[pdftex] Backtick producing a wrong character code in PDF output

[Ilya Zakharevich]

On Fri, Jan 28, 2022 at 10:59:20PM +0100, Reinhard Kotucha wrote:
>  > Did you notice that I do not complain about “the shape of the glyph”?
>  > I discuss only how cut-and-paste, search etc. would behave.
>  > 
>  >    (It seems that Adobe Acrobat is special-casing font “Tt”.  Should
>  >     not PDFTeX do this instead when it marks glyphs with character
>  >     codes in PDF?)
> 
> You can't cut-and-paste to pdftex.

And I never said I do this.  Since it seems that I was not clear: I’m
discussing the following pipeline:

      A snippet of code in TeX document.
                    ↓
    PDFTeX converts it to PDF document (with glyphs — ignored below — and character codes).
    	   	    ↓
     A PDF viewer uses these character codes for copy&paste and search.

① On the first arrow: apparently, PDFTeX creates wrong character codes (by default).

② On the second arrow: Adobe Acrobat seems to have workarounds against
  these bugs in PDFTeX.  However, as the discussion I referred to
  shows, it seems that a lot of other PDF viewers take what PDFTeX
  produces LITERALLY, without workarounds.

The result: the PDF viewers do not allow copy&paste of code
fragments, and a correct search in these fragments.

> pdftex does exactly what you tell it.  No more, no less.

The developers of a PDF viewer claim otherwise.  I do not know enough
to argue with them.  Can you please take part in the discussion I mentioned?  

> The distinct behaviour is on purpose.  \verb is for computer code and
> \texttt is for normal text.

If you say that \textt is “for normal text”, then what do I use for
code fragments in (say) footnotes?

Thanks,
Ilya