not the number of characters. Explicit UTF-8 and the typical browser recovery (that many BBS type systems now rely upon) of sending numeric entities for unencodeable characters, means that it may be necessary to change the number by more than one. I'd suggest that it is slightly more acceptable to have discontinuities in the remaining character count than to report an engineering bytes used figure and pretend it was then number of characters typed. Really, the best solution is to use data structures that are not restricted in capacity. If one doesn't want problems from variable length codes, you will have to explicitly validate each character for being in the allowed character set. Systems that impose limits can impose their own fragilities, e.g. one I have used gets confused and keeps objecting if one tries to correct an overflow by deleting characters in the middle of the textarea.
Received on Monday, 5 January 2004 02:39:45 UTC