[tex-live] Technical showstoppers for TL2003

[Vladimir Volovich]

"PO" == Petr Olsak writes:

 >> with encTeX, you put the input encoding into the format file. so
 >> the TeX engine is only able to process documents in ONE FIXED
 >> INPUT ENCODING without regeneration of the format file. do you
 >> find it elegant and useful? i don't. in my opinion, encTeX is
 >> totally useless,

 PO> This is not true! Please, if you are argumenting against encTeX,
 PO> do not write nonsenses.  Macro packages (like inputenc) can read
 PO> the encoding and is_printable setting everytime via encTeX
 PO> primitives and they can set new values. The whole engine
 PO> co-operates with the present-day TCX tables.

sorry, i got this impression after reading your manual (enctex-e.pdf):

> If your web2c distribution implements encTEX then you can initialize
> it by the -enc option in command line. You have to use this option
> during iniTEX because encTEX stores its primitives and its data to the
> format file. When the format is used, the encTEX is initialized from
> format file automatically and you need not use the -enc option
> again. If you are using a format without encTEX initialized in it and
> you write -enc option then the warning is printed and this option is
> ignored.

from reading that, i've got the impression that all reencoding is
stored in the format file and cannot be changed. so, all enctex's
primitives work in normal (non-ini) run too, and it's possible to
redefine all mappings on the fly?

 >> what we were talking about is EXPLICITLY the question whether to
 >> output some characters in 8-bit form instead of ^^-notation. we
 >> were NOT talking about any re-encoding like encTeX or TCX does.

 PO> encTeX solves the is_printable reading/setting too.  This was a
 PO> reason, why I mentioned it in this thread.

TCX solves it too, and is already in the web2c/tex live. so why do we
need enctex for that?

Best,
v.