Hi,
I have started some work on translation of documentation. I think that using utf-8 instead of latin-1 for encoding could help much translation effort. If everyone use the same encoding, it will also help dealing with special characters like & et > issues.
I also looked for a translation based on PO files.This kind of translation is really easier to keep up to date than integral translation of document.
What do you think about this ?
Jean-Denis
Jean-Denis Giguere wrote:
I have started some work on translation of documentation. I think that
using utf-8 instead of latin-1 for encoding could help much translation
effort. If everyone use the same encoding, it will also help dealing
with special characters like & et > issues.
The encoding should be appropriate for the specific locale. E.g. Alex'
Russian translations all use KOI8-R, which is the encoding which is
most commonly used in Russia (far more common than ISO-8859-5 or
UTF-8).
In spite of the hype, UTF-8 is far from being in widespread use. It's
widespread support (in libraries and applications) springs more from a
desire by programmers to avoid having to deal with encoding issues
than actual usage.
IMHO, it would be preferable to use the ISO-8859-* encodings where
possible. ISO-8859-* text can be automatically converted to UTF-8
using iconv, but the reverse isn't necessarily true.
--
Glynn Clements <glynn.clements@virgin.net>
On Thu, 29 Apr 2004 16:51:18 +0100, "Glynn Clements"
<glynn.clements@virgin.net> said:
Jean-Denis Giguere wrote:
> I have started some work on translation of documentation. I think that
> using utf-8 instead of latin-1 for encoding could help much translation
> effort. If everyone use the same encoding, it will also help dealing
> with special characters like & et > issues.The encoding should be appropriate for the specific locale. E.g. Alex'
Russian translations all use KOI8-R, which is the encoding which is
most commonly used in Russia (far more common than ISO-8859-5 or
UTF-8).In spite of the hype, UTF-8 is far from being in widespread use. It's
widespread support (in libraries and applications) springs more from a
desire by programmers to avoid having to deal with encoding issues
than actual usage.
Is it not a good reason ? I think docbook is a very good choice for
documentation. There are useful tools to translate this format. POT and
PO files are easy to manage and help to keep documentation translation up
to date. How can you deal with translation template (.POT) if each
translator work with his own encoding. It can be painful to maintain
translation if we don't have common tools.
IMHO, it would be preferable to use the ISO-8859-* encodings where
possible. ISO-8859-* text can be automatically converted to UTF-8
using iconv, but the reverse isn't necessarily true.
I never see this restriction, but it may be possible.
--
Glynn Clements <glynn.clements@virgin.net>
--
Jean-Denis Giguere
jdenisgiguere@fastmail.fm
--
http://www.fastmail.fm - Sent 0.000002 seconds ago
Jean-Denis Giguere wrote:
> > I have started some work on translation of documentation. I think that
> > using utf-8 instead of latin-1 for encoding could help much translation
> > effort. If everyone use the same encoding, it will also help dealing
> > with special characters like & et > issues.
>
> The encoding should be appropriate for the specific locale. E.g. Alex'
> Russian translations all use KOI8-R, which is the encoding which is
> most commonly used in Russia (far more common than ISO-8859-5 or
> UTF-8).
>
> In spite of the hype, UTF-8 is far from being in widespread use. It's
> widespread support (in libraries and applications) springs more from a
> desire by programmers to avoid having to deal with encoding issues
> than actual usage.Is it not a good reason ? I think docbook is a very good choice for
documentation. There are useful tools to translate this format. POT and
PO files are easy to manage and help to keep documentation translation up
to date. How can you deal with translation template (.POT) if each
translator work with his own encoding. It can be painful to maintain
translation if we don't have common tools.
The POT files should be in US-ASCII.
--
Glynn Clements <glynn.clements@virgin.net>