John Turner wrote:
> first time I've tried one of the recent versions with the new
> configure process
>
> o love the move to a more standard config/install process - would be
> very nice to have the (reasonably-standard) --exec-prefix option to
> configure to allow separation of arch-dependent and arch-independent
> pieces
Yes. Marek did some nice work with this.
>
> o one change from the previous versions (from 99.1 back as far as I
> can remember) is that now specifying -dir <foo> causes l2h to fail
> if <foo> doesn't already exist - before it would just create it and
> proceed
>
Hmm. Probably I'm being too cautious, but isn't it a bit dangerous
to allow the directory to be created automatically ?
For single jobs, it isn't much work to ensure the directory is
already available before trying to write into it with LaTeX2HTML.
In any case, it wouldn't be hard to have a -mkdir switch
(and/or $ALLOW_MKDIR variable) which overrides the safety check.
I can see how this would be useful with similar jobs being done
either together, or in a repetitive way.
> seeing a couple of other oddities that I'll save for separate msgs
Please send ASAP, as I'm trying to make the final clean-ups
before a complete (beta) release.
The latest features include:
* the failure with Perl 5.005 + threads seems to be *fixed*
(this needs some more testing; volunteers please!)
* recognising more input character-sets, in terms of 8-bit
(non-ascii) characters; windows, macroman, decmulti
as well as the various latin encodings.
* options to control the output encoding (charset)
e.g. for an accented char or math symbol either:
&#<num>; entity,
&<name>; entity,
8-bit character,
UTF-8 character pair/group,
the unaccented letter, or an image.
Full UTF8 encoding seems to work in NS4.5 and later,
as do named entities. However not all the latin2 characters
are recognised, and similarly for latin3, 4, ...
MSIE5 is better ?
...but there is a new NS on the way.
All the best,
Ross Moore
