Re: Win32 :hardcopy and UTF-8 hassles, and why Schroedinger's Cat has something to do with 'printencoding' ;-)
- Bram Moolenaar did utter on 22/12/2005 13:58:
> Timon Christl wrote:On Win32 you specify a character set for the printer font the same way
>>I wanted to share some discoveries I recently made:
>>First, printing UTF-8 on Win32:
>>On Win32, 'printencoding' is empty by default, so a UTF-8 file gets
>>printed with whatever 'encoding' is set to (which is UTF-8 on my
>>system). Printing on Win32 gvim obviously sends the text as raw text
>>(as opposed to printing through the GDI), and, since printers expect
>>some extended ASCII variant for legacy reasons when fed with data in
>>this way, UTF-8 text will always be printed wrong. It would therefore
>>be better if 'printencoding' is made available even when compiled with
>>-postscript, and that it defaulted to Latin1 on Win32 instead of being
> This simply has not been implemented yet. I hope someone looks into
> printing non-latin1 text on Win32.
you do for the display font. For example:
This will only work if the buffer character set is the same, no encoding
conversion is currently done for Win32 printing.
>>Last, the 'printencoding' oddity: Something strange is going on with
>>that option. The option ought not to be available on Win32 because of
>>-postscript (or so says the help). But doing
>>does not give an error, while it should give me E518. Doing
>>gives me E519, which is expected. But then again, doing
>>does not give me an error, while it should give me E113. When I try
>>the same game with some option name that does definitely not exist
>>(like 'foobarquux'), I get E518, E519 and E113, just as I would
>>expect. As a last note: the option does not show up in completion when
>>which is also what I would expect.
>>It seems that the existence of 'printencoding' is similar to the
>>Schroedinger's Cat thought experiment (See for example
>>http://en.wikipedia.org/wiki/Schr%C3%B6dinger%27s_cat). The option is
>>a superposition of "does exist" and "does not exist", until you
>>interfere with it, at which it collapses into one of the two states.
> See ":help hidden-options". Thus the cat is not dead or alive, it's
> hidden :-).
For a moment there it all made sense...