[tlbuild] Installing the current snapshot

Angelo Graziosi angelo.graziosi at alice.it
Tue Oct 13 10:56:22 CEST 2009

Reinhard Kotucha ha scritto:
> On 13 October 2009 Angelo Graziosi wrote:
>  > Marco Pallante ha scritto:
>  > > Hi all,
>  > > 
>  > > I used the UTF-8 file encoding both in the LEGGIMI (README) file
>  > > and in the texlive-it.tex documentation source. If there is any
>  > > problem, I could switch to Latin-1.
>  > > 
>  > > Nevertheless, I can confirm this problem: when downloading, my
>  > > browser (Firefox) assumes Latin-1 encoding and doesn't show any
>  > > of the accented characters.  However, manually switching to the
>  > > UTF-8 encoding, everything goes right.
>  > 
>  > Just for the sake of completeness...
>  > 
>  > Emacs says that in the ISO image[1], the file
>  > '/cdrom/readme-txt.dir/README.IT' is encoded as 'iso-latin-1-unix'.
> Do you have a recent version of Emacs?  Emacs 23.1 assumes that the
> file encoding is Unicode.  The file contains the word "varietà".  The
> letter "à" consists of two bytes, thus, it can't be ISO-8859.

Obviously: it is a build from CVS, 2009-10-11!! :-)

> Please note that Emacs 23 has *much* better Unicode support than any
> previous release.

I know. I follow Emacs development since 22.0.50-CVS.

>  > This seems to be confirmed by Firefox and IE8, which are configured
>  > to detec automatically the encoding systems of web pages.
> No.  They don't detect anything automatically.  If there is no
> adequate HTML tag, they use a default encoding.

Those browsers, in my case, are configured to "Allow pages to choose 
their own characters...". Indeed, a few tests show that a few web pages 
are in UTF-8 and others in ...ISO-8859...

>  > Indeed, [2] is detected as UTF-8 and looks right, instead [3,4] are 
>  > detected as Western ISO-8859-1, and look wrong.
> They didn't detect ISO-8859-1.  It was an assumption.  It's almost
> impossible to detect ISO-8859-1 (or any other 8-bit encoding).

Firefox show that!!! :-)

> What does "Western ISO-8859-1" mean?  Do they maintain different
> variants of ISO-8859-1?  What is "Eastern ISO-8859-1" then?
> ISO-8859-2 or ISO-2022-JP?  :)

"Western" is the translation of "Occidentale" (I have Firefox 3.5-it, 
obviously). In my Firefox, the menu "Visualizza | Codifica dei 
caratteri" (i.e. View | Characters encoding) shows many encodings, also 
"Europa Centrale ISO-8859-2", "Occidentale ISO-8859-15"...

> Regarding character encodings, I fully agree with Norbert.  UTF-8
> allows you to have Russian, Japanese, Devanagari, Mongolian, Chinese,
> Arabic,... in one and the same file.  UTF-8 is the future.  Forget
> everything else.

As I have written, I haven't to choose explicitly that: it is 
automatically selected if the page is in UTF. :-)

It seems we are talking about different things/systems :-)

Anyway this is not very important to the release. Who does not visualize 
correctly that documentation can always change manually the encoding, 
eventually by trying.


More information about the tlbuild mailing list