[EGIT] UTF-8 in author names in history views read as ISO-8859-1

[EGIT] UTF-8 in author names in history views read as ISO-8859-1

[José Fonseca]

Hi,

First, thanks for the work on egit -- it is progressively becoming more 
and more useful.

I found that spite UTF-8 being the default encoding, egit is interpreting 
(or displaying) names as ISO-8859-1 in the history view.

BTW, something I really miss from egit is "Team > Compare With > HEAD 
revision" or something like that.

Keep up the good work!

Cheers,

José Fonseca

Re: [EGIT] UTF-8 in author names in history views read as ISO-8859-1

[Robin Rosenberg]

torsdag 27 september 2007 skrev José Fonseca:
> Hi,
> 
> First, thanks for the work on egit -- it is progressively becoming more 
> and more useful.
> 
> I found that spite UTF-8 being the default encoding, egit is interpreting 
> (or displaying) names as ISO-8859-1 in the history view.

Encoding is largeley neglected. Actually conversions are made because one
has to, but there is no consistency there. Git mostly neglect encodings and
so does anything that attempt to be compatible with it. I have some patches
towards improvement. A problem is that conversion to or from UTF-8 and back
are not reversible, especially when the input is invalid for the conversion. This
*will* happen when dealing with git repos that contain non-ascii comments or
file names.

> BTW, something I really miss from egit is "Team > Compare With > HEAD 
> revision" or something like that.

Coming.. but no plan on when as for most ideas that I have. Lack of time means
most things come in small pieces that can be completed in an hour or two. OTOH
some things become clear and simple over time when not rushing them. This is
probably one of the simple features. Actually it's a good way to get started with
hacking the plugin ;]

> Keep up the good work!

Thanks

-- robin

Re: [EGIT] UTF-8 in author names in history views read as ISO-8859-1

[Robin Rosenberg]

torsdag 27 september 2007 skrev Robin Rosenberg:
> Encoding is largeley neglected. Actually conversions are made because one
> has to, but there is no consistency there. Git mostly neglect encodings and
> so does anything that attempt to be compatible with it. I have some patches
> towards improvement. A problem is that conversion to or from UTF-8 and back
> are not reversible, especially when the input is invalid for the conversion. This
> *will* happen when dealing with git repos that contain non-ascii comments or
> file names.

A confession. The only thing we do cater for is the encoding header in commits.

-- robin