Re: no worries
Gioele Barabucci <[email protected]>
| Newsgroups | gmane.text.tei.general |
|---|---|
| Message-ID | <[email protected]> |
Am 03.02.2017 um 00:39 schrieb Paterson, Duncan: > The other problem relates to choosing utf8 or utf16 encoding, but I > don't think double byte characters applies to Arabic. Allow me a technical consideration and advice. UTF-16 is only a useless historical artefact. It should not be used in any new project. The only sane choices are UTF-8 and (if really needed) UCS-4. UTF-16 is the worst of all encodings: it wastes bits like UCS-2/4, is incompatible with ASCII and is as computationally hard to work with as UTF-8. One should use either UTF-8 for ASCII compatibility and space savings or UCS-4 for speed of computation (but only under certain particular circumstances). Regards, -- Gioele Barabucci <[email protected]>