Re: texi2dvi failes to encode 'a' and 'e' with ogonek

[Karl Berry]

    Moreover there's eogonek instead of aogonek 

Thanks for finding my typo.  I installed that change.

    I've also observed that in the PDF index all diacritics have been
    removed.

By the index I presume you mean the bookmarks/outlines pane.  (An actual
document index out fine as far as I can see, it was in my test file ...)

Yes, that is so.  There is no good solution for that.  Another long-time
Polish TeX user (Staszek Wawrykiewicz) told me that he observed that the
Latin 2 oegonek ends up getting displayed as | sometimes, apparently
because the bookmark pane uses system fonts and the encoding is
indeterminate.  I'm not sure if that's still true, but anyway, it's way
too hard to deal with when we don't even have good font support yet.

In order to avoid garbage from being in the outline pane, we basically
wipe out all diacritics, etc.  (Technically it's the same
"ASCII-ification" we use to produce index sort strings, which is another
well-known i18n problem we aren't going to solve any time soon ...)

Best,
karl